Hamilton College Hamilton Digital Commons Articles Works by Type 11-1999 A spin network primer Seth Major Hamilton College, [email protected] Follow this and additional works at: https://digitalcommons.hamilton.edu/articles This document is the publisher's version of an article published in: American Journal of Physics, vol. 67, no. 11 (1999): 972-980. doi: 10.1119/1.19175 Citation Information Major, Seth, "A spin network primer" (1999). Hamilton Digital Commons. https://digitalcommons.hamilton.edu/articles/188 This work is made available by Hamilton College for educational and research purposes under a Creative Commons BY-NC-ND 4.0 license. For more information, visit http://digitalcommons.hamilton.edu/about.html or contact [email protected]. A spin network primer Seth A. Major Citation: American Journal of Physics 67, 972 (1999); View online: https://doi.org/10.1119/1.19175 View Table of Contents: http://aapt.scitation.org/toc/ajp/67/11 Published by the American Association of Physics Teachers Articles you may be interested in QUANTUM MEASUREMENTS American Journal of Physics 85, 5 (2016); 10.1119/1.4967925 Sliding down an arbitrary curve in the presence of friction American Journal of Physics 85, 108 (2017); 10.1119/1.4966628 AN APOLOGY FROM THE FORMER ASSOCIATE EDITOR American Journal of Physics 85, 405 (2017); 10.1119/1.4981790 Toward an understanding of the spin-statistics theorem American Journal of Physics 66, 284 (1998); 10.1119/1.18860 On the role of self-adjointness in the continuum formulation of topological quantum phases American Journal of Physics 84, 858 (2016); 10.1119/1.4961500 Deep learning for teaching university physics to computers American Journal of Physics 85, 311 (2017); 10.1119/1.4977792 A spin network primer Seth A. Majora) Institut fu¨r Theoretische Physik, Der Universita¨t Wien, Boltzmanngasse 5, A-1090 Wien, Austria ͑Received 8 February 1999; accepted 15 April 1999͒ Spin networks, essentially labeled graphs, are ‘‘good quantum numbers’’ for the quantum theory of geometry. These structures encompass a diverse range of techniques which may be used in the quantum mechanics of finite dimensional systems, gauge theory, and knot theory. Though accessible to undergraduates, spin network techniques are buried in more complicated formulations. In this paper a diagrammatic method, simple but rich, is introduced through an association of 2ϫ2 matrices with diagrams. This spin network diagrammatic method offers new perspectives on the quantum mechanics of angular momentum, group theory, knot theory, and even quantum geometry. Examples in each of these areas are discussed. © 1999 American Association of Physics Teachers. I. INTRODUCTION methods to model space–time. As the absolute space and time of Newton is a useful construct to apply in many every- Originally introduced as a quantum model of spatial day calculations, perhaps continuous space–time is simply geometry,1 spin networks have recently been shown to pro- useful as a calculational setting for a certain regime of phys- vide a basis for the states of quantum geometry—kinematic ics. 2 states in the Hamiltonian study of quantum gravity. At their Motivated by these difficulties, Penrose constructed a dis- roots, spin networks provide a description of the quantum crete model of space. The goal was to build a consistent mechanics of two-state systems. Even with this humble foun- model from which classical, continuum geometry emerged dation, spin networks form a remarkably diverse structure only in a limit. Together with John Moussouris, he was able which is useful in knot theory, the quantum mechanics of to show that spin networks could reproduce the familiar angular momentum, quantum geometry, and other areas. three-dimensional angles of space—a ‘‘theory of quantized Spin networks are intrinsically accessible to undergradu- directions.’’ 4 In this setting, spin networks were trivalent ates, but much of the material is buried in more complex graphs labeled by spins. For applications in quantum geom- formulations or lies in hard-to-find manuscripts. This article is intended to fill this gap. It presents an introduction to the etry it is better to work with spin networks with higher va- diagrammatic methods of spin networks, with an emphasis lence vertices. on applications in quantum mechanics. In so doing, it offers These suitably generalized spin networks have been undergraduates not only a fresh perspective on angular mo- shown to form the eigenspace of operators measuring geo- 5 mentum in quantum mechanics but also a link to leading metric quantities such as area and volume. These new spin edge research in the study of the Hamiltonian formulation of network techniques arose out of a powerful suite of methods quantum gravity. One quantum operator of geometry is pre- for background-independent quantization that has been de- sented in detail; this is the operator which measures the area veloped over the past few years. Spin networks are fantasti- of a surface. cally useful both as a basis for the states of quantum geom- The history of spin networks goes back to the early sev- etry and as a computational tool. Spin network techniques enties when Penrose first constructed networks as a funda- were used to compute the spectrum of area and volume mentally discrete model for three-dimensional space. Diffi- operators.6 Spin networks, first used as a combinatorial basis culties inherent in the continuum formulation of physics led for space–time, now find uses in quantum gravity, knot 3 Penrose to explore this possibility. These difficulties come theory, and group theory. from both quantum and gravitational theory as seen from This spin network primer begins by associating 2ϫ2 ma- three examples: First, while quantum physics is based on trices with diagrams. The first goal is to make the diagram- noncommuting quantities, coordinates of space are commut- matics ‘‘planar isotopic,’’ meaning the diagrams are invari- ing numbers, so it appears that our usual notion of space ant under smooth deformations of lines in the plane. It is conflicts with quantum mechanics. Second, on a more prag- analogous to the manipulations which one would expect for matic level, quantum calculations often yield divergent an- ordinary strings on a table. Once this is completed, the struc- swers which grow arbitrarily large as one calculates physical quantities on finer and smaller scales. A good bit of machin- ture is enriched in Sec. II C to allow combinations and inter- ery in quantum field theory is devoted to regulating and sections between lines. This yields a structure which includes renormalizing these divergent quantities. However, many of the rules of addition of angular momentum. It is further ex- these difficulties vanish if a smallest size or ‘‘cutoff’’ is in- plored in Sec. III with the diagrammatics of the usual angular troduced. A discrete structure, such as a lattice, provides momentum relations of quantum mechanics. ͑A reader more such a cutoff. Thus, were space–time built from a lattice or familiar with the angular momentum states of quantum me- network, then quantum field theory would be spared many of chanics may wish to go directly to this section to see how the problems of divergences. Third, there is a hint coming spin networks are employed in this setting.͒ In Sec. IV this from general relativity itself. Since regular initial data, say a connection to angular momentum is used to give a diagram- collapsing shell of matter, can evolve into a singularity, rela- matic version of the Wigner–Eckart theorem. The article fin- tivity demonstrates that the space–time metric is not always ishes with a discussion on the area operator of quantum grav- well-defined. This suggests that it is profitable to study other ity. 972 Am. J. Phys. 67 ͑11͒, November 1999 © 1999 American Association of Physics Teachers 972 II. A PLAY ON LINE ͑2͒ This section begins by building an association between the ϫ ͑ ␦B Kronecker delta functions, the 2 2 identity matrix or A), and a line. It is not hard to ensure that the lines behave like However, these ‘‘topological’’ difficulties are fixed by modi- elastic strings on a table. The association and this require- fying the definition of a bent line. One can add an i to the ment lead to a little bit of knot theory, to the full structure of antisymmetric tensors spin networks, and to a diagrammatic method for the quan- tum mechanics of angular momentum. Since each of the two awkward features contains a pair of A. Line, bend, and loop ⑀’s, the i fixes these sign problems. However, there is one more property to investigate. The Kronecker ␦B is the 2ϫ2 identity matrix in compo- ␦D␦C⑀ ϭϪ⑀ ͑ A On account of the relation A B˜ CD ˜ AB one has the nent notation. Thus, indices C and D are added to the diagram for clarity͒ 10 ͑␦B͒ϭͩ ͪ A 01 —not what one would expect for strings. This final problem ␦0ϭ␦1ϭ ␦1ϭ␦0ϭ can be cured by associating a minus sign with each crossing. and 0 1 1 while 0 1 0. The indices A and B in this ⑀ expression may take one of two values, 0 or 1. The diagram- Thus, by associating an i with every and a sign with matics begins by associating the Kronecker ␦ with a line every crossing, the diagrams behave as continuously de- formed lines in a plane. The more precise name of this con- cept is known as planar isotopy. Structures which can be moved about in this way are called topological. What this association of curves with ␦’s and ˜⑀’s accomplishes is that it allows one to perform algebraic calculations by moving lines The position of the indices on ␦ determines the location of in a plane.