[Paper Review] Three mathematical faces of SU(2) - spin networks

Andrei Tyurin|ArXiv.org|Nov 5, 2000

Noncommutative and Quantum Gravity Theories16 references3 citations

TL;DR

This paper establishes a threefold mathematical correspondence for SU(2)-spin networks by linking combinatorial spin networks to classical theta functions, Riemann surfaces, and unitary representations. It identifies SU(2)-spin networks of level $k$ with non-Abelian theta functions via a geometric realization on the unitary Schottky space, showing that spin network states embed naturally into $L^2(\mathrm{SU}(2)^g, \mathrm{d} \vec{x})^{\mathrm{Ad}_{\mathrm{diag}} \mathrm{SU}(2)}$, thus unifying discrete quantum gravity structures with continuous geometry.

ABSTRACT

Spin networks are at the core of quantum gravity. Our aim is to plug the mathematical community at large into the procedures turn to create a finite quantum theory of general relativity. For this, because of the different cultural backgraund, we would like to change the tack: to relate discrete (combinatorial) objects to the standard "contineous" geometry.

Motivation & Objective

• To bridge discrete combinatorial structures (spin networks) with continuous geometry (Riemann surfaces, Jacobians) via non-Abelian theta functions.
• To establish a mathematical framework where SU(2)-spin networks of level $k$ correspond to holomorphic sections of line bundles on Jacobians.
• To generalize the classical theory of theta functions to the non-Abelian setting using spin networks as analogs of theta characteristics.
• To provide a geometric realization of spin network states as functions on the unitary Schottky space $\mathrm{SU}(2)^g$, compatible with quantum gravity formalisms.

Proposed method

• Define an SU(2)-spin network as a trivalent graph $\Gamma$ with edges labeled by half-integer spins $j_i \in \frac{1}{2}\mathbb{Z}^+$, satisfying Clebsch–Gordan conditions at vertices.
• Introduce the concept of 'coloring' by doubling spins ($2j_i$) to treat them as integer labels, facilitating combinatorial analysis.
• Construct a map $\pi_\Gamma: J_{\Sigma_\Gamma} \to \mathrm{U}(1)^{E(\Gamma)}$ from the Jacobian of a Riemann surface to the edge group, identifying the image as a $g$-torus $T^g_-$.
• Realize spin network states $f_{\Gamma_j}$ as functions in $L^2(\mathrm{SU}(2)^g, \mathrm{d} \vec{x})^{\mathrm{Ad}_{\mathrm{diag}} \mathrm{SU}(2)}$, using the identification $uS_g = Q_\Gamma$.
• Use the doubling construction $\Gamma \# \overline{\Gamma}$ to extend spin network invariants to closed graphs, enabling consistent state assignment.
• Establish a correspondence between level-$k$ SU(2)-spin networks and Bohr–Sommerfeld tori in the Abelian case, showing $N^k_A(\Gamma) = k^g$.

Experimental results

Research questions

• RQ1 How can SU(2)-spin networks be interpreted as non-Abelian analogs of theta characteristics in the theory of theta functions?
• RQ2 What is the geometric realization of spin network states in terms of continuous manifolds such as the unitary Schottky space?
• RQ3 How do level-$k$ SU(2)-spin networks relate to holomorphic sections of line bundles on Jacobians of Riemann surfaces?
• RQ4 What is the role of the Jacobian $J_{\Sigma_\Gamma}$ and the map $\pi_\Gamma$ in connecting spin networks to classical theta functions?
• RQ5 In the Abelian case, how does the number of level-$k$ spin networks relate to the number of $k$-torsion points on a $g$-torus?

Key findings

• The space of SU(2)-spin networks of genus $g$ and level $k$ is finite, with $N^k_g$ such networks existing for each $g$ and $k$.
• Spin network states $f_{\Gamma_j}$ are embedded in $L^2(\mathrm{SU}(2)^g, \mathrm{d} \vec{x})^{\mathrm{Ad}_{\mathrm{diag}} \mathrm{SU}(2)}$, establishing a bridge to quantum gravity formalisms.
• The identification $uS_g = Q_\Gamma$ realizes the unitary Schottky space as the image of the spin network map, linking discrete and continuous geometry.
• For the Abelian case ($\mathrm{U}(1)$-spin networks), the number of level-$k$ spin networks is $k^g$, matching the number of $k$-torsion points on a $g$-torus.
• The map $\pi_\Gamma: J_{\Sigma_\Gamma} \to \mathrm{U}(1)^{E(\Gamma)}$ realizes the Jacobian as a $2g$-torus, with image $T^g_-$, and the fibration is Lagrangian.
• Bohr–Sommerfeld fibers of the real polarization $\pi_\Gamma$ correspond exactly to level-$k$ Abelian spin networks, with $|\mathrm{BS}_k(\Gamma)| = k^g$.