1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2: %% ws-procs975x65.tex   :   10 October 2003

3: %% Text file to use with ws-procs975x65.cls written in Latex2E.   

4: %% The content, structure, format and layout of this style file is the 

5: %% property of World Scientific Publishing Co. Pte. Ltd. 

6: %% Copyright 1995, 2002 by World Scientific Publishing Co. 

7: %% All rights are reserved.

8: %%

9: %% Proceedings Trim Size: 9.75in x 6.5in

10: %% Text Area: 8in (include runningheads) x 5in

11: %% Main Text is 10/13pt					  

12: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

13: %%

14: 

15: %\documentclass[draft]{ws-procs975x65}

16: \documentclass{ws-procs975x65}

17: \def\IR{{\hbox{{\rm I}\kern-.2em\hbox{\rm R}}}}

18: \def\IC{{\hbox{{\rm I}\kern-.2em\hbox{\rm C}}}}

19: \def\IR{{\hbox{{\rm I}\kern-.2em\hbox{\rm R}}}}

20: \def\II{{\hbox{{\rm I}\kern-.2em\hbox{\rm I}}}}

21: \def\beq{\begin{equation}}

22: \def\eeq{\end{equation}}

23: \def\be{\begin{equation}}

24: \def\ee{\end{equation}}\def\um{\frac{1}{2}}

25: \def\bea{\begin{eqnarray}}

26: \def\eea{\end{eqnarray}}

27: \def\b{\beta}

28: \def\a{\alpha}

29: \def\al{\alpha}

30: \def\ga{\gamma}

31: \def\g{\gamma}

32: \def\La{\Lambda}

33: %\def\ref{##}{(\ref{##})}

34: \begin{document}

35: 

36: \title{Quantum Holonomies in (2+1)-Dimensional Gravity}

37: 

38: \author{J.~E.~Nelson\footnote{speaker at the conference}}

39:        

40: \address{Dipartimento di Fisica Teorica, Universit\`a degli Studi di Torino\\

41: and Istituto Nazionale di Fisica Nucleare, Sezione di Torino\\

42: via Pietro Giuria 1, 10125 Torino, Italy\\E-mail: nelson@to.infn.it}

43: 

44: \author{R.~F.~Picken}

45: 

46: \address{Departamento de Matem\'{a}tica and 

47: CEMAT - Centro de Matem\'{a}tica e Aplica\c{c}\~{o}es \\

48: Instituto Superior T\'{e}cnico\\Avenida Rovisco Pais, 1049-001 

49: Lisboa, Portugal\\E-mail: rpicken@math.ist.utl.pt} 

50: 

51: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

52: % You may repeat \author \address as often as necessary      %

53: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

54: 

55: \maketitle

56: 

57: \abstracts{We describe an approach to the quantization of (2+1)--dimensional

58: gravity with topology $\IR \times T^2$ and negative cosmological constant,

59: which uses two quantum holonomy matrices satisfying a $q$--commutation

60: relation. Solutions of diagonal and upper--triangular form are constructed,

61: which in the latter case exhibit additional, non--trivial {\it internal}

62: relations for each holonomy matrix. This leads to the notion of quantum matrix

63: pairs. These are pairs of matrices with non-commuting entries, which have the

64: same pattern of internal relations, q-commute with each other under matrix

65: multiplication, and are such that products of powers of the matrices obey the

66: same pattern of internal relations as the original pair. This has implications

67: for the classical moduli space, described by ordered pairs of commuting

68: ${\rm SL}(2, \IR)$ matrices modulo simultaneous conjugation by 

69: ${\rm SL}(2, \IR)$ matrices.} 

70: 

71: \section{Introduction}

72: It is known that the phase space of $(2+1)$-dimensional gravity

73: with topology

74: $\IR \times T^2$ and negative cosmological constant $\Lambda$

75: is described by the space of gauge equivalence classes of flat

76: $SO(2,2)$ (de Sitter) connections on the torus,

77: or equivalently by the space of conjugacy classes of homomorphisms from

78: $\pi_1(T^2)$ to $SO(2,2)$ \cite{wit}.

79: Since the fundamental group of $T^2$ is generated by two classes

80: $\gamma_1$ and $\gamma_2$, subject to the relation

81: $\gamma_1\cdot\gamma_2\cdot\gamma_1^{-1}\cdot\gamma_2^{-1}={\II}$, the

82: phase space may be identified with pairs of commuting

83: $SO(2,2)$ elements $(S_1,S_2)$,

84: identified up to simultaneous conjugation by the same group element

85: $(S_1,S_2)\sim (g^{-1}S_1g, g^{-1}S_2g)$ for any $g\in SO(2,2)$. These

86: matrix pairs are the holonomies of the flat connection along the two

87: generators of the fundamental group.  Using the isomorphism

88: $\hbox{SO}(2,2)\cong \hbox{SL}(2,\IR)\otimes \hbox{SL}(2,\IR)$ in Ref. \refcite{NRZ} the

89: Poisson algebra of elements $({U_1}^{\pm},{U_2}^{\pm})$ of

90: $\hbox{SL}(2,\IR)$ was calculated. It follows from the Poisson brackets

91: of the connection at points where the curves

92: intersect. The intersection number  between $\ga_1$ and $\ga_2$ is taken

93: to be $+1$, and $\gamma_1\cdot\gamma_2$ has intersection number $-1$ with

94: $\gamma_1$ and $+1$ with $\gamma_2$. The phase space was then

95: described in terms of the six gauge-invariant (normalized) traces

96: $~ T_i^{\pm}, i=1,2,3$, where $T_i^{\pm}= {\frac 1 2}{\rm

97: tr} U_i^{\pm}$, $i=1,2$, $T_3^{\pm}={\frac 1 2}{\rm tr}(U_1^{\pm}U_2^{\pm})$

98: which satisfy the non--linear cyclical Poisson bracket algebra

99: \beq

100: \{T_i^{\pm},T_j^{\pm}\}=\mp{\frac {\sqrt {-\Lambda}} 4}({\epsilon_{ij}}^k 

101: T_k^{\pm} - T_i^{\pm}T_j^{\pm}), \quad \epsilon_{123}=1

102: \label{pbr}

103: \eeq 

104: where the superscript

105: $\pm$ refers to the two copies (real and independent) of

106: $\hbox{SL}(2,\IR)$. The six holonomies $T_i^{\pm}$ of (\ref{pbr}) provide an overcomplete description

107: of the spacetime geometry of $\IR\!\times\!T^2$.  To see this, consider the 

108: cubic polynomials

109: \begin{equation}

110: F^{\pm}=1-(T_1^{\pm})^2-(T_2^{\pm})^2-(T_3^{\pm})^2 +

111:  2 T_1^{\pm}T_2^{\pm}T_3^{\pm} .

112: \label{b9}

113: \end{equation}

114: which have vanishing Poisson brackets with all of the

115: traces $T_i^{\pm}$, are cyclically symmetric in the $T_i^{\pm}$, and

116: vanish classically by the $\hbox{SL}(2, \IR)$ Mandelstam identities;

117: setting $F^\pm = 0$ removes the redundancy.

118: 

119: The Poisson algebra (\ref{pbr}) and its generalization \cite{NR3} to more

120: complicated spatial topologies can be quantized for any value of the

121: cosmological constant \footnote[1]{There is an analogous discussion for $\Lambda$

122: positive or zero \cite{NR2}. For example, for $\Lambda$ positive,

123: the spinor group of the de Sitter

124: group $\hbox{SO}(3,1)$ is $\hbox{SL}(2,C)$, and the two $\pm$ copies

125: refer to complex conjugates.}. For a generic topology, one obtains an abstract

126: quantum algebra \cite{NR1,NR0}.  For genus $1$ with $\Lambda < 0$, the

127: quantum theory has been worked out explicitly \cite{cn1}.

128: 

129: The holonomies of (\ref{pbr}) can be represented classically as

130: \beq

131: T_1^\pm = \cosh{\frac {r_1^\pm} 2} , \quad T_2^\pm  = \cosh{\frac {r_2^\pm} 2} , \quad

132: T_3^\pm = \cosh{\frac{(r_1^\pm+r_2^\pm)} 2} , \label{cc6}

133: \eeq

134: where

135: $r_{1,2}^{\pm}$ are also real, global, time-independent (but undetermined) 

136: parameters which, from (\ref{pbr}) satisfy the Poisson brackets

137: \be\{r_1^\pm,r_2^\pm\}=\mp \sqrt {-\Lambda}, \qquad \{r_{1,2}^+,r_{1,2}^-\}= 0.

138: \label{pb}

139: \ee

140: In this case the cubic polynomials (\ref{b9}) are identically zero, 

141: %With this parametrisation, the Chern-Simons action \cite{Wit, Achu} is \cite{cn1}

142: %\begin{equation}

143: %I_2 = \int \alpha(r_1^-dr_2^- - r_1^+ dr_2^+)

144: %\label{i2}

145: %\end{equation}

146: and quantization \footnote[2]{Direct quantization of the algebra (\ref{pbr})

147: gives an algebra related to the Lie algebra of the quantum group

148: $\hbox{SU}(2)_q$ \cite{NRZ,NR5}, where $q=\exp{(4i\theta)}, \tan\theta=

149: - {\frac {\hbar \sqrt {-\Lambda}} 8}$. A (scaled) representation of the operators (\ref{cc6}) leads to

150: the commutators $[\hat r_1^{\pm}, \hat r_2^{\pm}] = \pm 8i\theta$, which differ

151: from (\ref{dc1}) by terms of order $\hbar^3$.}  is achieved by replacing

152: Eqs.(\ref{pb}) with the commutators

153: \begin{equation}

154: [\hat r_1^\pm, \hat r_2^\pm] = \mp  i\hbar \sqrt {-\Lambda}, \quad [\hat r_{1,2}^+, 

155: \hat r_{1,2}^-] = 0 .

156: \label{dc1}

157: \end{equation}

158: 

159: Previous quantizations \cite{cn1,cn2} have

160: concentrated entirely on the traces $T_i^{\pm}$ and their

161: representation Eq.(\ref{cc6}). Here we observe that we may

162: regard the quantized traces $\hat T_i^\pm$ as traces of 

163: operator-valued holonomy matrices $\hat T_i^{\pm}= \um {\rm tr}

164: \hat U_i^{\pm}$, $i=1,2$, $\hat T_3^{\pm}= \um {\rm tr}(\hat

165: U_1^{\pm} \hat U_2^{\pm})$, where (for the (+)  matrices, dropping

166: the superscript) the matrices $\hat U_i$ have, for example, the diagonal form 

167: \beq \hat U_i

168: = \left(\begin{array}{clcr}e^{\frac {{\hat r}_i} 2}&~~0\\0& e^{-{\frac {{\hat

169: r}_i} 2}} \end{array}\right)\quad 

170: = \exp{({\frac {\hat r_i \sigma_3} 2})} \label{diag} 

171: \eeq 

172: where $\sigma_3$ is one of the Pauli matrices. Now, from Eq.(\ref{dc1}) and the 

173: identity

174: $$ e^{\hat X}

175: e^{\hat Y}= e^{\hat Y} e^{\hat X} e^{[ \hat X, \hat Y ]},

176: $$

177: valid when

178: $[ \hat X, \hat Y ]$ is a $c$--number, one finds that the 

179: matrices (\ref{diag}) satisfy, {\it by both matrix and operator

180: multiplication}, the $q$--commutation relation\footnote[3]{the (-) matrices

181: satisfy a similar relation but with $q$ replaced by $q^{-1}$}:

182: \beq

183: \hat U_1 \hat U_2 = q \hat U_2 \hat U_1, \quad {\rm with} 

184: \label{fund}

185: \eeq

186: \beq

187: q=\exp (-{\frac {i \hbar \sqrt{-\Lambda}}  4})

188: \label{q}

189: \eeq

190: i.e. a deformation of the classical equation stating that the holonomies

191: commute.

192: 

193: Equations of the form (\ref{fund}) appear abundantly in the quantum group and

194: quantum geometry literature, as Weyl relations or $q$--commutators, or as the

195: defining relation for the quantum plane \cite{man}, but normally the symbols

196: $\hat U_1$ and $\hat U_2$ stand for scalar operators, as opposed to $2\times 2$

197: matrices with operator entries. 

198: 

199: Our approach is based on the fundamental equation

200: (\ref{fund}). Instead of representing the algebra of traces, Eq.

201:  (\ref{pbr}), we find representations of matrices $\hat U_1$ and $\hat

202: U_2$ satisfying Eq. (\ref{fund}) that generalize the choices

203: (\ref{diag}), for a general $q$--parameter. This constitutes a new

204: approach to quantization that is consistent with previous

205: approaches for this model \cite{NR1,cn1,cn2}, namely a deformation

206: of classical holonomies that consequently satisfy a

207: $q$--commutation relation. The gauge-invariance of

208: the traces is replaced by the gauge-covariance of Eq. (\ref{fund}) under

209: the replacements $\hat U_i\rightarrow g^{-1}\hat U_ig$, $i=1,2$ for

210: $g\in \hbox{SL}(2,\IR )$ an ordinary, i.e. not operator-valued,

211: matrix. We argue that working directly with the matrices $U_i$,

212: rather than with the indirect information contained in their

213: traces, gives a clearer insight into the structure of the phase

214: space, both classically and after quantization.

215: 

216: A more detailed account of these results, including representations, is given

217: in Ref. \refcite{NP1}. The matrices $\hat U_1$ and $\hat U_2$ determine a new

218: quantum--group--like structure, which is studied from the algebraic perspective

219: in Ref. \refcite{NP2}. The description of the classical phase space in terms of

220: pairs of matrices $U_i$ is given in Ref. \refcite{NP3}. The generalization to supergroups in the context of (2+1)--supergravity is described in Ref. \refcite{mik:pic}. 

221: 

222: \section{An algebraic solution}

223: We give just one example of a purely algebraic solution to Eq. (\ref{fund}). For

224: others see Ref. \refcite{NP2}. Consider the upper--triangular matrices 

225: \beq

226: U_i=\left( \begin{array}{cc} \a_i &\b_i\\ 0&\g_i\end{array}

227: \right), i=1,2

228: \label{Ui}

229: \eeq

230: which generalize Eq. (\ref{diag}). It can be checked that they will satisfy

231: Eq. (\ref{fund}) provided their non--commuting elements satisfy the following

232: mutual relations 

233: \be

234: \a_1\a_2=q\a_2\a_1,\quad \g_1\g_2=q\g_2\g_1, \quad {\rm and}

235: \label{ab1}

236: \ee 

237: \be

238:  \a_1\b_2 = q\b_2\g_1, \quad \b_1\g_2 = q \a_2\b_1

239: \label{ab11}

240: \ee

241: and the following {\it internal} relations for each ($i=1,2$) matrix

242: \be

243: \a_i\g_i=\g_i\a_i=1,\quad \a_i\b_i=\b_i\g_i \label{int}

244: \ee

245: 

246: Note that the mutual relations (\ref{ab1}) are standard $q$--commutation 

247: relations \cite{man}, which clearly become commutative in the classical limit 

248: $q \to 1$, whereas relations (\ref{ab11}) have a different structure, involving 

249: three elements not two. The relations (\ref{ab1}) also imply that, for example

250: \be

251: \a_1{\a_2}^{-1}=q^{-1}{\a_2}^{-1}\a_1,

252: \quad {\a_1}^{-1}{\a_2}^{-1}=q{\a_2}^{-1}{\a_1}^{-1},

253: \label{ab12}

254: \ee

255: and similarly for $\g_1,\g_2$. The internal relations Eqs. (\ref{int}) are a 

256: new feature, since, in the Poisson algebra of (2+1)--dimensional gravity

257: \cite{NRZ}, only matrix elements from different holonomies have

258: non--zero brackets and would therefore not commute on quantization.

259: Elements of a single holonomy commute.  Note that the $q$

260: parameter does not appear in the internal relations (\ref{int})

261: which therefore persist in the classical limit $q\rightarrow 1$,

262: when the matrices (\ref{Ui}) commute.

263: 

264: The internal relations (\ref{int}) are also not standard $q$--commutation

265: relations, but are preserved under matrix multiplication,

266: and in this sense they are analogous to the internal relations for quantum

267: groups \cite{man}. For example, the product $U_1U_2$ is given by

268: \be

269: U_1U_2 = \left(\begin{array}{clcr}{\a_1\a_2}&{~~\a_1\b_2 + \b_1\g_2}\\

270: 0&{~~\g_1\g_2}

271: \end{array}\right)

272: \label{matrvv}

273: \ee

274: whose internal relations are analogous to Eq. (\ref{int}), by using 

275: (\ref{ab1})--(\ref{ab11}) and (\ref{int}). This feature is discussed in 

276: greater detail in Ref. \refcite{NP1}.

277: 

278: \section{The classical moduli space} 

279: The new, non--trivial {\it internal} relations, Eq. (\ref{int}) have important

280: implications for the classical phase space, which consists of pairs of 

281: commuting $\hbox{SL}(2,\IR)$ matrices, identified up to simultaneous

282: conjugation by elements of $\hbox{SL}(2,\IR)$, since the classical counterpart

283: to Eq. (\ref{fund}) is the statement that the two matrices $U_1$ and $U_2$

284: commute. This space was studied in detail in Ref. \refcite{eza}. The algebraic analysis

285: of the classical case, in terms of the eigenvalues and eigenspaces of the two

286: matrices, does not carry over in any straightforward way to quantum matrices.

287: For instance, an upper--triangular matrix with two distinct diagonal entries

288: can be diagonalized as an ordinary matrix, but this is not in general true if

289: the matrix has non--trivial {\it internal} relations between non--commuting

290: entries. In Ref. \refcite{NP3} we give another parametrization of the classical

291: phase space, which is more appropriate to the present context. It consists of

292: sectors where both matrices are diagonalizable, but also sectors where both are

293: non-diagonalizable but can be simultaneously conjugated into upper triangular

294: form, as well as other sectors. A spectral analysis of commuting

295: $\hbox{SL}(2,\IR)$ matrices allows a classification of the equivalence classes,

296: and a unique canonical form is given for each of these. In this way the moduli

297: space becomes explicitly parametrized, and has a simple structure, resembling

298: that of a cell complex, allowing it to be depicted. Full details are given in

299: Ref. \refcite{NP3}.

300: 

301: \section*{Acknowledgments} This work was supported by the Istituto Nazionale

302: di Fisica Nucleare (INFN) of Italy, Iniziativa Specifica FI41, the Italian

303: Ministero dell'Universit\`a e della Ricerca Scientifica e Tecnologica (MIUR),

304: and by the programme {\em Programa Operacional 

305: ``Ci\^{e}ncia, Tecnologia, Inova\c{c}\~{a}o''} (POCTI) of the 

306: {\em Funda\c{c}\~{a}o para a Ci\^{e}ncia e a Tecnologia} (FCT), 

307: cofinanced by the European Community fund FEDER. 

308: 

309: \begin{thebibliography}{00}

310: 

311: %ibitem{gol}  W. M. Goldman, Adv. Math. {\bf 54}, 200--225 (1984).

312: 

313: \bibitem{wit} E. Witten, Nucl. Phys. {\bf B311}, 46-78 (1988/89).

314:  

315: \bibitem{NRZ} J.~E.~Nelson, T.~Regge and F.~Zertuche, Nucl.\ Phys.\

316:  {\bf B339}, 516 (1990).

317: 

318: \bibitem{NR3} J.~E.~Nelson and T.~Regge, Commun.\ Math.\ Phys.\ {\bf 141},

319:  211 (1991).

320: 

321: \bibitem{NR2} J.~E.~Nelson and T.~Regge, Nucl.\ Phys.\ {\bf B328}, 190 (1989).

322: 

323: \bibitem{NR1} J.~E.~Nelson and T.~Regge, Phys.\ Lett.\ {\bf B272}, 213--216 (1991).

324: 

325: \bibitem{NR0} J.~E.~Nelson and T.~Regge, Phys.\ Rev.\ {\bf D50}, 5125

326:  (1994).

327: 

328: \bibitem{cn1} S.~Carlip and J.~E.~Nelson, Phys.\ Lett.\ {\bf B324}, 299

329: (1994); Phys. Rev. {\bf D51}, 10, 5643-5653 (1995).

330: 

331: \bibitem{NR5} J.~E.~Nelson and T.~Regge, in ``Integrable Systems and

332:  Quantum Groups'', Pavia 1990, eds.~ M.~ Carfora,  M.~Martellini,

333:  A.~Marzuoli, (1992 World Scientific Publ. Co., Singapore).

334: 

335: 

336: 

337: \bibitem{cn2} S.~ Carlip and J.~E.~Nelson, Phys.\ Rev.\ {\bf D59} 024012-1

338: (1999).

339: 

340: \bibitem{man} Yu. I. Manin, {\it Quantum Groups and Non-commutative

341: Geometry}, Universit\'{e} de Montr\'{e}al, Centre de Recherches

342: Math\'{e}matiques, Montr\'{e}al, 1988. 

343: 

344: 

345: \bibitem{NP1} J.~E.~Nelson and R. F. Picken,  Phys.\ Lett.\ {\bf B471}, 

346: 367--372 (2000).

347: 

348: \bibitem{NP2} J.~E.~Nelson and R. F.  Picken, Lett. Math. Phys. {\bf 52},

349: 277--290 (2000).

350: 

351: \bibitem{NP3} J.~E.~Nelson and R. F. Picken, Lett. Math. Phys. {\bf 59}, 215--226

352: (2002).

353: 

354: \bibitem{mik:pic} A. Mikovic and R. F. Picken, Adv. Theor. Math. Phys. {\bf 5}, 243--263 (2001).

355: 

356: \bibitem{eza} K. Ezawa, ``Reduced Phase Space of the First Order Einstein

357: Gravity on $\bf R \times \bf {T^2}$'', Osaka preprint OU-HET-185

358: (1993), hep-th/9312151; Class. Quant. Grav. {\bf 12}, 373-392 (1995); Phys.

359: Rev. {\bf D49} 5211-5226 (1994); Addendum-ibid {\bf D50}, 2935-2938 (1995).

360: 

361: 

362: \end{thebibliography}

363: 

364: \end{document}

365: