Bouncing geodesics, black hole singularities, and singularities of thermal correlators

This paper utilizes Hadamard theory to establish a rigorous holographic framework linking bouncing geodesics and curvature singularities to the divergence of thermal correlators, while also demonstrating through explicit examples that the absence of bouncing geodesics does not necessarily preclude the existence of black hole singularities.

The Gist

Imagine the universe as a giant, complex video game. In this game, there are two layers: the World (the inside of the game, where gravity and black holes live) and the Screen (the edge of the game, where we, the observers, live and watch the action).

This paper is about trying to understand the most dangerous, broken parts of the game world—Black Hole Singularities—by only looking at the data on the screen.

Here is the story of what the authors discovered, explained in simple terms.

1. The Problem: The "Glitch" in the Code

In physics, a black hole has a center called a singularity. This is like a "glitch" in the universe's code where the rules of geometry break down, and things get infinitely dense. We can't go there to see it; it's too dangerous.

However, thanks to a magical rule called Holography, everything that happens inside the black hole is secretly encoded in the data on the "Screen" (the boundary of the universe). The scientists wanted to know: Can we look at the data on the screen to tell if there is a glitch (singularity) inside?

2. The Old Idea: The "Bouncing Ball" Detective

For a while, physicists had a clever detective tool. They imagined throwing a ball (a particle) into the black hole.

• If the black hole is "normal," the ball falls in and hits the center, disappearing forever.
• But in some black holes, the gravity near the center is so weird that the ball hits the singularity and bounces back out, like a rubber ball hitting a trampoline.

These "bouncing balls" leave a specific fingerprint on the Screen. If you see this fingerprint, you know there is a singularity inside. This was the old theory: No bounce = No singularity.

3. The New Discovery: The "Hadamard Flashlight"

The authors of this paper decided to shine a very bright, mathematical flashlight (called Hadamard theory) on this problem. They wanted to prove exactly why the bouncing ball leaves a fingerprint.

The Big Rule They Found:
They proved that whenever two points on the Screen can be connected by a path that touches the singularity and bounces back, the data on the Screen will "crash" (become infinite).

Think of it like this: If you shout at a wall, and the sound bounces off a broken piece of the wall and comes back to you, your ears hear a weird echo. The authors proved that this "echo" (a mathematical singularity) must exist if the path exists. It doesn't matter how heavy the ball is; if the path exists, the echo exists.

4. The Twist: The "Fake Glitch" (The Phantom)

Here is where the paper gets really interesting. They asked: Is the opposite true? If we see the echo, does it mean there is a real singularity?

They tested this with two specific types of black holes:

1. The BTZ Black Hole: A simple, 3D black hole.
2. The Linear Axion Black Hole: A 4D black hole that looks very similar to the first one but has a real curvature singularity (a true glitch).

The Surprise:

• In the BTZ case, there is no real singularity, and there is no bouncing ball. No echo. (Everything works as expected).
• In the Linear Axion case, there IS a real singularity. But... there is NO bouncing ball.

The Metaphor:
Imagine a room with a broken floor (the singularity).

• In the first room, if you drop a ball, it falls through the hole and stops. It never bounces.
• In the second room, the floor is also broken, but the "gravity" of the hole is shaped differently. If you drop a ball, it falls in, but it gets stuck or slows down so much it never actually hits the bottom and bounces back.

The Conclusion:
The authors found that just because a black hole has a singularity, it doesn't mean a "bouncing ball" path exists. Therefore, the "bouncing ball" test is not a perfect detector. You can have a broken universe (a singularity) without the specific "echo" on the screen.

5. The "Phantom" Echoes

The paper also looked at the data in a different way (using "momentum space" instead of "position space"). They found that sometimes the math predicts "echoes" that look like they should be there, but when you translate them back to real space, they disappear.

They call these "Phantom Singularities."

• Analogy: It's like seeing a reflection in a mirror that looks like a monster. You get scared, but when you walk over to the mirror, there's nothing there. The "monster" was just a trick of the light (or in this case, a trick of the math integration).

Summary: What Does This Mean for Us?

1. We have a better flashlight: We now have a rigorous mathematical proof (Hadamard theory) that explains exactly when and why black holes leave "echoes" on the boundary of the universe.
2. The old test isn't perfect: We used to think "If we see the echo, there's a singularity. If we don't, there isn't." The authors proved this is wrong. You can have a singularity without the echo.
3. The universe is tricky: Some black holes hide their broken centers so well that even our best mathematical "bouncing ball" probes can't detect them.

In short, the paper tells us that while the "bouncing geodesic" (the bouncing ball) is a useful tool, it's not the whole story. The universe is more complex, and sometimes the most dangerous glitches leave no trace on the surface at all.

Technical Summary

1. Problem Statement

The paper addresses the challenge of diagnosing and classifying spacetime curvature singularities (specifically inside black holes) using the dual boundary theory in the context of the gauge/gravity duality (holography).

• Context: While the physics of event horizons (Hawking radiation) is well-understood, the nature of spacetime singularities remains a central open problem. In holography, the event horizon maps to a thermal state in the Conformal Field Theory (CFT), but the imprint of the bulk singularity on CFT observables (like thermal correlation functions) is difficult to extract.
• The "Bouncing Geodesic" Hypothesis: Previous research suggested that "bouncing geodesics"—null limits of spacelike geodesics that approach a curvature singularity, turn around, and return to the boundary—leave distinct imprints (singularities) in boundary correlation functions.
• Key Questions:
  1. Why do these singularities exist for operators of any conformal dimension ($\Delta$), not just in the geodesic (large mass) limit?
  2. Can these singularities be identified on the principal sheet of the physical retarded propagator without complex analytic continuations?
  3. Do bouncing geodesics exist in general spacetimes (including non-AdS)?
  4. Is the existence of a bouncing geodesic (and its associated singularity) strictly equivalent to the existence of a curvature singularity?

2. Methodology

The authors employ a rigorous mathematical framework based on the Hadamard theory of hyperbolic differential equations to analyze the analytic structure of propagators.

• Hadamard Theorems: The core methodology relies on Theorem 1 (Hadamard), which describes the local singularity structure of the retarded Green's function $G(X, Y)$ for a scalar field in a $D$-dimensional spacetime.
  • For odd $D$: $G \sim W(X,Y) / L^{D-2}$, where $L$ is the geodesic distance.
  • For even $D$: $G \sim U(X,Y)\delta(L^2) + Q(X,Y)$.
  • Crucial Insight: The singularities of the Green's function occur if and only if the points $X$ and $Y$ can be connected by a null geodesic ($L=0$).
• Holographic Renormalization: The authors analyze how the bulk-to-boundary limit and holographic renormalization affect the location and power of these singularities. They demonstrate that while renormalization changes the power of the singularity (depending on $\Delta$), it does not alter the location determined by the null geodesic condition.
• Momentum Space Analysis: The paper investigates the relationship between singularities in momentum space (fixed $k$) and position space (fixed $x$). They utilize Quasinormal Modes (QNMs) and Fourier integrals to distinguish between "real" singularities and "phantom" singularities that cancel out upon integration over spatial momentum.
• Counter-Example Construction: The authors construct explicit counter-examples using specific gravity models (BTZ black hole and the self-dual linear axion model) to test the equivalence between curvature singularities and bouncing geodesics.

3. Key Contributions and Results

A. Rigorous Proof of Singularity Existence (Theorems 1 & 2)

The authors prove that in any holographic theory, the retarded thermal propagator has a singularity whenever two boundary points can be connected by a null geodesic passing through the bulk.

• Significance: This establishes that "bouncing singularities" are present for operators of any dimension $\Delta$, not just in the geodesic limit.
• Principal Sheet: Unlike the Wightman function (which requires analytic continuation to non-physical sheets to see these singularities), the bouncing singularities appear directly on the principal sheet of the retarded propagator.

B. General Characterization of Bouncing Geodesics

The paper provides a general definition and existence criterion for bouncing geodesics:

• Definition: A null limit of a spacelike geodesic that approaches a curvature singularity, turns, and returns to a finite distance.
• Existence Condition: For spacetimes with a blackening factor $f(r) \sim \pm b/r^a$ near the singularity ($r \to 0$), a bouncing geodesic exists if the potential allows the geodesic to turn back. Specifically, if $f(r) \to -\infty$ (as in Schwarzschild), spacelike geodesics bounce; if $f(r) \to +\infty$ (as in Reissner-Nordström), timelike geodesics bounce.
• Result: Any spacetime where the blackening factor diverges at the singularity leads to a bouncing singularity in the correlator.

C. The "Phantom Singularity" Phenomenon

A major finding is the distinction between singularities in momentum space ($t, k$) and position space ($t, x$).

• Phantom Singularities: Singularities that appear in the complex time plane at fixed momentum $k$ but are eliminated by the Fourier integration over $k$ when moving to position space $x=0$.
• Mechanism: These often correspond to piecewise null geodesics (glued segments) rather than true smooth geodesics. The Hadamard theorem dictates that only true null geodesics generate singularities in the physical propagator.

D. Counter-Example: Self-Dual Linear Axion Model

The authors answer Question 4 (Equivalence of singularity and bouncing geodesic) with a definitive No.

• The Model: They analyze a 4D black hole solution in the self-dual linear axion model.
• Properties: This black hole possesses a genuine curvature singularity (Ricci scalar diverges as $1/r^2$) but has a blackening factor identical to the BTZ black hole ($f(r) = r^2 - r_0^2$).
• Result: Because the potential well at the singularity is finite (unlike the infinite well in Schwarzschild), no bouncing geodesics exist for this geometry. Consequently, the dual retarded correlator has no bouncing singularities, despite the presence of a curvature singularity.
• Implication: The existence of a curvature singularity is not sufficient to guarantee the existence of bouncing geodesics or their associated correlator singularities.

4. Significance and Implications

1. Diagnostic Tool Refinement: The paper clarifies that while bouncing geodesics are a powerful probe for many black hole singularities (like Schwarzschild), they are not a universal diagnostic. The absence of a bouncing singularity in a correlator does not necessarily imply the absence of a curvature singularity.
2. Rigorous Framework: By grounding the discussion in Hadamard theory, the authors move beyond heuristic WKB approximations, providing a mathematically rigorous explanation for why singularities appear on the principal sheet of the retarded propagator for all $\Delta$.
3. Momentum vs. Position Space: The work highlights the danger of inferring position-space physics solely from momentum-space pole structures (QNMs). "Phantom" singularities in $k$-space can be artifacts of the integration measure and do not always correspond to physical features in $x$-space.
4. Beyond AdS/CFT: The Hadamard theorems apply to any Lorentzian manifold, suggesting that the link between null geodesics and propagator singularities is a fundamental property of quantum field theory in curved spacetime, applicable even in asymptotically flat or de Sitter spacetimes where a dual CFT may not exist.

Conclusion

The paper successfully demystifies the relationship between bulk geometry and boundary correlators. It proves that null-connectability is the necessary and sufficient condition for singularities in retarded propagators. However, it crucially demonstrates that the existence of a curvature singularity does not automatically imply the existence of a null-connectable path (bouncing geodesic), thereby limiting the scope of using bouncing geodesics as a universal probe for black hole interiors. The authors identify the "self-dual linear axion" black hole as a concrete example where a curvature singularity exists without a corresponding bouncing singularity in the dual field theory.