Nonperturbative Anomalous Thresholds
Miguel Correia
TL;DR
This work addresses anomalous thresholds—nonperturbative branch points that appear below normal thresholds in heavy-particle scattering—by deriving a nonperturbative framework grounded in extended unitarity and mass analyticity. In $d=2$, it yields explicit results: a simple pole at $s=a$ with a residue tied to the nonperturbative $m\,m\rightarrow m\,m$ amplitude, and a three-step construction that recovers the triangle diagram and passes a stringent check against the $E_8$ integrable model. The analysis is generalized to $d>2$, where the anomalous threshold manifests as a left-hand cut whose discontinuity matches Mandelstam's results (e.g., in $d=4$, $J=0$), and a nonperturbative double-pole formula for $M M \to M M$ is obtained in two dimensions. A nonperturbative test with the $E_8$ model reveals that overlapping singularities (such as a box diagram) are essential to reproduce the full double-pole structure. Overall, the paper provides a principled, nonperturbative derivation of Coleman-Thun-type poles and connects S-matrix bootstrap concepts with exact integrable-model data.
Abstract
Feynman diagrams (notably the triangle diagram) involving heavy enough particles contain branch cuts on the physical sheet - anomalous thresholds - which, unlike normal thresholds and bound-state poles, do not correspond to any asymptotic $n$-particle state. ``Who ordered that?" We show that anomalous thresholds arise as a consequence of established S-matrix principles and two reasonable assumptions: unitarity below the physical region and analyticity in the mass. We find explicit nonperturbative formulas for the anomalous threshold singularity and test them against the Coleman-Thun poles of the exactly solvable $E_8$ integrable model.
