Parametric holomorphy? Triviality versus Duality in Sinh-Gordon
Abstract
Suppose a regularised functional integral depends holomorphically on a parameter that receives only a finite renormalization. Can one expect the correlation functions to retain the analyticity in the parameter after removal of the cutoff(s)? We examine the issue in the Sinh-Gordon theory by computing the intrinsic 4-point coupling as a function of the Lagrangian coupling \beta. Drawing on the conjectured triviality of the model in its functional integral formulation for \beta^2 > 8\pi, and the weak-strong coupling duality in the bootstrap formulation on the other hand, we conclude that the operations: ``Removal of the cutoff(s)'' and ``analytic continuation in \beta'' do not commute.
Cite
@article{arxiv.hep-th/9909170,
title = {Parametric holomorphy? Triviality versus Duality in Sinh-Gordon},
author = {M. Niedermaier},
journal= {arXiv preprint arXiv:hep-th/9909170},
year = {2007}
}
Comments
9 pages, Latex, 2 Figures; typos corrected, small B slope computed analytically