SciPost Submission Page
OffShell Strings II: Black Hole Entropy
by Amr Ahmadain, Aron C. Wall
Submission summary
Authors (as registered SciPost users):  Amr Ahmadain 
Submission information  

Preprint Link:  scipost_202308_00048v1 (pdf) 
Date submitted:  20230831 12:44 
Submitted by:  Ahmadain, Amr 
Submitted to:  SciPost Physics 
Ontological classification  

Academic field:  Physics 
Specialties: 

Approach:  Theoretical 
Abstract
In 1994 , Susskind and Uglum argued that it is possible to derive the BekensteinHawking entropy $A / 4 G_N$ from string theory. In this article we explain the conceptual underpinnings of this argument, while elucidating its relationship to induced gravity and ER=EPR. Following an offshell calculation by Tseytlin, we explicitly derive the classical closed string effective action from sphere diagrams at leading order in $\alpha^{\prime}$. We then show how to use this to obtain black hole entropy from the RG flow of the NLSM on conical manifolds. (We also briefly discuss the more problematic "open string picture" of Susskind and Uglum, in which strings end on the horizon.) We then compare these offshell results with the rival "orbifold replica trick" using the onshell $\mathbb{C} / Z_N$ background, which does not account for the leading order BekensteinHawking entropy  unless perhaps tachyons are allowed to condense on the orbifold. Possible connections to the ER=EPR conjecture are explored. Finally, we discuss prospects for various extensions, including prospects for deriving holographic entanglement entropy in the bulk of AdS.
Current status:
Submission & Refereeing History
You are currently on this page
Reports on this Submission
Report 1 by Lorenz Eberhardt on 2023927 (Invited Report)
Report
The authors make a brave attempt at one of string theory's most notorious problems: to give a general direct derivation of Hawking's black hole entropy formula $S_\text{BH}=\frac{A}{4G}$.
The paper does not contain major new ideas, but instead clarifies several existing approaches and puts them into perspective, as well as discusses their strength and weaknesses. The literature on this subject is filled with vague and partially contradictory statements and the authors do a great job of informing the reader about their validity from a modern point of view.
The main approach that the authors pursue is a stringy uplift of the derivations of black hole entropy from the gravitational path integral. This entails computing the leading contribution of order $\frac{1}{G}$ to the string partition function, which is captured by the sphere in perturbative string theory. Computing the sphere partition function of a string worldsheet theory is a subtle problem and the authors use the technology developed in their previous paper which in turn is based on old papers of Tseytlin. In those papers, it is proposed to allow offshell string backgrounds, i.e. string backgrounds that do not satisfy the equations of motion and which are not represented by CFTs on the worldsheet. This requires the introduction of a UVcutoff $\varepsilon$ on the worldsheet and the effect of dividing by the volume of the Moebius group is roughly realized in the offshell approach by taking a derivative with respect to $\log \varepsilon$. The authors use this prescription and demonstrate that one can derive the full offshell spacetime effective action order by order in $\alpha'$ from the formalism. They carry this program explicitly out to first order in $\alpha'$ (which already involves 2loop computations in the dilaton sector).
I greatly appreciate that the authors attack such a difficult problem in string theory. Even though I would argue that their main contribution is to flesh out and compare various partially existing proposals and computations in the literature, I think that this paper is a very valuable resource for the community. It is a bit anticlimactic that the computation is essentially reduced from a string theory computation to a gravity computation and is thus not `inherently stringy'. In particular these methods will presumably not give new insight into the nature of black hole microstates in quantum gravity. I also thought that although one of the main goals of the paper was to clarify Susskind's \& Uglum's open string picture, they did not add much more to it.
I think that this paper is suitable for publication in SciPost.
I have some mostly minor remarks:
1. Eq. (8): I don't understand the last step, why is it still $z$dependent? Is this some sort of zero mode of $\log G(X(z))$?
2. The first term in eq. (10) should presumably read $\frac{1}{4\pi \epsilon^2}$.
3. Page 9, last paragraph of section III: Please provide a reference for the mentioned renormalization theorem (the same appears again in the discussion section VI.B without reference).
4. Please also provide a reference for eq. (45). In what sense are the expectation value brackets around $\frac{A}{4G}$ to be interpreted?
5. Eq. (53): $t \to \tau$
6. Paragraph after eq. (60): The claim that RG flow on the worldsheet corresponds to a Ricci flow in the target manifold should be explained better. I would have expected a generalized version of a Ricci flow, since there is also a dilaton field that flows (the authors put the Bfield to zero by assumption). It should also be mentioned that this statement is only true to leading order in $\alpha'$, there would be corrections at higher orders. It should perhaps also be explained that this statement is true in the string frame, not the Einstein frame. This distinction is never made in the article.
7. Eq. (63): Please think about using another variable than $\beta$, as it can be very easy to mix up with the inverse temperature.
8. Eq. (65): I think the biggest issue with such a factorization is that once we take back reaction into account, the inner and outer regions cannot be defined anymore since the horizon itself fluctuates. Thus (65) can at best only hold approximately in the limit of weak string coupling.
9. The explanation before eq. (68) is in my view incorrect or misleading. The embedding of the worldsheet into Euclidean target space $X: \Sigma \longrightarrow \mathfrak{M}$ is always a continuous map (but not necessarily differentiable). The horizon $H \subset \mathfrak{M}$ is a closed subset in the Euclidean spacetime. Thus by simple topology $X^{1}(H)$ is closed and hence compact (since $\Sigma$ is compact). Since $\Sigma$ intersects $H$ generically in a number of points, $X^{1}(H)$ consists of a number of points and is thus by compactness finite. Thus the worldsheet intersects the horizon actually only a finite number of times.
Consequently, I also think that the following discussion about adding stiffness terms is moot. (Incidentally the inconsistent use of $m$ and $n$ around eqs. (69) and (70) is also confusing to the reader.)
10. Section V.A: The argument that the sphere contribution in the onshell approach to the orbifold replica trick vanishes is unconvincing to me. The authors point to their previous paper, section II.B for this. The argument their is a supergravity argument which shows that for smooth and compact target spaces, the sphere partition function vanishes since the action is onshell a total derivative. This argument does not apply to the orbifold, since the orbifold is neither smooth nor compact. As far as I am aware it is not known how to compute the sphere contribution and it is currently unknown whether it is zero or nonzero. The only reliable argument for a vanishing sphere partition function that I am aware of only works at $N=1$, where at least for the superstring, spacetime supersymmetry requires a vanishing onshell action. I urge the authors to be more honest about the comparison. It is fair to say that the onshell orbifold approach is not able to produce quantitative results for the sphere contribution at the moment.
11. The first sentence of the last paragraph on the left column of page 17 presumably has a typo and I don't understand what the authors want to say.
Requested changes
Please address my remarks.