A stringy dispersion relation for field theory

This paper derives a novel, local, and crossing-symmetric dispersion relation for 2-2 scattering amplitudes motivated by string theory, demonstrating its utility in bounding Wilson coefficients for gravitational effective field theories and providing convergent series representations for Veneziano, Virasoro-Shapiro, and multi-variable generalizations.

The Gist

Imagine you are trying to describe a complex, multi-sided object, like a crystal. In physics, this object is a scattering amplitude—a mathematical recipe that tells us the probability of particles smashing into each other and bouncing off in different directions.

For decades, physicists have had two main ways to look at this crystal:

1. The "Side A" View: You look at it from the front (the $s$-channel). You see a specific pattern of poles (peaks in the data).
2. The "Side B" View: You look at it from the side (the $t$-channel). You see a different pattern of poles.

In the old days of string theory, there was a famous hypothesis called "duality." It said, "Hey, these two views are actually the same object, just described differently." But there was a problem: standard math tools (dispersion relations) could only show you one side at a time. If you tried to look at both sides simultaneously, the math would break down, or you'd have to do messy manual corrections to make it work.

Enter this paper: The "Stringy" Dispersion Relation.

The authors (Faizan Bhat, Arnab Priya Saha, and Aninda Sinha) have invented a new, magical lens. Let's break down what they did using some everyday analogies.

1. The Elastic Rubber Sheet (The Parameter $\lambda$)

Imagine the scattering amplitude is drawn on a rubber sheet.

• Traditionally, to see the "Front View," you had to stretch the sheet flat in one direction.
• To see the "Side View," you had to stretch it flat in the other direction.
• You couldn't see both at once without tearing the sheet.

The authors introduce a knob (a parameter they call $\lambda$). Think of this knob as a "stretchiness" control.

• Turn the knob one way, and the rubber sheet stretches to show you the Front View perfectly.
• Turn it the other way, and it stretches to show the Side View.
• The Magic: If you leave the knob in the middle, the sheet is in a "superposition." It shows you the Front View and the Side View simultaneously, without tearing. It reveals the poles (the peaks) of all channels at the same time.

This is what they call a "Crossing Symmetric Dispersion Relation." It's a single formula that respects the symmetry of the universe (that physics shouldn't care which way you label the particles) and works everywhere, not just in a small, restricted corner of the math.

2. The "Double-Counting" Problem Solved

In the old "interference model," trying to add the Front View and Side View together was like trying to count the same apples twice. Physicists thought, "If I sum the poles from both sides, I'm double-counting the particles!"

The authors show that with their new "elastic" formula, you can sum the poles from all directions without double-counting. It's like realizing that the "Front View" and "Side View" are just different shadows cast by the same 3D object. Their formula reconstructs the whole 3D object from the shadows, showing that the "double counting" was just an illusion caused by looking at the shadows separately.

3. The Gravity Problem (The "Graviton Pole")

One of the biggest headaches in modern physics is dealing with gravity.

• When particles interact via gravity, there is a "pole" (a singularity) at zero momentum. It's like a black hole in the math that swallows everything.
• Standard tools fail here. If you try to use the old "Front View" math to study gravity, the math explodes (diverges) because of this gravity pole.
• Previous solutions required physicists to do a complicated dance, moving into "impact parameter space" (a weird, abstract coordinate system) to avoid the explosion.

The Paper's Solution:
The new "elastic" knob ($\lambda$) acts as a regulator. It's like putting a safety net under the black hole. By turning the knob to a specific non-zero value, the math stays stable. The gravity pole is still there, but it's tamed. This allows physicists to finally derive strict "bounds" (rules) on how gravity behaves at low energies without getting lost in the math.

4. The "Universal Translator" for String Theory

The paper also shows that this new formula works beautifully for famous string theory amplitudes (like the Veneziano and Virasoro-Shapiro amplitudes).

• Old way: You had to write one series of numbers for the Front View and a totally different series for the Side View.
• New way: You get one single, elegant series of numbers that works for everything. It converges (adds up to the right answer) everywhere, even in places where the old methods failed.

It's like having a universal translator that can speak every language of the string theory crystal simultaneously, rather than needing a different translator for every corner of the room.

5. Looking Ahead: The $n$-Particle Puzzle

Finally, the authors take a first step toward the "Holy Grail": describing collisions with many particles (not just 2).

• Imagine trying to describe a collision of 5 particles. The number of angles and variables explodes.
• They propose a generalization of their "elastic sheet" idea that could handle these complex, multi-particle collisions. It's a blueprint for a future where we can map out the entire landscape of particle interactions, not just the simple 2-particle ones.

Summary

In short, this paper provides a new, flexible mathematical lens that allows physicists to see all sides of a particle collision at once.

• It fixes old math problems where different views didn't match.
• It tames the dangerous "gravity pole" that usually breaks calculations.
• It offers a single, unified way to describe string theory amplitudes that is more accurate and easier to use than previous methods.

It's a bit like discovering that the "Front," "Side," and "Top" views of a building are actually just different slices of a single, perfectly symmetrical 3D hologram, and now we have the tool to view the whole hologram at once.

Technical Summary

Here is a detailed technical summary of the paper "A stringy dispersion relation for field theory" by Faizan Bhat, Arnab Priya Saha, and Aninda Sinha.

1. Problem Statement

The paper addresses the challenge of deriving local, crossing-symmetric dispersion relations (CSDRs) for scattering amplitudes that are free of spurious non-local terms and possess a parametric ambiguity motivated by string theory.

• Context: Traditional fixed-$t$ or fixed-$s$ dispersion relations are powerful tools for constraining Effective Field Theories (EFTs) via the S-matrix bootstrap. However, they suffer from two main limitations:
  1. Restricted Convergence: They only manifest poles in one channel (e.g., $s$-channel), requiring analytic continuation to see poles in other channels, which restricts the domain of convergence.
  2. Gravitational EFTs: In theories with gravity, the presence of the massless graviton pole ($1/t$) causes the standard fixed-$t$dispersion relation to diverge in the forward limit ($t \to 0$). This prevents the derivation of bounds on Wilson coefficients using standard forward-limit techniques.
• Goal: The authors aim to construct a dispersion relation that is:
  • Manifestly crossing-symmetric (treating $s, t, u$ equally).
  • Local (free of spurious singularities).
  • Parametric (containing a free parameter $\lambda$ that interpolates between different fixed-channel limits).
  • Capable of handling the graviton pole without requiring a shift to impact parameter space.

2. Methodology

The authors derive the dispersion relations using Cauchy's residue theorem in the complex plane, employing a specific ansatz for the kinematic variables to ensure crossing symmetry.

A. Two-Channel Derivation ($s \leftrightarrow t$)

• Ansatz: They consider a 2-2 scattering amplitude $M(s, t)$ symmetric under $s \leftrightarrow t$. They introduce a parameter $\lambda$ and define a mapping for the second variable:
  $$\hat{s}_2(\sigma, s, t) = \frac{(s+\lambda)(t+\lambda)}{\sigma+\lambda} - \lambda$$
  This mapping ensures that as $\sigma \to \infty$, $\hat{s}_2 \to -\lambda$.
• Kernel Construction: By demanding crossing symmetry and the recovery of fixed-$t$ (when $\lambda = -t$) and fixed-$s$ (when $\lambda = -s$) limits, they derive a specific kernel:
  $$H^{(\lambda)}(\sigma, s, t) = \frac{1}{\sigma - s} + \frac{1}{\sigma - t} - \frac{1}{\sigma + \lambda}$$
• Resulting Relation: The amplitude is expressed as an integral over the discontinuity $A^{(s)}$ along a contour $C_1$:
  $$M(s, t) = \frac{1}{\pi} \int_{C_1} d\sigma \, H^{(\lambda)}(\sigma, s, t) A^{(s)}\left(\sigma, \frac{(s+\lambda)(t+\lambda)}{\sigma+\lambda} - \lambda\right)$$
• Higher Subtractions: For amplitudes growing faster than constants at infinity, they derive subtracted versions by expanding the integrand around $s, t = -\lambda$ or by dividing the amplitude by $(s-\alpha)^N(t-\alpha)^N$.

B. Three-Channel Derivation ($s, t, u$)

• The method is generalized to amplitudes with $s, t, u$ symmetry (e.g., Virasoro-Shapiro).
• They define shifted Mandelstam variables $s_1, s_2, s_3$ such that $\sum s_i = 0$.
• The kernel becomes:
  $$H^{(\lambda)}(\sigma, s_1, s_2, s_3) = \sum_{i=1}^3 \frac{1}{\sigma - s_i} - \frac{1}{\sigma + \lambda}$$
• The mapping for the other variables involves solving a quadratic equation derived from the condition that the kernel's singularities map to the amplitude's singularities symmetrically.

C. Application to String Amplitudes

• The authors apply these relations to the Veneziano and Virasoro-Shapiro amplitudes.
• They demonstrate that the parametric dispersion relation generates series expansions that manifest poles in all channels simultaneously and converge everywhere (for appropriate $\lambda$), unlike fixed-channel expansions which have restricted domains.

3. Key Contributions

1. Derivation of the "Stringy" CSDR:
   The paper provides a self-contained, straightforward derivation of the parametric CSDR, avoiding the complex parametrisations of earlier works. It explicitly shows how the parameter $\lambda$ acts as an "elasticity" parameter, deforming the worldsheet between $s$-channel and $t$-channel representations.

2. Resolution of the Graviton Pole Problem:
   A major breakthrough is the application of the parametric CSDR to weakly-coupled gravitational EFTs.

   • In standard fixed-$t$ approaches, the graviton pole at $t=0$ makes the forward limit ill-defined.
   • In the parametric approach, choosing a non-zero $\lambda$ effectively regulates the IR divergence. The graviton pole contribution is absorbed into the subtraction constant or regulated by the kernel, allowing for a convergent expansion around the forward limit ($t \to 0$) without needing to work in impact parameter space.

3. Extended Domain of Convergence:
   The authors prove that for amplitudes with Regge behavior (typical of string theory), the unsubtracted parametric CSDR converges for all $s$ and $t$ provided $\text{Re}(\lambda) > \alpha_0/\alpha'$. This is a significant improvement over fixed-$t$ relations, which often fail in the forward limit for high-energy string amplitudes.

4. Multi-Variable Generalization ($n$-particle):
   The paper outlines a strategy for $n$-particle scattering. They propose a totally symmetric $n$-variable dispersion relation with a kernel involving $n$ poles and a parameter $\lambda$. They explicitly check this for $n=5$ (5-particle case) and discuss the algebraic complexity of finding the roots required for the mapping, setting the stage for future work on Koba-Nielsen amplitudes.

4. Results

• Series Representations:
  • Veneziano Amplitude: Derived a series representation (Eq. 1.2 / 5.2) that converges for $\lambda > 0$ and manifests poles in both $s$ and $t$ channels. Numerical checks show this converges much faster than fixed-$t$ expansions (e.g., 99% accuracy with 10 terms vs. 500 terms for fixed-$t$).
  • Virasoro-Shapiro (Dilaton): Derived a 3-channel series representation (Eq. 5.5) that converges for $\text{Re}(\lambda) > 1$.
• Wilson Coefficient Bounds:
  • The authors successfully derived dispersive representations for Wilson coefficients ($W_{p,q}$) in gravitational EFTs.
  • Using the SDPB (Semidefinite Program Solver) package, they computed bounds on the space of meromorphic amplitudes with a graviton pole in $d=5$.
  • The results (Fig. 8) show that the bounds converge as the truncation $k_{max}$ increases, proving the viability of the method for bootstrapping gravity.
• Numerical Validation:
  • The paper includes numerical tests for one-loop box diagrams and various string amplitudes, confirming that the final results are independent of the parameter $\lambda$, while the convergence rate depends on the choice of $\lambda$.

5. Significance

• Theoretical Unification: The work bridges the gap between the "dual resonance" model (string theory) and local quantum field theory (Feynman diagrams) by providing a representation that is simultaneously crossing-symmetric, local, and capable of summing over all channels.
• Bootstrap Program Advancement: It offers a more efficient basis for the numerical S-matrix bootstrap, particularly for theories involving gravity where previous methods failed or were cumbersome.
• New Tools for EFTs: By resolving the IR divergence issue in the forward limit, this formalism opens the door to deriving rigorous bounds on higher-derivative gravitational corrections (e.g., $R^4$ terms) directly from unitarity and analyticity, without relying on specific UV completions or impact parameter limits.
• Foundation for $n$-Particle Physics: The proposed generalization to $n$-particle amplitudes represents a crucial first step toward a comprehensive dispersive framework for multi-particle scattering, a long-standing open problem in high-energy physics.

In summary, this paper introduces a robust, parametric dispersion relation that overcomes the limitations of traditional fixed-channel approaches, successfully handles gravitational singularities, and provides a powerful new tool for exploring the space of consistent quantum field theories and string theories.