Exact solution of D = 1 Kazakov-Migdal induced gauge theory

Michele Caselle, A. D’Adda, Stefano PanzeriPhysics 1992 · DOI · arXiv

Abstract

This paper presents the first exact solution for the Kazakov-Migdal (KM) induced gauge theory on a one-dimensional compactified lattice. By solving the model for an arbitrary number of colors ($N$) and lattice sites ($S$), the authors demonstrate that the theory provides a rigorous description of the vortex-free sector of 1D bosonic strings. The solution establishes a mathematical bridge between strong and weak coupling phases, offering a unified prescription for handling the instabilities inherent in string theory's "upside-down oscillator" regime.

Introduction

In the early 1990s, Kazakov and Migdal proposed a novel approach to Yang-Mills theory: instead of defining the gauge field directly, it could be "induced" by massive scalar fields in the adjoint representation. The original motivation was to find a more tractable path toward solving Quantum Chromodynamics (QCD) in four dimensions. However, early investigations primarily focused on the "large $N$ limit," where the number of colors approaches infinity. To fully understand the theory's phase transitions and its deep connection to string theory, exact solutions for finite $N$ are required.

This study addresses the simplest case—a one-dimensional lattice—to determine if the KM model can reproduce known results from string theory and to observe its behavior across different coupling regimes. The result is a complete mathematical derivation that confirms the KM model's utility as a proxy for the singlet sector of compactified bosonic strings.

Theoretical Framework and Methodology

The KM model on a 1D lattice consists of $S$ sites and $S$ links, with scalar fields $\Phi_s$ at each site and gauge fields $U_s$ on each link. The partition function is defined by integrating over these fields. The authors employ several sophisticated mathematical techniques to reduce this complex integral into a solvable form:

  1. HCIZ Formula: The Harish-Chandra-Itzykson-Zuber integral is used to integrate out the gauge fields (unitary matrices). This reduces the problem from a matrix integral to an integral over the eigenvalues of the scalar fields.
  2. Permutation Group Theory: The partition function is reformulated as a sum over permutations of $N$ objects. By decomposing these permutations into their cyclic representations, the authors simplify the combinatorial complexity of the system.
  3. Grand-Canonical Mapping: The authors introduce a fugacity-like variable ($\theta$) to transform the sum over permutations into a grand-canonical partition function. This effectively maps the system onto a set of $N$ non-interacting fermions, a standard technique in matrix model physics.
  4. Analytic Continuation: The original integral converges only when the scalar mass $m > 1$ (the strong coupling regime). To explore the physically significant "weak coupling" regime ($m < 1$), the authors utilize analytic continuation. This allows them to reach the "upside-down oscillator" phase, which is central to the study of 2D gravity and string theory.

Key Insights

Space Independence

A significant finding of the paper is that the core information of the partition function is essentially space-independent. While the model is defined on $S$ sites, the results for any $S$ can be rescaled to the $S=1$ case. This suggests that the vacuum density of the eigenvalues does not depend on the lattice size, simplifying the continuum limit significantly.

The Vortex-Free Connection

The 1D KM model specifically describes the "singlet" or vortex-free sector of compactified bosonic strings. In higher dimensions, the presence of "vortices" (singularities) often complicates the mathematical treatment. The KM model naturally filters these out in one dimension, providing a clean environment to study the underlying string dynamics.

Phase Transitions as Oscillator Dynamics

The transition at the critical mass $m_c = 1$ is interpreted as a transition between standard harmonic oscillators (stable) and "upside-down" oscillators (unstable/imaginary frequency). The KM model provides a rigorous prescription for handling this instability: the properties of the unstable phase can be derived from the stable phase via the substitution of frequency $\omega \to i\omega$, validated by the authors' analytic continuation.

Results

The primary result of the paper is the derivation of a closed-form expression for the partition function $Z^{(N)}$. For a system with $N$ colors, the partition function is expressed as:

$$Z^{(N)}(q) = \frac{q^{N^2/2}}{\prod_{k=1}^N (1-q^k)}$$

where $q$ is a function of the mass and the number of lattice sites. This elegant $q$-series result allows for the calculation of all thermodynamic properties of the model.

Furthermore, the authors prove that as the number of sites $S \to \infty$, the lattice model perfectly converges to the continuum string theory results previously obtained by Boulatov and Kazakov. This convergence validates the KM model as a robust discretization of 1D string theory.

Limitations and Future Directions

While the exact solution for $D=1$ is a significant milestone, the authors note several limitations:

  • Dimensionality: The reduction to eigenvalues is specific to $D=1$. In higher dimensions, vortex contributions become essential, and the mathematical simplicity of the 1D case is lost.
  • Singlet Sector Limitation: The model captures the vortex-free sector but does not account for the physics of non-singlet sectors of the 1D string.

Despite these limitations, the method of integrating over matrix fields to obtain a partition function of gauge fields only is proposed as a more robust way to study the transition between weak and strong coupling. Future work may involve extending these techniques to more complex lattice geometries or exploring the inclusion of non-singlet states.

Summary updated

BibTeX
@inproceedings{caselle1992exact,
  author = {Michele Caselle and A. D’Adda and Stefano Panzeri},
  title = {{Exact solution of D = 1 Kazakov-Migdal induced gauge theory}},
  booktitle = {Physics},
  year = {1992},
  doi = {10.1016/0370-2693(92)91496-v},
}
Figure