[Paper Review] All Loop Scattering As A Counting Problem

Nima Arkani–Hamed, H. Robert Frost+3 more|arXiv (Cornell University)|Sep 27, 2023

Advanced Mathematical Theories and Applications4 citations

TL;DR

This paper introduces a novel combinatorial framework for computing all-loop scattering amplitudes in a theory of colored scalar particles with cubic interactions, replacing Feynman diagrams with a counting problem over curves on a Riemann surface. The key result is a curve integral formulation where amplitudes are determined by solving a system of nonlinear equations for variables $ u_C $, which encode binary geometry and compactify Teichmüller space, yielding amplitudes that factorize into tree and loop components at large $ n $ and satisfy novel differential equations.

ABSTRACT

This is the first in a series of papers presenting a new understanding of scattering amplitudes based on fundamentally combinatorial ideas in the kinematic space of the scattering data. We study the simplest theory of colored scalar particles with cubic interactions, at all loop orders and to all orders in the topological 't Hooft expansion. We find a novel formula for loop-integrated amplitudes, with no trace of the conventional sum over Feynman diagrams, but instead determined by a beautifully simple counting problem attached to any order of the topological expansion. These results represent a significant step forward in the decade-long quest to formulate the fundamental physics of the real world in a radically new language, where the rules of spacetime and quantum mechanics, as reflected in the principles of locality and unitarity, are seen to emerge from deeper mathematical structures.

Motivation & Objective

• To reformulate all-loop scattering amplitudes in a theory of colored scalar particles without relying on Feynman diagrams or spacetime-based formalisms.
• To identify a fundamental counting problem over curves on a surface that determines loop-integrated amplitudes at all orders in the topological ’t Hooft expansion.
• To derive a new geometric and algebraic structure—defined by $ u_C $-variables satisfying $ u_C + extstyleigprod_D u_D^{n(C,D)} = 1 $—that compactifies Teichmüller space and encodes scattering amplitudes.
• To demonstrate that amplitudes factorize into tree and loop components at large $ n $, enabling reconstruction from low-point amplitudes.
• To uncover a new family of differential equations in kinematic space that generate efficient recursion relations for all-loop amplitudes.

Proposed method

• Define a curve integral formulation where each curve $ C $ on a surface is associated with a variable $ u_C $, constrained by the nonlinear equation $ u_C + extstyleigprod_D u_D^{n(C,D)} = 1 $, with $ n(C,D) $ the intersection number.
• Work in the positive region where $ 0 eq u_C eq 1 $, ensuring compactness and invariance, leading to a natural compactification of Teichmüller space.
• Construct the amplitude as a sum over curve configurations, with integrand defined via headlight functions $ eta_C $ and $ eta_{C'} $, derived from tropicalizations of polynomial matrices.
• Use the global Schwinger formula and Mirzakhani kernel insertion to derive a forward-limit-like formula for 1-loop amplitudes, expressing them as sums of disk amplitudes with linearized propagators.
• Apply recursion relations (e.g., Eq. 330) to evaluate integrands and recover known results like the sunrise diagram with a $ 1/3 $ factor from automorphism counting.
• Compute tropicalizations $ f_n = ext{Trop}(F_n) $, $ g_n = ext{Trop}(G_n) $, and derive headlight functions $ eta_n $ from differences of tropicalized polynomials.

Figure 1 : Fat graphs at tree-level, 1-loop single trace, 1-loop double trace, and 2-loop single trace, respectively.

Experimental results

Research questions

• RQ1 Can scattering amplitudes at all loop orders be formulated without reference to spacetime, Hilbert space, or Feynman diagrams?
• RQ2 Is there a combinatorial counting problem over curves on a surface that fully determines loop-integrated amplitudes in a non-supersymmetric, non-planar theory?
• RQ3 How do the $ u_C $-variables, defined by nonlinear equations, give rise to a compactification of Teichmüller space that is physically meaningful?
• RQ4 Can the amplitudes be reconstructed from low-point tree and loop amplitudes in the large $ n $ limit, and what is the structure of this factorization?
• RQ5 Do the new curve integral formulas generate a natural family of differential equations in kinematic space that yield efficient recursion relations for amplitudes?

Key findings

• The curve integral formulation yields a complete all-loop amplitude formula without any trace of Feynman diagrams, replacing them with a counting problem over curves.
• The $ u_C $-variables satisfy $ u_C + extstyleigprod_D u_D^{n(C,D)} = 1 $, and in the positive region, $ 0 eq u_C eq 1 $, defining a compact, invariant geometry that realizes a new compactification of Teichmüller space.
• For the 1-loop sunrise diagram, the curve integral reproduces $ I_ ext{sunrise} = rac{1}{3} rac{1}{X_{0/1}X_{1/0}X_{1/1}} $, matching the known result and correctly accounting for the $ | ext{Aut}( ext{graph})| = 3 $ symmetry factor.
• The 1-loop planar amplitude is expressed as a sum over $ n $ disk amplitudes with linearized propagators: $ I_{ ext{1loop}} = rac{1}{3} rac{1}{X_{0i}} ext{sum} $, derived via insertion of the Mirzakhani kernel.
• The tropicalized polynomials $ f_n $ and $ g_n $ yield explicit expressions: $ f_n = ext{max}(0, (n+1)x, (n+1)x + ny) $ for $ n eq 0 $, and $ g_n = ext{max}(0, -(n+1)x, -(n+1)x - ny) $ for $ n < 0 $, which define the headlight functions.
• The headlight functions $ eta_n $ are derived as $ eta_n = -f_n + 2f_{n-1} - f_{n-2} $ for $ n eq 0 $, and similarly for negative $ n $, enabling the full integrand construction.

Figure 2 : A tree fat graph with momenta assigned to all edges.