[Paper Review] The c=1 String Theory S-Matrix Revisited

Bruno Balthazar, Victor A. Rodriguez+1 more|arXiv (Cornell University)|May 19, 2017

Black Holes and Theoretical Physics3 citations

TL;DR

This paper revisits the S-matrix of c=1 string theory by numerically computing tree-level and genus-one amplitudes using Liouville theory and Virasoro conformal blocks. It resolves the longstanding puzzle of why resonance momentum analytic continuation works for 1→3 but not 2→2 scattering, and confirms the matrix model S-matrix at genus one via numerical integration of DOZZ structure constants and conformal blocks, achieving full agreement with matrix model predictions including unitarity constraints.

ABSTRACT

We revisit the perturbative S-matrix of c=1 string theory from the worldsheet perspective. We clarify the origin of the leg pole factors, the non-analyticity of the string amplitudes, and the validity as well as limitations of earlier computations based on resonance momenta. We compute the tree level 4-point amplitude and the genus one 2-point reflection amplitude by numerically integrating Virasoro conformal blocks with DOZZ structure constants on the sphere and on the torus, with sufficiently generic complex Liouville momenta, and find agreement with known answers from the c=1 matrix model.

Motivation & Objective

• To resolve the inconsistency in analytic continuation methods used to compute 2→2 scattering amplitudes in c=1 string theory.
• To systematically compute the c=1 string S-matrix beyond tree level using exact Liouville theory and conformal block integration.
• To verify the equivalence of the string theory S-matrix with the c=1 matrix model at genus one, particularly for the 1→1 reflection amplitude.
• To clarify the origin of 'leg pole factors' in the S-matrix as arising from proper normalization of vertex operators in the c=25 Liouville theory.
• To numerically validate the unitarity of the S-matrix by confirming the real part of the genus-one 1→1 amplitude from the tree-level 1→2 amplitude.

Proposed method

• Formulates the c=1 string amplitude as an integral over internal Liouville momenta of Virasoro conformal blocks multiplied by DOZZ structure constants.
• Employs Zamolodchikov's recursive representation of sphere 4-point conformal blocks for efficient numerical evaluation of the tree-level 4-point amplitude.
• Applies generalized recursive conformal block representations to compute torus 2-point blocks in both OPE and necklace channels for genus-one amplitudes.
• Performs numerical integration over moduli space by patching OPE and necklace channel results near singularities, with careful contour deformation and residue tracking during analytic continuation.
• Uses modular covariance checks and fitting functions in the large τ₂ regime to accurately compute the τ-integral over the fundamental domain.
• Verifies consistency between OPE and necklace channel computations by including residue contributions when poles cross integration contours during analytic continuation of ω.

Experimental results

Research questions

• RQ1 Why does the resonance momentum analytic continuation method correctly reproduce the 1→3 amplitude but fail for the 2→2 amplitude in c=1 string theory?
• RQ2 How can the full S-matrix of c=1 string theory be systematically computed for general momenta beyond the resonance regime?
• RQ3 What is the precise origin of the 'leg pole factors' in the S-matrix, and how do they relate to vertex operator normalization in the c=25 Liouville theory?
• RQ4 To what extent does the genus-one 1→1 reflection amplitude in c=1 string theory agree with the matrix model prediction, and how is unitarity preserved?
• RQ5 How can numerical integration of Liouville correlators be made stable and accurate, especially near singularities in moduli space?

Key findings

• The paper explains that leg pole factors arise from proper normalization of vertex operators in the c=25 Liouville theory, resolving a long-standing ambiguity in the S-matrix formulation.
• The 1→3 amplitude computed via analytic continuation at resonance momenta is correctly reproduced, confirming the known matrix model result.
• The 2→2 amplitude cannot be recovered via resonance momentum methods because the analytic structure of the Liouville correlator does not match the piecewise-analytic structure of the matrix model amplitude.
• The genus-one 1→1 reflection amplitude is numerically computed and found to agree with the matrix model prediction, including the imaginary part fixed by perturbative unitarity.
• The real part of the genus-one 1→1 amplitude is confirmed to be fixed by the tree-level 1→2 amplitude via unitarity, providing a nontrivial consistency check.
• Numerical consistency is verified between OPE and necklace channel computations, including residue contributions when poles cross integration contours during analytic continuation of ω.