1: \documentclass[12pt,a4wide]{article}

2: 

3: \bibliographystyle{lrl}

4: 

5: \usepackage{epsf}

6: \usepackage{longtable}

7: \usepackage{amsmath}

8: \usepackage{amssymb}

9: \usepackage{epubtk}

10: \usepackage{fancybox}

11: \usepackage{graphicx}  

12: 

13: \newcommand{\hypref}[2]{{#2}}

14: \newcommand{\gym}{g_{\rm YM}}

15: \newcommand{\gs}{g_S}

16: \newcommand{\Nfour}{${\cal N}=4$ }

17: \newcommand{\Nf}{${\cal N}=4$ }

18: \newcommand{\AdSS}{$AdS_5\times S^5$ }

19: \newcommand{\cO}{{\cal O}}

20: \newcommand{\Tr}{\mbox{Tr}}

21: \newcommand{\ft}[2]{{\textstyle \frac{#1}{#2}}}

22: \newcommand{\eqn}[1]{(\ref{#1})}

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

24: \newcommand{\ee}{\end{equation}}

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

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

27: \newcommand\beal{\begin{align}}

28: \newcommand\eeal{\end{align}}

29: \newcommand\nn\nonumber

30: \newcommand{\up}{\uparrow}

31: \newcommand{\down}{\downarrow}

32: \newcommand{\Zint}{\mathbb{Z}}      

33: \newcommand{\Real}{\mathbb{R}}            

34: \newcommand{\eins}{{\bf 1}}      

35: 

36: \newlength{\intwidth}

37: \DeclareRobustCommand{\fpint}[2]

38:   {\mathop{%

39:      \text{%

40:        \settowidth{\intwidth}{$\int$}%

41:        \makebox[0pt][l]{\makebox[\intwidth]{$-$}}%

42:        $\int_{#1}^{#2}$}}}

43: 

44: 

45: \begin{document}

46: 

47: \title{Spinning strings and integrable spin chains in the AdS/CFT correspondence}

48: 

49: \author{%

50: Jan Plefka\\

51: Max-Planck-Institut f\"ur Gravitationsphysik\\

52: (Albert-Einstein-Institut)\\

53: Am M\"uhlenberg 1, 14476 Potsdam, Germany\\

54: {\tt jan.plefka@aei.mpg.de}}

55: 

56: 

57: \date{{\small AEI-2005-124, hep-th/0507136}}

58: \maketitle

59: 

60: \begin{abstract}

61: In this introductory review we discuss dynamical tests of the \AdSS string/\Nfour super

62: Yang-Mills duality. After a brief introduction to AdS/CFT we argue that semiclassical string energies

63: yield information on the quantum spectrum of the string in the limit

64: of large angular momenta on the $S^5$.

65: The energies of the folded and circular spinning string solutions

66: rotating on a $S^3$ within the $S^5$ are derived,

67: which yield all loop predictions for the dual gauge

68: theory scaling dimensions. These follow from the eigenvalues of the dilatation operator of 

69: \Nfour super Yang-Mills in a minimal $SU(2)$ subsector and we display its reformulation in terms of a

70: Heisenberg $s=1/2$ spin chain along with the coordinate Bethe ansatz for its explicit diagonalization. 

71: In order to make

72: contact to the spinning string energies we then study the thermodynamic limit of the one-loop gauge

73: theory Bethe equations and demonstrate the matching with the folded and closed string result at this

74: loop order. Finally the known gauge theory results at higher-loop orders are reviewed and the 

75: associated long-range spin chain Bethe ansatz is introduced, leading to an asymptotic all-loop

76: conjecture for the gauge theory Bethe equations. This uncovers discrepancies at the three-loop

77: order between gauge theory scaling  dimensions and string theory energies and the implications

78: of this are discussed. Along the way we comment on further developments and generalizations

79: of the subject and point to the relevant literature. 

80: \end{abstract}

81: 

82: \epubtkKeywords{String Theory, Supersymmetric Yang-Mills, AdS/CFT Correspondence}

83: 

84: \newpage

85: 

86: 

87: %==============================================================================

88: 

89: \section{Introduction}

90: \label{section:introduction}

91: 

92: String theory was initially discovered in  an attempt to describe the physics

93: of the strong interactions prior to the advent of gauge field theories and

94: QCD. Today it has matured to a very promising candidate for a unified 

95: quantum theory of gravity and all the other  forces of nature.

96: In this interpretation gauge fields arise as the low energy

97: excitations of fundamental open strings and are therefore

98: derived, non-fundamental objects, just as the theory of gravity itself. 

99: Ironically though, advances in our understanding

100: of non-perturbative string theory and of D-branes has led to a resurrection of gauge fields

101: as fundamental objects. Namely it is now generally believed that

102: string theory in suitable space-time backgrounds can have a dual, holographic description in 

103: terms of gauge field theories and thus the question

104: of which of the two is the fundamental one becomes redundant. This

105: belief builds on a remarkable proposal due to Maldacena \cite{hep-th/9711200} known

106: as the Anti-de-Sitter/Conformal Field Theory (AdS/CFT) 

107: correspondence (for reviews see \cite{hep-th/9905111,hep-th/0201253}).

108: 

109: The initial idea of a string/gauge duality is due to 't Hooft \cite{'tHooft:1973jz}, who 

110: realized that the perturbative expansion of $SU(N)$ gauge field

111: theory in the large $N$ limit can be reinterpreted as a genus expansion of

112: discretized two dimensional surfaces built from the field theory Feynman diagrams.

113: Here $1/N$ counts the genus of the Feynman diagram, while the 't Hooft coupling

114: $\lambda:=\gym^2\, N$ (with $\gym$ denoting the gauge theory coupling constant)

115: enumerates quantum loops. 

116: E.g.~the genus expansion of the free energy $F$ of a  $SU(N)$ gauge theory in the 't

117: Hooft limit ($N\to\infty$ with $\lambda$ fixed) takes the pictorial form

118: %

119: \begin{equation}

120: F=N^2\,\raisebox{-.3cm}{\epsfysize=1cm\epsfbox{g0.eps}}

121: + {1}\,  \,\raisebox{-.3cm}{\epsfysize=1cm\epsfbox{g1.eps}}

122: + {\frac{1}{N^2}}\,  \,\raisebox{-.3cm}{\epsfysize=1cm\epsfbox{g2.eps}}

123: +\ldots 

124:  =\sum_{g=0}^\infty \frac{1}{{N^{2g-2}}}\, \sum_{l=0}^\infty c_{g,l}\, 

125: {\lambda}^l

126: \end{equation}

127: %

128: with suitable coefficients $c_{g,l}$ denoting the contributions at genus $g$ and loop

129: order $l$. Obviously this $1/N$ expansion resembles the perturbative expansion of a 

130: string theory in the string coupling constant $g_S$.

131: 

132: The AdS/CFT correspondence is the first concrete realization of this

133: idea for four dimensional gauge theories.

134: In its purest form -- which shall also be the setting we will be

135: interested in -- it identifies the `fundamental'

136: type IIB superstring in a ten dimensional

137:  anti-de-Sitter cross sphere ($AdS_5\times S^5$) space-time background with the 

138: maximally supersymmetric Yang-Mills theory with gauge group $SU(N)$

139: (\Nfour SYM) in four dimensions. The \Nfour super Yang-Mills model is a

140: quantum conformal field theory, as its $\beta$-function vanishes exactly. The string model 

141: is controlled by two parameters: The string coupling constant $\gs$

142: and the `effective' string tension

143: $R^2/\alpha'$ where $R$ is the common radius of the 

144: $AdS_5$ and $S_5$ geometries. The gauge theory, on the other hand, is parametrized by the rank

145: $N$ of the gauge group and the coupling constant $\gym$, or equivalently the 't Hooft coupling

146: $\lambda:=\gym^2\, N$. According to the AdS/CFT proposal, these two sets of parameters are to be

147: identified as

148: %

149: \begin{equation}

150: \label{eq:AdS/CFT rels}

151: \frac{4\pi\lambda}{N}=\gs \qquad \sqrt{\lambda} = \frac{R^2}{\alpha'} \, .

152: \end{equation}

153: %

154: We see that in the AdS/CFT proposal the string coupling constant is

155: not simply given by $1/N$, but comes with a linear factor in

156: $\lambda$. This, however, does not alter the genus expansion and its

157: interpretation in form of string worldsheets.

158: 

159: The equations \eqn{eq:AdS/CFT rels} relate the coupling constants but there is

160: also a dictionary between the excitations of the two theories.

161: The correspondence identifies the energy eigenstates of the \AdSS

162: string, 

163: which we denote schematically as $|\cO_A\rangle$ with $A$ being a multi-index,  

164: with (suitable) composite gauge

165: theory operators of the form 

166: $\cO_{A}=\Tr(\phi_{i_1}\, \phi_{i_2}\ldots\phi_{i_n})$ where

167: $(\phi_{i})_{ab}$ are the

168: elementary fields of \Nfour SYM (and their covariant derivatives) in the adjoint

169: representation of $SU(N)$, i.e.~$N\times N$ hermitian matrices . 

170: The energy eigenvalue $E$ of  a string state with respect to time in

171: global coordinates 

172: is conjectured to be equal to the scaling dimension $\Delta$ of the dual gauge theory operator,

173: which in turn is determined from the two point function of the conformal field theory

174: \footnote{For simplicity we take ${\cal O}_A(x)$ to be a scalar

175:   operator here.}

176: %

177: \begin{equation}

178: \label{eq:DvsE}

179: \langle \cO_A(x)\, \cO_B(y) \rangle = \frac{M\,

180:   \delta_{A,B}}{(x-y)^{2\, \Delta_A(\lambda,\ft 1 N)}}

181: \quad\Leftrightarrow\quad

182: {\cal H}_{\rm String}\, |\cO_A\rangle = E_A(\ft{R^2}{\alpha'},\gs)\,

183: |\cO_A\rangle 

184: \end{equation}

185: %

186: with $\Delta(\lambda,\ft 1 N) \stackrel{!}{=} E(\ft{R^2}{\alpha'},\gs)$.

187: 

188: A zeroth order test of the conjecture is then the agreement of the

189: underlying symmetry supergroup $PSU(2,2|4)$ of the

190: two theories, which furnishes the representations under which $\cO_A(x)$

191: and $|\cO_A\rangle$  transform. This then yields 

192: a hint on how one could set up an explicit string state/gauge operator dictionary. 

193: 

194: Clearly there is little hope of determining either the all genus 

195: (all orders in $\gs$) string spectrum or the complete $1/N$ dependence of the gauge theory scaling

196: dimensions $\Delta$. But the identification of the planar gauge theory with the free ($\gs=0$)

197: string seems feasible and fascinating: Free \AdSS string theory

198: should give the exact all loop

199: gauge theory scaling dimensions in the large $N$ limit! Unfortunately though, 

200: our knowledge of the string spectrum in 

201: curved backgrounds, even in such a highly symmetric one as \AdSS,

202: remains scarce.  Therefore until very recently

203: investigations on the string side of the correspondence were limited to the

204: domain of the low energy effective field

205: theory description of \AdSS strings in terms of type IIB supergravity. This, however, is necessarily

206: limited to weakly curved geometries in string units, i.e.~to the domain of $\sqrt{\lambda}\gg 1$

207: by virtue of \eqn{eq:AdS/CFT rels}. On the gauge theory side one has control only

208: in the perturbative regime where $\lambda\ll 1$, which is perfectly incompatible

209: with the accessible supergravity regime $\sqrt{\lambda}\gg 1$. Hence one is facing 

210: a strong/weak coupling 

211: duality, in which strongly coupled gauge fields are described by {\sl classical} supergravity and

212: weakly coupled gauge fields correspond to strings propagating in a highly curved background

213: geometry. This insight is certainly fascinating, but at the same time strongly hinders any dynamical tests

214: (or even a proof) of the AdS/CFT conjecture in regimes which are not protected by the

215: large amount of symmetry in the problem.

216: 

217: This situation has profoundly changed since 2002 by performing studies of the correspondence in 

218: novel limits where quantum numbers (such as spins or angular momenta in the geometric \AdSS language) 

219: become large in a correlated fashion as $N\to\infty$. 

220: This was initiated in the work of Berenstein, Maldacena

221: and Nastase \cite{hep-th/0202021} who considered the quantum fluctuation expansion of the

222: string around a degenerated point-like configuration, 

223: corresponding to a particle rotating with a large angular momentum $J$ on a

224: great circle of the $S^5$ space. In the limit of $J\to\infty$ with $J^2/N$ held fixed (the `BMN limit')

225: the geometry seen by the fastly moving particle is a

226: gravitational plane-wave, which allows for an exact quantization of the

227: free string in the light-cone-gauge \cite{hep-th/0112044,hep-th/0202109}. 

228: The resulting string spectrum leads to a formidable prediction

229: for the all loop scaling dimensions of the dual gauge theory operators in the corresponding

230: limit, i.e.~the famous formula $\Delta_n= J +2\, \sqrt{1+\frac{\lambda\,n^2}{J^2}}$ for the simplest

231: two string oscillator mode excitation.

232: The key point here is the emergence of the effective gauge theory 

233: loop counting parameter $\lambda/J^2$ in the BMN limit.

234: These scaling dimensions have by now been firmly reproduced up to the three loop order

235: in gauge theory \cite{hep-th/0303060,hep-th/0310252,hep-th/0409009}. 

236: This has  also led to important structural information for higher (or all loop)

237: attempts in gauge theory, which maximally employ the uncovered  

238: integrable structures to be discussed below. 

239: Moreover, 

240: the plane wave string theory/\Nfour SYM duality could be extended to the interacting string 

241: ($g_S\neq 0$) respectively non-planar gauge theory regime providing us with the most concrete realization 

242: of a string/gauge duality to

243: date (for reviews see \cite{hep-th/0307027,hep-th/0307101,hep-th/0310119,hep-th/0401155}). 

244: 

245: In this review we shall discuss developments beyond the BMN plane-wave

246: correspondence which employ more

247: general sectors of large quantum numbers in the AdS/CFT duality. The key point from the

248: string perspective is that such a limit can make the semiclassical (in the plane-wave case) or even

249: classical (in the `spinning string' case) computation of the string

250: energies also quantum exact \cite{hep-th/0304255,hep-th/0306130}, 

251: i.e. higher $\sigma$-model

252: loops are suppressed by inverse powers of the total angular momentum

253: $J$ on the five sphere\footnote{This suppression

254:   only occurs if at least one angular momentum on the five sphere

255:   becomes large. If this is not the case, e.g.~for a spinnning string

256:   purely on the $AdS_5$ \cite{hep-th/0204051}, quantum corrections are not

257:   suppressed by inverse powers of the spin on $AdS_5$

258:   \cite{hep-th/0204226}.}. 

259: These considerations on the string

260: side then (arguably) yield all loop predictions

261: for the dual gauge theory. Additionally the perturbative gauge

262: theoretic studies at the first few orders in $\lambda$ led to

263: the discovery that the spectrum of scaling dimensions of the planar

264: gauge theory is identical to that of an integrable long-range spin

265: chain \cite{hep-th/0212208,hep-th/0303060,hep-th/0307042}. 

266: Consistently the $AdS_5\times S^5$ string

267: is a classically integrable model \cite{hep-th/0305116}, which has been heavily exploited in the

268: construction of spinning string solutions.

269: 

270: This review aims at a more elementary introduction to this very active area of research, which

271: in principle holds the promise of finding the exact quantum spectrum of the $AdS_5\times S^5$  

272: string or equivalently 

273: the all loop scaling dimensions of planar \Nfour super Yang-Mills. It is

274: intended as a first guide to the field for students and interested `newcomers'

275:  and points to the relevant literature for deeper studies.

276: We will discuss the simplest solutions of the $AdS_5\times S^5$ string corresponding to folded

277: and circular string configurations propagating in a $\Real\times

278: S^3$ subspace, with the $S^3$ lying within the $S^5$.

279: On the gauge theory side we will motivate the emergence of

280: the spin chain picture at the leading one loop order 

281: and discuss the emerging Heisenberg XXX${}_{1/2}$ model and its diagonalization

282: using the coordinate Bethe ansatz technique. This then enables us to perform a comparison 

283: between the classical string predictions in the limit of large angular momenta 

284: and the dual thermodynamic limit of the spin chain spectrum. Finally we turn to higher-loop

285: calculations in the gauge-theory and discuss conjectures for the all-loop form of the Bethe

286: equations, giving rise to a long-range interacting spin chain. Comparison with the obtained 

287: string results uncovers a discrepancy from loop order three onwards and the 

288: interpretation of this result is also discussed.

289: 

290: There already exist a number of more detailed reviews on spinning

291: strings, integrability and spin chains in the AdS/CFT correspondence:

292: Tseytlin's review \cite{hep-th/0311139} mostly focuses on the string side

293: the correspondence, whereas Beisert's Physics Report \cite{hep-th/0407277} concentrates 

294: primarily on the gauge side. See also Tseytlin's second review

295: \cite{hep-th/0409296}, on the so-called coherent-state

296: effective action approach, which we will not discuss in this review. 

297: Recommended is also the shorter review by

298: Zarembo \cite{hep-th/0411191} on the $SU(2)$ respectively $\Real\times S^3$

299: subsector, discussing the integrable structure appearing on the classical string -- not

300: covered in this review. For a detailed account of the near plane-wave superstring, its

301: quantum spectroscopy and integrability structures see Swanson's thesis \cite{hep-th/0505028}.

302: 

303: %==============================================================================

304: 

305: \section{The setup}

306: \label{section:setup}

307: 

308: With the embedding coordinates $X^m(\tau,\sigma)$ and

309: $Y^m(\tau,\sigma)$ the Polyakov action of the \AdSS 

310: string in conformal gauge ($\eta_{ab}=\mbox{diag}(-1,1)$) takes the

311: form ($m,n=1,2,3,4,5$)

312: \be

313: \label{eq:AdSSaction}

314: I=-\frac{\sqrt{\lambda}}{4\pi}\, \int d\tau d\sigma\, \Bigl [ G_{mn}^{(AdS_5)}\, \partial_aX^m\, \partial^aX^n

315: + G_{mn}^{(S^5)}\, \partial_a Y^m\, \partial^a  Y^n\, \Bigr ] + \mbox{fermions}\, ,

316: \ee

317: where we have suppressed the fermionic terms in the action, as they will not be relevant in our

318: discussion of classical solutions (the full fermionic action is stated in

319: \cite{hep-th/9805028,hep-th/0010104}). A natural choice of 

320: coordinates for the \AdSS space (``global coordinates'')

321: is

322: \footnote{This metric arises from a parametrization of the five sphere

323: $x_1^2+x_2^2+x_3^2+\ldots + x_6^2=1$ and anti-de-Sitter space

324: $-y_{-1}^2-y_0^2+y_1^2+y_2^2+\ldots y_4^2=-1$ through

325: \begin{align}

326: x_1+i\, x_2&= \sin\gamma\, \cos\psi \, e^{i\,\phi_1},\quad

327: x_3+i\, x_4= \sin\gamma\, \sin\psi \, e^{i\,\phi_2},\quad

328: x_5+i\, x_6= \cos\gamma\, \, e^{i\,\phi_3};\nn\\

329: y_1+i\, y_2&= \sinh\rho\, \cos\bar\psi \, e^{i\,\varphi_1},\quad

330: y_3+i\, y_4= \sinh\rho\, \sin\bar\psi \, e^{i\,\varphi_2},\quad

331: y_{-1}+i\, y_0= \cosh\rho\, \, e^{i\,t}\, .\nn

332: \end{align}

333: The embedding coordinates $X^m(\tau,\sigma)$ and $Y^m(\tau,\sigma)$ in

334: \eqn{eq:AdSSaction} are hence given

335: by $X^m=(\rho,\bar\psi,\varphi_1,\varphi_2,t)$ and $Y^m=(\gamma,\psi,\phi_1,\phi_2,\phi_3)$.}

336: \begin{align}

337: \label{eq:metrics}

338: ds^2_{AdS_5} &= d\rho^2 -\cosh^2 \rho \, dt^2 +\sinh^2\rho\, (\,

339: d\bar\psi^2 +\cos^2\bar\psi\, d{\varphi_1}^2

340: +\sin^2\bar\psi\, d{\varphi_2}^2\, )\nn\\

341: ds^2_{S^5} &= d\gamma^2 +\cos^2\gamma\, d{\phi_3}^2 +\sin^2\gamma\, (\, d\psi^2 +\cos^2\psi\, d{\phi_1}^2

342: +\sin^2\psi\, d{\phi_2}^2\, )\, .

343: \end{align}

344: Moreover, we have directly written the string

345: action with the help of the effective string tension $\sqrt{\lambda}=R^2/\alpha'$ 

346: of \eqn{eq:AdS/CFT rels}. It is helpful to picture the $AdS_5$

347: space-time as a bulk cylinder with a four dimensional boundary of the

348: form $\Real\times S^3$ (see figure \ref{fig:one}).

349: \begin{figure}[t]

350: \label{fig:one}

351: \begin{center}

352: \includegraphics[width=2cm]{AdS5PIC.eps}

353: \quad \raisebox{0.8cm}{$\times$} \quad

354: \includegraphics[width=2cm]{S5PIC.eps}

355: \end{center}

356: \caption{Cartoon of the $AdS_5$ (bulk cylinder with boundary $R\times

357:   S^3$) space-time and the $S^5$ (sphere) space.}

358: \end{figure}

359: A consequence of the conformal gauge choice are the Virasoro

360: constraints, which take the form

361: \be

362: \label{VirConstr}

363: 0\stackrel{!}{=} \dot X^m\, X'_m + \dot Y^p\, {Y'_p} \qquad

364: 0\stackrel{!}{=} \dot X^m\, \dot X_m + \dot Y^p\, \dot Y_p 

365: + {X^m}'\, X_m' +  {Y^p}'\,  Y_p' \, ,

366: \ee

367: where the dot refers to $\partial_\tau$ and the prime to

368: $\partial_\sigma$ derivatives. Of course the contractions in the

369: above are to be performed with the metrics of \eqn{eq:metrics}.

370: 

371: As mentioned in the introduction it is

372: presently unknown how to perform an exact quantization of this model.

373: It is, however, possible to perform a quantum fluctuation expansion

374: in $1/\sqrt{\lambda}$. For this one expands 

375: around a classical solution of \eqn{eq:AdSSaction} and integrates out

376: the fluctuations in the path-integral loop order by order. 

377: This is the route we will follow. Of course in doing so one will

378: only have a patch-wise access to the full spectrum of the theory,

379: with each patch given by the solution expanded around.

380: 

381: \subsection{The rotating point-particle}

382: \label{sect:particle}

383: It is instructive to sketch this procedure by considering the perhaps

384: simplest solution to the equations of motion of \eqn{eq:AdSSaction}:

385: the rotating point-particle on $S^5$, which is a degenerated string

386: configuration.

387: 

388: \begin{minipage}{6cm}

389: \begin{align}

390: \label{eq:pp}

391: t&=\kappa\,\tau \quad \rho=0 \quad \gamma=\frac{\pi}{2}\nn\\

392: \phi_1&=\kappa\,\tau \quad\phi_2=\phi_3=\psi=0\nn

393: \end{align}

394: \end{minipage}

395: \qquad\qquad

396: \begin{minipage}{6cm}

397: \raisebox{-2cm}{\epsfxsize=1.2cm\epsfbox{PunktAdS.eps}}

398: \raisebox{-1.3cm}{$\times$}

399: \raisebox{-2cm}{\epsfxsize=1.4cm\epsfbox{PunktS5.eps}}\\[1cm]

400: \end{minipage}

401: 

402: One easily shows that this configurations satisfies the equations of

403: motion and the Virasoro constraints. 

404: 

405: A glance at

406: eqs.~\eqn{eq:AdSSaction} and \eqn{eq:metrics} reveals that the cyclic

407: coordinates of the action $I=\int dt\, L$ are

408: $(t,\varphi_1,\varphi_2;\phi_1,\phi_2,\phi_3)$

409: leading to the conserved charges $(E,S_1,S_2;J_1,J_2,J_3)$,

410: corresponding to the energy $E$ and two spins $(S_1,S_2)$ on 

411: $AdS_5$ as well as the three angular momenta $(J_1,J_2,J_3)$

412: on the five sphere respectively. The energy $E$ and

413: the first angular momentum $J_1$ are the only non vanishing conserved

414: quantities of the above point-particle configuration

415: and take the values (we also spell out $J_2$ for later use)

416: \bea

417: E&:=&\frac{\partial L}{\partial \dot t}=

418: \sqrt{\lambda}\int_0^{2\pi}\ft{d\sigma}{2\pi}\, 

419: \cosh^2\rho\, \dot t = {\sqrt{\lambda}}\, \kappa \, ,\nn\\

420: J_1&:=& -\frac{\partial L}{\partial \dot \phi_1}=

421: \sqrt{\lambda}\int_0^{2\pi}\ft{d\sigma}{2\pi}\,\sin^2\gamma\,

422: \cos^2\psi\,\dot \phi_1 = {\sqrt{\lambda}}\, \kappa\, , \label{EJ1J2}\\

423: J_2&:=& -\frac{\partial L}{\partial \dot \phi_2}=

424: \sqrt{\lambda}\int_0^{2\pi}\ft{d\sigma}{2\pi}\,\sin^2\gamma\,

425: \sin^2\psi\,\dot \phi_2 = 0\, ,\qquad

426: J:= J_1+J_2\, . \nn

427: \eea

428: Hence classically $E=J$. One may now consider quantum fluctuations

429: around this solution, i.e. $X^\mu=X^\mu_{\rm{solution}}(\tau) +

430: \frac{1}{\lambda^{1/4}}\, x^\mu(\tau,\sigma)$, and can integrate out

431: the quantum field $x^\mu(\tau,\sigma)$ in a perturbative fashion. 

432: This will result in `quantum' corrections to the classical energy $E_0$  in terms of

433: an  expansion in $1/\sqrt{\lambda}$

434: \be

435: E=\sqrt{\lambda}\, \kappa + E_2(\kappa) + \frac{1}{\sqrt{\lambda}}\,

436: E_4(\kappa) +\ldots

437: \ee

438: They key idea of Berenstein, Maldacena and Nastase \cite{hep-th/0202021} was to

439: consider the limit $J\to\infty$ with the parameter $\kappa=J/\sqrt{\lambda}$ held

440: fixed. This limit of a large quantum number 

441: suppresses all the higher loop contributions beyond one-loop, i.e.

442: \be

443: E-J= E_2(\kappa) + \frac{1}{\sqrt{J}}\, \tilde E_4(\kappa) +\ldots

444: \stackrel{J\to\infty}{\longrightarrow}  E_2(\kappa)\, .

445: \ee 

446: Hence the quadratic approximation becomes exact! This quadratic

447: fluctuation action (including the fermions) is nothing but the IIB

448: superstring in a plane wave background \cite{hep-th/0112044}, which

449: arises from the \AdSS geometry through a so-called Penrose limit

450: (see \cite{hep-th/0110242,hep-th/0201081} for this construction). The quantization of

451: this string model is straightforward in the light-cone gauge 

452: \cite{hep-th/0112044,hep-th/0202109} and leads to a free, massive two

453: dimensional theory for the transverse degrees of freedom ($i=1,\ldots,

454: 8$)

455: \begin{equation}

456: I_2=\int d\tau d\sigma (\frac 12 \partial_ax^i\, \partial_ax^i

457: -\frac{\kappa^2}{2}\, x^i\, x^i + \mbox{fermions})

458: \end{equation}

459: with a compact expression for the spectrum 

460: \be

461: E_2=\frac{1}{\sqrt{\lambda'}}\,

462: \sum_{n=-\infty}^\infty\sqrt{1+\lambda'\, n^2}\, \widehat N_n

463: \qquad \lambda':= \frac{1}{\sqrt{\kappa}}=\frac{\lambda}{J^2}

464: \ee

465: where $\widehat N_n:=\alpha^{\dagger\, i}_n\, \alpha_n^i$ is the

466: excitation number operator for transverse  string excitations

467: $\alpha^{\dagger\, i}_n\, |0\rangle$ with

468: $[\alpha_m^i,\alpha^{\dagger\, j}_n]=\delta_{nm}\, \delta_{ij}$. The

469: Virasoro constraints \eqn{VirConstr} reduce to the level matching condition

470: $\sum_n n \, \widehat N_n=0$ known from string theory in flat Minkowski space-time. 

471: Hence the first stringy excitation is

472: $\alpha_n^\dagger\, \alpha_{-n}^\dagger\, |0\rangle$ with

473: $\sqrt{\lambda'}\, E_2=2\,

474: \sqrt{1+\lambda'\, n^2}$. For a more detailed treatment of the plane

475: wave superstring see \cite{hep-th/0307027,hep-th/0307101,hep-th/0310119,hep-th/0401155}.

476: 

477: \subsection{\Nfour Super Yang-Mills}

478: 

479: The conjectured dual gauge theory of the $AdS_5\times S^5$ superstring is

480: the maximally supersymmetric (${\cal N}=4$)

481: Yang-Mills theory in four dimensions \cite{NUPHA.B122.253,NUPHA.B121.77}.

482: Its field content is given by a gluon field $A_\mu(x)$, six scalars $\phi_i(x)$ 

483: ($i=1,\ldots,6$) as well as 4 Majorana gluinos, which we choose to write as a 16 component

484: 10d Majorana-Weyl spinor  $\chi_\alpha(x)$ ($\alpha=1,\ldots 16$). 

485: All fields are in the adjoint representation of $SU(N)$.

486: The action of ${\cal N}=4$ Super Yang-Mills is uniquely determined by two parameters, the

487: coupling constant $g_{\rm YM}$ 

488: and the rank of the gauge group $SU(N)$

489: \be

490: S=\frac{2}{\gym^2}\, \int d^4x\, \Tr\, \Bigl\{ \frac{1}{4}\, (F_{\mu\nu})^2

491: +\frac{1}{2}\, (D_\mu\phi_i)^2 -\frac{1}{4}\, [\phi_i,\phi_j]\,

492: [\phi_i,\phi_j] + \frac{1}{2}\, \bar\chi D\!\!\!\!/\,\, \chi - \frac{i}{2}

493: \, \bar \chi\, \Gamma_i\, [\phi_i,\chi]\, \Bigr \}

494: \ee

495: with the covariant derivative $D_\mu= \partial_\mu

496: -i[A_\mu,\,\,\,]$. Furthermore, $(\Gamma_\mu,\Gamma_i)$ are the

497: ten dimensional Dirac matrices.

498: 

499: Due to the large amount of supersymmetry present, the conformal invariance 

500: of the classical field theory survives the quantization procedure: The

501: coupling constant $\gym$ is not renormalized and its $\beta$-function

502: vanishes to all orders in perturbation theory

503: \cite{PHLTA.B100.245,NUPHA.B236.125,PHLTA.B123.323}.

504: This is why one often refers to ${\cal N}=4$ Super Yang-Mills

505: as a ``finite'' quantum field theory. Nevertheless composite gauge invariant

506: operators, i.e.~traces of products of fundamental 

507: fields and their covariant derivatives at the same space-point, 

508: e.g.~${\cal O}_{i_1\ldots i_k}(x)=\Tr[\phi_{i_1}(x)\,\phi_{i_2}(x)\ldots 

509: \phi_{i_k}(x)]$, are renormalized and acquire anomalous dimensions. These

510: may be read off from the two point functions (stated here for the case

511: of scalar operators)

512: \be

513: \langle {\cal O}_A(x)\, {\cal O}_B(y)\rangle = \frac{\delta_{AB}}{(x-y)^{2\,

514: \Delta_{{\cal O}_A}}}

515: \label{CFT2PtFct}

516: \ee

517: where $\Delta_{{\cal O}_A}$ is the  scaling dimension of the composite

518: operator ${\cal O}_A$. Classically these scaling dimensions are simply

519: the sum of the individual dimensions 

520: of the constituent fields ($[\phi_i]=[A_\mu]=1$ and $[\chi]=3/2$). 

521: In quantum theory the scaling dimensions 

522: receive anomalous corrections, organized in

523: a double expansion in $\lambda=\gym^2\, N$ (loops) and $1/N^2$ (genera)

524: \be

525: \Delta = \Delta_0 + \sum_{l=1}^\infty \lambda^l\, \sum_{g=0}^\infty

526: \frac{1}{N^{2\, g}}\, \Delta_{l,g} \, .

527: \label{anexp}

528: \ee

529: Determining the scaling dimensions in perturbation theory is a difficult task

530: due to the phenomenon of operator mixing: One has to identify the correct

531: basis of (classically) degenerate gauge theory operators $\cO_A$ in which

532: \eqn{CFT2PtFct} indeed becomes diagonal. This task is greatly facilitated

533: through the use of the gauge theory dilatation operator, to be discussed in

534: section \ref{sec:gt}.

535: 

536: The core statement of the AdS/CFT correspondence is that the scaling dimensions

537: $\Delta(\lambda,1/N^2)$ are equal to the energies $E$ of the \AdSS string excitations.

538: A central problem, next to actually computing these quantities on either side of the 

539: correspondence, is to establish a 'dictionary' between states in the string

540: theory and their dual gauge theory operators. Here the underlying symmetry structure

541: of $SU(2,2|4)$ is of help, whose bosonic factors are $SO(2,4)\times SO(6)$. 

542: $SO(2,4)$ corresponds to the isometry group of $AdS_5$  or 

543: the conformal group  in four dimensions respectively. $SO(6)$, on the

544: other hand, emerges from the isometries of the five sphere and the 

545: $R$-symmetry group of internal rotations of the six scalars and four gluinos in 

546: SYM. Clearly then any state or operator can be labeled by the eigenvalues of

547: the six Cartan generators of $SO(2,4)\times SO(6)$

548: \be

549: (E;\underbrace{S_1,S_2}_{S^3}; \underbrace{J_1,J_2,J_3}_{S^5})

550: \ee

551: where we shall denote the $S_i$ as the commuting 'spins' on the three sphere within 

552: $AdS_5$, the $J_k$ as the commuting angular momenta on the $S^5$ and $E$ is the 

553: total energy. These are the conserved quantities of the string

554: discussed above eq.~\eqn{EJ1J2}.

555: 

556: The strategy is now to search for string solutions with

557: energies 

558: $$E=E(S_1,S_2,J_1,J_2,J_3)$$ 

559: and to identify these with scaling dimensions of Super Yang-Mills operators carrying the

560: identical set of charges ($S_1$,$S_2$,$J_1$,$J_2$,$J_3$). Generically the string energies

561: will depend on additional data, such as winding numbers or oscillator levels, which we have

562: suppressed in the above.

563: 

564: 

565: %==============================================================================

566: 

567: \section{Spinning string solutions}

568: \label{section:spinning_string_solutions}

569: 

570: We shall now look for a solution of the \AdSS string with $S_1=S_2=0$ and $J_3=0$,

571: i.e.~a string configuration rotating in the $S^3$ within $S^5$

572: and evolving only with the time coordinate of the $AdS_5$ space-time. This was first

573: discussed by Frolov and Tseytlin in \cite{hep-th/0304255,hep-th/0306143}. 

574: For this let us consider the

575: following ansatz in the global coordinates of

576: \eqn{eq:metrics} 

577: \be

578: t=\kappa\, \tau \qquad \rho=0 \qquad \gamma= \ft \pi 2 \qquad

579: \phi_3=0  \qquad \phi_1=\omega_1\, \tau \qquad  \phi_2=\omega_2\, \tau

580: \qquad \psi=\psi(\sigma)\, ,

581: \label{ansatz}

582: \ee 

583: with the constant parameters $\kappa$, $\omega_{1,2}$ and the profile $\psi(\sigma)$ to be

584: determined. The string action \eqn{eq:AdSSaction} then becomes 

585: \be

586: I=-\frac{\sqrt\lambda}{4\pi}\, \int d\tau\, \int_0^{2\pi}d\sigma\, 

587: \Bigl [ \, \kappa^2 + {\psi'}^2 - \cos^2\psi\, {\omega_1}^2 - \sin^2\psi\, {\omega_2}^2 \,

588: \Bigr ]\, ,

589: \ee

590: leading to an equation of motion for $\psi(\sigma)$

591: \be

592: \psi'' + \sin\psi\, \cos\psi\, ({\omega_2}^2-{\omega_1}^2) =0\, .

593: \label{EOMpsi}

594: \ee

595: We define ${\omega_{21}}^2:= {\omega_2}^2-{\omega_1}^2$, which we take to be positive

596: without loss of generality. Integrating this equation once yields the ``string pendulum''

597: equation

598: \be

599: \frac{d\psi}{d\sigma} = \omega_{21}\,\sqrt{q - \sin^2\psi}\, ,

600: \label{stringpendulum}

601: \ee

602: where we have introduced an integration constant $q$. Clearly there are two qualitatively

603: distinct situations for $q$ larger or smaller than one: For $q\leq 1$ we have a folded string

604: with $\psi$ ranging from $-\psi_0$ to $\psi_0$, where $q=\sin^2\psi_0$, and $\psi'=0$ at

605: the turning points where the string folds back onto itself (see figure

606: \ref{fig:foldedstring}).

607: \begin{figure}[t]

608: \begin{center}

609: \includegraphics[width=4cm]{stringbanana.eps}

610: \end{center}

611: \caption{The folded sting extending from $\psi=-\psi_0$ to

612:   $\psi=\psi_0$, where $\sin^2\psi_0:=q$.}

613: \label{fig:foldedstring}

614: \end{figure} 

615: If, however, $q>1$ then $\psi'$ never vanishes and we have a circular

616: string configuration embracing a full circle on the $S^3$: The energy stored in the system is

617: large enough to let the pendulum overturn.

618: 

619: In addition we have to fulfill the Virasoro constraint equations \eqn{VirConstr}. One checks that

620: our ansatz \eqn{ansatz} satisfies the first constraint of \eqn{VirConstr}, whereas the second

621: constraint equation leads to

622: \be

623: q=\frac{\kappa^2 -{\omega_1}^2}{{\omega_{21}}^2}\qquad \qquad({\omega_{21}}^2\neq 0)\, ,

624: \label{F1}

625: \ee

626: relating the integration constant $q$ to

627: the parameters of our ansatz\footnote{The case $\omega_{21}=0$

628: is discussed in subsection \ref{sect:simplesol} below.}.

629: Our goal is to 

630: compute the energy $E$ of these two solutions as a function of the commuting angular momenta

631: $J_1$ and $J_2$ on the three sphere within $S^5$. Upon using

632: eqs.~\eqn{EJ1J2} and inserting the ansatz \eqn{ansatz} these are given by

633: \bea

634: E&=& \sqrt{\lambda} \, ,\\

635: \label{J1J2def}

636: J_1&=&  \sqrt{\lambda}\,\omega_1\,\int_0^{2\pi}

637: \ft{d\sigma}{2\pi}\,\cos^2\psi(\sigma)\, ,\qquad

638: J_2=  \sqrt{\lambda}\, \omega_2\,\int_0^{2\pi}

639: \ft{d\sigma}{2\pi}\,\sin^2\psi(\sigma)  \, .

640: \eea

641: From this we learn that

642: \be

643: \label{F3}

644: \sqrt{\lambda}= \frac{J_1}{\omega_1}+ \frac{J_2}{\omega_2} \, .

645: \ee

646: \subsection{The special case: $\omega_1=\omega_2$}

647: \label{sect:simplesol}

648: It is instructive to first discuss the particularly simple special case of a circular 

649: string, where $\omega_1=\omega_2$ and \eqn{F1} does not apply. This

650: will turn out to be a limiting case of the $q>1$ scenario. For $\omega_{21}=0$

651:  the equation of motion for $\psi(\sigma)$ \eqn{EOMpsi} immediately yields 

652: \be

653: \psi''=0 \qquad \Rightarrow \qquad \psi(\sigma)=n\, \sigma

654: \ee

655: with

656: $n$ being the integer winding number of this ciruclar string $\psi(\sigma+2\pi)=

657: \psi(\sigma)+2\pi\, n$.

658: In this case the Virasoro constraints \eqn{VirConstr} yield 

659: $\kappa=\sqrt{n^2+{\omega_1}^2}$.

660: A little bit of algebra quickly shows that the energy $E$ may be reexpressed as

661: a function of $J_1=J_2$ and reads

662: \be

663: \label{C1en}

664: E=2\, J_1\, \sqrt{1+\frac{n^2\, \lambda}{4\, {J_1}^2}}

665: \ee

666: which is analytic in $\frac{\lambda}{{J_1}^2}$! This amounts to an all loop prediction for

667: the dual gauge theory scaling dimension in the BMN limit $J_1\to\infty$ with $\lambda/{J_1}^2$

668: fixed, quite similar to the result for the plane-wave superstring discussed above.

669: 

670: Let us now discuss the folded and circular string solutions in turn.

671: 

672: \subsection{The folded string: $q\leq 1$}

673: \label{subsect:3.2}

674: 

675: In the folded case $J_1$ may be expressed in terms of an elliptic

676: integral\footnote{Our conventions are ($x<1$) 

677: $$

678: E(x):=\int_0^{\pi/2}d\psi \, \sqrt{1-x\, \sin^2\psi} \qquad

679: K(x):=\int_0^{\pi/2}d\psi \, \frac{1}{\sqrt{1-x\, \sin^2\psi}}\, .

680: $$} by substituting (using \eqn{stringpendulum})

681: \begin{equation}

682: d\sigma=\frac{d\psi}{\omega_{21}\sqrt{q-\sin^2\psi}}

683: \end{equation}

684: into \eqn{J1J2def}

685: and performing some elementary transformations to find ($q=\sin^2\psi_0$) 

686: \be

687: \label{F2}

688: J_1=\frac{\sqrt{\lambda}\, \omega_1}{2\pi}\, 4\, \int_0^{\psi_0}

689:   d\psi\, \frac{\cos^2\psi}{\omega_{21}\, \sqrt{\sin^2\psi_0 -\sin^2\psi}}

690: = \frac{2\, \sqrt{\lambda}\, \omega_1}{\pi\, \omega_{21}}\,

691: E(\sin^2\psi_0)\, ,

692: \ee

693: where we only need to integrate over one quarter of the folded string

694: due to symmetry considerations (see figure \ref{fig:foldedstring}). 

695: Additionally we have

696: \be

697: \label{F4}

698: 2\pi= \int_0^{2\pi}d\sigma = 4\,

699: \int_0^{\psi_0}\frac{d\psi}{\omega_{21}\, \sqrt{\sin^2\psi_0

700:     -\sin^2\psi}} = \frac{4}{\omega_{21}}\, K(\sin^2\psi_0) \, .

701: \ee

702: The four equation \eqn{F1}, \eqn{F3}, \eqn{F2} and \eqn{F4} may then be

703: used to eliminate the parameters of our solution $\kappa$,

704: $\omega_1$ and $\omega_2$. For this rewrite \eqn{F2} and \eqn{F4} as ($\kappa=E/\sqrt{\lambda}$)

705: \be

706: \label{omegas}

707: \frac{\omega_1}{\omega_{21}}=\frac{\pi}{2\,\sqrt{\lambda}}\, \frac{J_1}{E(q)}\, ,\qquad

708: \frac{\kappa}{\omega_{21}} = \frac{\pi}{2\, \sqrt{\lambda}}\, \frac{E}{K(q)} \, ,

709: \ee

710: and use \eqn{F3} to deduce

711: \be

712: \label{omegas2}

713: \frac{\omega_2}{\omega_{21}}= \frac{\pi}{2\, \sqrt{\lambda}}\, \frac{J_2}{K(q)-E(q)} \, .

714: \ee

715: Then the Virasoro constraint equation \eqn{F1} and the identity $1=({\omega_2}^2-{\omega_1}^2)/{\omega_{21}}^2$

716: yield the two folded string equations

717: \be

718: \frac{4\, q\, \lambda }{\pi^2}= \frac{E^2}{K(q)^2} -

719: \frac{{J_1}^2}{E(q)^2}\, , \qquad

720: \frac{4\, \lambda}{\pi^2} = \frac{{J_2}^2}{(K(q)-E(q))^2} -

721: \frac{{J_1}^2}{E(q)^2}\, ,

722: \label{Fstringeqs}

723: \ee

724: which implicitly define the sought after energy function $E=E(J_1,J_2)$ upon further elimination of

725: $q$. This is achieved by assuming an analytic behavior of

726: $q$ and $E$ in the BMN type limit of large total angular 

727: momentum $J:=J_1+J_2\to\infty$ with $\lambda/J^2$ held fixed

728: \begin{align}

729: \label{BMNexp}

730: q&=q_0+\frac{\lambda}{J^2}\, q_1 + \frac{\lambda^2}{J^4}\, q_2 +\ldots

731: \nn\\

732: E&= J\, \Bigl (E_0 + \frac{\lambda}{J^2}\, E_1 +

733: \frac{\lambda^2}{J^4}\, E_2 +\ldots \Bigr ) \, .

734: \end{align}

735: Plugging these expansions into \eqn{Fstringeqs} one can solve for the $q_i$ and

736: $E_i$ iteratively. At leading order $E_0=1$ (as it should from the

737: dual gauge theory perspective) and $q_0$ is implicitly determined through

738: the ``filling fraction'' $J_2/J$

739: \be

740: \label{FE1}

741:  \frac{J_2}{J} = 1- \frac{E(q_0)}{K(q_0)} \, .

742: \ee

743: The first non-trivial term in the energy is then  expressed in terms of

744: $q_0$ through

745: \be

746: \label{FEprediction}

747: E_1=\frac{2}{\pi^2}\, K(q_0)\, \Bigl ( \, E(q_0) - (1-q_0)\, K(q_0)\,

748: \Bigr )

749: \ee

750: yielding a clear prediciton for one-loop gauge theory. The higher

751: order $E_i$ can then also be straightforwardly

752: obtained. To give a

753: concrete example, let us evaluate $E$ for the first few orders in

754: $\lambda/J^2$ in the ``half filled'' case $J_1=J_2$:

755: \be

756: E=2\, J_1\, ( 1 + \frac{0.71}{8}\, \frac{\lambda}{{J_1}^2} -

757: \frac{1.69}{32}\,  \frac{\lambda^2}{{J_1}^4} + \ldots )

758: \ee

759: which shows that the energy of the folded string configuration is

760: smaller than the one of the closed configuration of \eqn{C1en} for

761: single winding $n=1$.

762: 

763:  

764: \subsection{The circular string: $q> 1$}

765: \label{subsect:3.3}

766: 

767: For $q>1$ the string does not fold back onto itself and extends around

768: the full circle of $\psi\in[0,2\pi]$

769: as $\sigma$ runs from $0$ to $2\pi$. 

770: The initial expression for \eqn{F2} now changes only marginally to

771: \be

772: J_1=\frac{\sqrt{\lambda}\, \omega_1}{2\pi}\, \int_0^{2\pi}

773:   d\psi\, \frac{\cos^2\psi}{\omega_{21}\, \sqrt{q -\sin^2\psi}} \, .

774: \ee

775: Elementary transformations yield

776: \be

777: \label{C2}

778: J_1= \frac{2\, \sqrt{\lambda}\, \omega_1}{\pi\, \omega_{21}}\Bigr [

779: \, \frac{1-q}{\sqrt{q}}\, K(q^{-1}) + \sqrt{q}\, E(q^{-1})\, \Bigr ] \, .

780: \ee

781: Analogously \eqn{F4} now becomes

782: \be

783: \label{C4}

784: 2\pi = \int_0^{2\pi}d\sigma = 

785: 4\,

786: \int_0^{\pi/2}\, \frac{d\psi}{\omega_{21}\, \sqrt{\sin^2\psi_0

787:     -\sin^2\psi}} = \frac{4}{\omega_{21}\,\sqrt{q}}\, K(q^{-1}) \, .

788: \ee

789: The corresponding relations to \eqn{omegas} and \eqn{omegas2} then take  the form

790: \begin{align}

791: \frac{1}{\omega_{21}} &= \frac{\pi}{2}\, \frac{\sqrt{q}}{K(q^{-1})} \, , \qquad

792: \frac{\omega_1}{\omega_{21}}=\frac{\pi}{2\,\sqrt{\lambda}}\, 

793: \frac{\sqrt{q}\, J_1}{(1-q)\, K(q^{-1})+

794: q\, E(q)}\, ,\nn\\

795: \frac{\omega_2}{\omega_{21}}&= \frac{\pi}{2\, \sqrt{\lambda}}\, 

796: \frac{J_2}{\sqrt{q}[\, K(q^{-1})-E(q^{-1})\, ]} \, .

797: \end{align}

798: From these one deduces in complete analogy to \eqn{Fstringeqs} the two

799: circular string equations

800: \begin{align}

801: \frac{4\, \lambda }{\pi^2}&= \frac{E^2}{K(q^{-1})^2} -

802: \frac{{J_1}^2}{[\, (1-q)\, K(q^{-1})+q\, E(q^{-1})\, ]^2}\nn\\

803: \frac{4\, q\, \lambda}{\pi^2} &= \frac{{J_2}^2}{[\,K(q^{-1})-E(q^{-1})\,]^2} -

804: \frac{q^2\,{J_1}^2}{[\,(1-q)\, K(q^{-1})+q\, E(q^{-1})\, ]^2} \, ,

805: \label{Cstringeqs}

806: \end{align}

807: which encode the energy relation $E=E(J_1,J_2)$ upon elimination of $q$.

808: In order to do this we again make an analytic ansatz in $\lambda/J^2$ for $q$ and $E$ 

809: as we did in \eqn{BMNexp}. This yields

810: the following implicit expression for $q_0$ in terms of the filling fraction $J_2/J$

811: \be

812: \label{CS1}

813:  \frac{J_2}{J} = q_0\, \Bigl ( 1- \frac{E(q_0^{-1})}{K(q_0^{-1})} \Bigr )\, ,

814: \ee

815: The first two energy terms in the $\lambda/J^2$ expansion then take the form

816: \be

817: \label{CS2}

818: E_0=1\, , \qquad E_1= \frac{2}{\pi^2}\, E(q_0^{-1})\, K(q_0^{-1}) \, ,

819: \ee

820: which again gives a clean prediction for the dual gauge theory scaling dimensions

821: at one-loop.

822: 

823: In figure~\ref{Eplot} we have plotted the energies of the folded and circular string solutions

824: against the filling fraction $J_2/J$. As expected from the ``string pendulum'' picture of 

825: \eqn{stringpendulum} the folded string solution has lower energy for fixed filling

826: fraction $J_2/J$: The pendulum

827: does not perform a full turn, but oscillates back and forth. Note that the folded string

828: approaches $E_1=0$ in the limit $J_2/J\to 0$, which is the rotating point-particle solution

829: of subsection~\ref{sect:particle}.  The quantum fluctuations about it correspond to 

830: the plane-wave string domain. Note also that the simplest circular string solution with $\omega_1=\omega_2$

831: discussed in subsection~\ref{sect:simplesol} is the minimum ($J_2/J=0.5$) of the full circular

832: string solution for $n=1$. 

833: \begin{figure}[t]

834: \begin{center}

835: \includegraphics[width=9cm]{spinningstringsplot.eps}

836: \end{center}

837: \caption{The one-loop energies of the folded (dark) and circular (light) string solutions

838: plotted against the filling fraction $J_2/J$. The dashed curve is the mirrored folded string

839: solution where one interchanges $J_1\leftrightarrow J_2$.} 

840: \label{Eplot}

841: \end{figure}

842: 

843: These classical string energies will be reproduced in a dual one-loop gauge theory computation

844: in section~\ref{sec:gt}.

845: 

846: \subsection{Further Developments}

847: 

848: The discussed explicit solutions of the bosonic string on a $\Real\times S^3$

849: background are the first simple elements of the general set of

850: classical solutions of the bosonic $AdS_5\times S^5$  string

851: considered in the literature. 

852: Earlier examples not discussed in the text are 

853: \cite{hep-th/0204051,hep-th/0205244,hep-th/0209047}.

854: The bosonic part of the classical string action consists of a $SO(6)$

855: and a $SO(2,4)$ $\sigma$-model augmented by the conformal gauge

856: (Virasoro) constraints. The $O(p,q)$ $\sigma$-models have been known 

857: to be integrable for a long time \cite{CMPHA.46.207,NUPHA.B137.46,Zakharov:1973pp,Faddeev:1985qu} and

858: one would expect this to remain true also once one imposes the

859: conformal gauge constraints. And indeed in

860: \cite{hep-th/0307191,hep-th/0311004} more complicated solutions of the

861: $AdS_5\times S^5$ string were constructed by reducing the system to

862: the so-called Neumann integrable system through a suitable ansatz.

863: The solutions of \cite{hep-th/0307191,hep-th/0311004}

864: involve nonvanishing values for all spins and

865: angular momenta $(S_1,S_2,J_1,J_2,J_3)$. However, they do not

866: generically display an analytic behavior in the effective coupling

867: constant $\lambda'$. This is true only for configurations with at

868: least one large charge $J_i$ on the $S^5$.

869: 

870: In the context of the $O(p,q)$ $\sigma$-models integrability is based

871: on the existence of a Lax pair, a family of flat connections on the 2d string

872: worldsheet, giving rise to an infinite number of conserved

873: charges. These were first discussed in the context of the bosonic string

874: in \cite{hep-th/0206103} and for the full superstring in \cite{hep-th/0305116}.

875: The Lax pair for the string was put to use in \cite{hep-th/0402207} for

876: string configurations on $\Real\times S^3$ -- the sector we also considered in

877: the above. These investigations led to the construction of an

878: underlying algebraic curve parametrizing the solutions.

879: This enabled the authors of \cite{hep-th/0402207} to write down an integral equation of Bethe type 

880: yielding the associated energies of the solutions. Very similar equations

881: will appear below in our discussion in section~\ref{subsect:thermo}

882: on the thermodynamic limit of the gauge theory Bethe equations. 

883: The extraction of these integral equations from the string

884: $\sigma$-model then allows for a direct comparison to the gauge

885: theory Bethe equations. On this level of formalization, there is no

886: need to compare explicit solutions any longer -- as we are doing here for

887: pedagogical purposes. This construction based on an underlying algebraic

888: curve makes full use of the technology of integrable systems and has been nicely 

889: reviewed by Zarembo in \cite{hep-th/0411191}.

890: 

891: In the very interesting paper \cite{hep-th/0406256} these continuum string

892: Bethe equations were boldly discretized  leading to a conjectured set of

893: Bethe equations for the {\sl quantum} spectrum of the string. 

894: This proposal has been verified by comparing it to known quantum data 

895: of the $AdS_5\times S^5$ string: 

896: The near plane-wave spectrum of the superstring of 

897: \cite{hep-th/0307032,hep-th/0404007,hep-th/0405153,hep-th/0407096}

898: as well as the expected \cite{hep-th/9802109} generic scaling of the string energies with

899: $\lambda^{1/4}$ in the strong coupling limit agree with the

900: predictions of the quantum string Bethe equations. But there is more

901: quantum data for the $AdS_5\times S^5$ string available: In a series of papers by

902: Tseytlin, Frolov and collaborators one-loop corrections on the string

903: worldsheet to the energies of various spinning string solutions have been computed 

904: \cite{hep-th/0306130,hep-th/0408187,hep-th/0501203}. The one loop

905: correction for a circular string moving in $AdS_3\times S^1 \subset

906: AdS_5\times S^5$ obtained in \cite{hep-th/0501203} was recently

907: compared \cite{hep-th/0507189} to the

908: result obtained from the proposed quantum string Bethe equations of

909: \cite{hep-th/0406256}. 

910: The authors of \cite{hep-th/0507189} find agreement 

911: when they expand the results in $\lambda'$ (up to third order), but

912: disagreements emerge in different

913: limits (where $\lambda'$ is not small). The interpretation of this

914: result is unclear at present.

915: Finally the proposed

916: quantum string Bethe equations of \cite{hep-th/0406256} can also 

917: be microscopically attributed

918: to a $s=1/2$ spin chain model with long-range interactions up

919:  to (at least) order five in a small $\lambda$ expansion \cite{hep-th/0409054}.

920: 

921: The technically involved construction of algebraic curves solving the

922: classical $\Real\times S^3$ string $\sigma$-model 

923: has subsequently been generalized to larger sectors: In

924: \cite{hep-th/0410253} to $\Real\times S^5$ (or $SO(6)$ in gauge theory language) configurations,

925: in \cite{hep-th/0410105} to $AdS_3\times S^1$ (or $SL(2)$) string configurations

926: and finally in \cite{hep-th/0502226} to superstrings propagating in

927: the full $AdS_5\times S^5$ space.

928: 

929: There has also been progress on a number of possible paths towards the

930: true quantization of the classical integrable model of the \AdSS string in the works

931: \cite{hep-th/0410282,hep-th/0411089,hep-th/0502240,hep-th/0504144,hep-th/0409159,hep-th/0411170}, 

932: however, it is fair to say that this problem remains currently unsolved. 

933: 

934: \section{The dual gauge theory side}

935: \label{sec:gt}

936: 

937: Let us now turn to the identification of the folded

938: and circular string solutions in the dual gauge theory.

939: 

940: Our aim is to reproduce the obtained energy functions $E_1(J_1,J_2)$ 

941:  plotted in figure \ref{Eplot} 

942: from a dual gauge theory computation

943: at one-loop. For this we need to identify the gauge theory operators,

944: which are dual to the spinning strings on $\Real\times S^3$.

945: As here $J_2=0=S_1=S_2$ the relevant operators will

946: be built from the two complex scalars

947: $Z:=\phi_1+i\,\phi_2$ and $W:=\phi_3+i\,\phi_4$ with a total number of

948: $J_1$ $Z$-fields and $J_2$ $W$-fields, i.e.

949: \be

950: {\cal O}^{J_1,J_2}_\alpha= \Tr[\, Z^{J_1}\, W^{J_2}\, ] +\ldots \, ,

951: \label{calO} 

952: \ee

953: where the dots denote suitable permutations of the $Z$ and $W$ to be discussed. 

954: An operator of the form \eqn{calO}

955: may be pictured as a ring of black (``$Z$'') and red (``$W$") beads -- or equivalently as a

956: configuration of an $s=1/2$ quantum spin chain, where $W$ corresponds

957: to the state

958: $|\uparrow\,\rangle$ and $Z$ to $|\downarrow\,\rangle$. 

959: $$

960: \Tr[Z W^2 Z W^4 ] \quad \Leftrightarrow \quad

961: \raisebox{0.5cm}{\rotatebox{-90}{{\epsfysize=1.4cm\epsfbox{flatcirc0.eps}}}}

962: \quad \Leftrightarrow \quad 

963: |\downarrow\,\uparrow\,\uparrow\,\downarrow\,\uparrow\,\uparrow

964: \,\uparrow\,\uparrow\,\rangle_{\rm cyclic}

965: $$

966: 

967: How does one compute the associated scaling dimensions at (say) one

968: loop order for $J_1,J_2\to\infty$? Clearly one is facing a huge operator mixing

969: problem as all  ${\cal O}^{J_1,J_2}_i$ with arbitrary permutations of $Z$'s and $W$'s

970:  are degenerate at tree level where $\Delta^{{\cal O}^{J_1,J_2}_i}_0= J_1+J_2$. 

971: 

972: A very efficient tool to deal with this problem is 

973: the dilatation operator ${\cal D}$, which was introduced in 

974: \cite{hep-th/0212269,hep-th/0303060}. It acts on the trace

975: operators ${\cal O}^{J_1,J_2}_\alpha$ at a fixed space-time point $x$ and its eigenvalues are

976: the scaling dimensions $\Delta$

977: \be

978: {\cal D}\circ {\cal O}^{J_1,J_2}_\alpha(x) = \sum_{\beta}\, {\cal

979:   D}_{\alpha \beta}\, {\cal O}^{J_1,J_2}_\beta(x) \, .

980: \ee 

981: The dilatation operator is constructed in such a fashion as to attach the relevant 

982: diagrams to the open legs of the

983: \begin{figure}[t]

984: \begin{center}

985: $\displaystyle

986: {\cal D}\circ \raisebox{-.5cm}{\epsfysize=1.2cm\epsfbox{flatcirc0.eps}}

987: =

988: \raisebox{-.5cm}{\epsfysize=1.2cm\epsfbox{fcirc1.eps}}\, +\,

989: \raisebox{-.5cm}{\epsfysize=1.2cm\epsfbox{fcirc2.eps}}\, +\,

990: \raisebox{-.5cm}{\epsfysize=1.2cm\epsfbox{fcirc3.eps}}\,+\ldots

991: =

992: \sum_p N\, c_p\,\underbrace{\raisebox{-.3cm}{\epsfysize=1.2cm\epsfbox{fcircp.eps}}}_{\rm planar}

993: + \, \,d_p 

994: \underbrace{\raisebox{-.6cm}{\epsfysize=1.5cm\epsfbox{fcircp2.eps}}}_{\rm nonplanar}\nn

995: $

996: \end{center}

997: \caption{The action of the dilatation operator on a trace operator.} 

998: \label{dilopact}

999: \end{figure}

1000: ``incoming" trace operators as depicted in figure~\ref{dilopact} and

1001: may be computed in perturbation theory 

1002: \begin{equation}

1003: {\cal  D}

1004: =\sum_{n=0}^\infty  {\cal D}^{(n)} \, ,

1005: \end{equation}

1006: where ${\cal D}^{(n)}$ is of order $\gym^{2n}$.

1007: For the explicit computation of the one-loop piece ${\cal D}^{(1)}$ see e.g.~the

1008: review \cite{hep-th/0307101}, where the concrete relation to 

1009: two-point functions is also explained. In our `minimal' $SU(2)$ sector

1010: of complex scalar fields $Z$ and $W$ it takes the rather simple form

1011: \be

1012: {\cal D}^{(0)} = \Tr(Z\check Z+W\check W),\qquad 

1013: {\cal D}^{(1)}  =

1014: -\frac{g^2_{\rm YM}}{8\pi^2}\, \Tr\, [Z,W]\, [\check{Z},{\check{W}}] \, ,

1015: \qquad \mbox{where}\quad  \check{Z}_{ij}:=\frac{d}{dZ_{ji}} \, .

1016: \label{D0n1}

1017: \ee

1018: Note that the tree-level piece ${\cal D}^{(0)}$ simply measures the length of the incident

1019: operator (or spin chain) $J_1+J_2$.

1020: The eigenvalues of the dilatation operator then yield the 

1021: scaling dimensions we are looking for -- diagonalization of ${\cal D}$

1022: solves the mixing problem.

1023: 

1024: We shall be exclusively interested in the planar contribution to ${\cal D}$, as 

1025: this sector of the gauge theory corresponds to the ``free"

1026: (in the sense of $g_s=0$) \AdSS string. For this it is important to realize that the

1027: planar piece of ${\cal D}^{(1)}$ only acts on two neighboring fields

1028: in the chain of $Z$'s and $W$'s. This may be seen by evaluating explicitly the action

1029: of ${\cal D}^{(1)}$ on two fields $Z$ and $W$ separated by arbitrary matrices

1030: $\cal{A}$ and $\cal{B}$

1031: \be

1032: \Tr\ [Z,W]\, [\check{Z},{\check{W}}]\circ

1033: \Tr (Z\,{\cal A}\,  {W}\, {\cal B}) =

1034: -\Tr ({\cal A})\, \Tr ( [Z,{W}]\, {\cal B})+ 

1035: \Tr ({\cal B})\, \Tr ( [Z,{W}]\, {\cal A}) \, .

1036: \ee

1037: Clearly there is an enhanced contribution when ${\cal A}=1$ or ${\cal

1038:   B}=1$, i.e.~$Z$ and $W$ are nearest neighbours on the spin chain.

1039: From the above computation we learn that

1040: \be

1041: {\cal D}^{(1)}_{\rm planar} = \frac{\lambda}{8\pi^2}\, \sum_{i=1}^L\,

1042: ( \eins_{i,i+1} - P_{i,i+1} )

1043: \ee

1044: where $P_{i,j}$ permutes the fields (or

1045: spins) at sites $i$ and $j$ and periodicity $P_{L,L+1}=P_{1,L}$ is understood. 

1046: Remarkably, as noticed by Minahan and Zarembo

1047: \cite{hep-th/0212208}, this spin-chain operator is the Heisenberg

1048: XXX${}_{1/2}$ quantum spin chain Hamiltonian, which is the prototype of an 

1049: integrable spin-chain. Written in terms of the Pauli matrices

1050: $\vec\sigma_i$ acting on the spin at site $i$ one finds

1051: \be

1052: {\cal D}^{(1)}_{\rm planar}= \frac{\lambda}{8\pi^2}{\cal H}_{{\rm

1053:     XXX}_{1/2}} = \frac{\lambda}{4\pi^2}\, \sum_{i=1}^L\,

1054: (\frac{1}{4} -\vec\sigma_i\cdot\vec\sigma_{i+1}) \, .

1055: \ee

1056: Due to the positive sign in front of the sum, the spin chain is

1057: ferromagnetic and its ground state is $|\down\down\ldots\down\,\rangle_{\rm cyclic}

1058: \Leftrightarrow \Tr(Z^{L})$: The gauge dual of the rotating point particle of

1059: section \ref{sect:particle}. Excitations of the ground state are

1060: given by spin flips or ``magnons''. 

1061: Note that a one-magnon excitation

1062: $|\down\ldots\down\up\down\ldots \down\,\rangle_{\rm cyclic}$ has vanishing energy due

1063: to the cyclic property of the trace, it corresponds to a zero mode plane-wave string

1064: excitation $\alpha_0^{\dagger}\,|0\rangle$. Two-magnon excitations are the first stringy

1065: excitations which are dual to the $\alpha_n^{\dagger}\alpha_{-n}^\dagger\, |0\rangle$

1066: state of the plane-wave string in the BMN limit.

1067: 

1068: The integrability of the spin-chain amounts to the existence of $L-1$ higher charges $Q_k$ which

1069: commute with the Hamiltonian (alias dilatation operator) and amongst themselves, 

1070: i.e.~$[Q_k,Q_l]=0$. Explicitly the first two charges of the Heisenberg chain are given by

1071: \be

1072: Q_2:={\cal H}_{{\rm XXX}_{1/2}}\qquad

1073: Q_3=\sum_{i=1}^L  (\vec\sigma_i\times \vec\sigma_{i+1})\cdot

1074: \vec\sigma_{i+2} \, .

1075: \ee

1076: The explicit form of all the higher $Q_k$ may be found in \cite{hep-th/9403149}. Note that

1077: $Q_k$ will involve up to $k$ neighboring spin interactions. 

1078: 

1079: \subsection{The coordinate Bethe ansatz}

1080: \label{sect:coordBethe}

1081: 

1082: We now discuss the ansatz that enabled Bethe to diagonalize the Heisenberg model

1083: in 1931 \cite{Bethe:1931hc}\footnote{For a nice and detailed review on this topic

1084: see \cite{muller}. The technology of the

1085: algebraic Bethe ansatz is reviewed in \cite{hep-th/9605187}.}. For this we will for 

1086: the moment drop the cyclicity 

1087: constraint imposed on us from the underlying trace structure of the gauge theory

1088: operators and treat a general non-cyclic, but periodic, spin

1089: chain. The vacuum state of the Heisenberg chain is then given by

1090: $|\down\ldots\down\rangle$.

1091: Let $|x_1,x_2,\ldots x_J\rangle$ with $x_1<x_2<\ldots <x_J$ 

1092: denote a state of the chain with up-spins (magnons)

1093: located at sites $x_i$, i.e.~ $|1,3,4\rangle_{L=5} =|\down\up\down\down\up\,\rangle$. 

1094: It is useful to think of these spin flips as particles located at the sites $x_i$. Note that the

1095: Hamiltonian preserves the magnon or particle number.

1096: 

1097: The one magnon sector is then trivially diagonalized by Fourier transformation

1098: \begin{align}

1099: |\psi(p_1)\rangle &:= \sum_{x=1}^L\, e^{i\, p_1\, x}\, |x\rangle ,\qquad

1100: \mbox{with} \quad Q_2\, |\psi(p_1)\rangle = 4\, \sin^2(\frac{p_1}{2})

1101: \, |\psi(p_1)\rangle\\

1102: &\mbox{where} \quad Q_2= \sum_{i=1}^L\, ( \eins_{i,i+1} - P_{i,i+1} )

1103: \end{align}

1104: as $2-e^{ip}-e^{-ip}=4\,\sin^2(p/2)$. The periodic boundary condidtions

1105: require the one-magnon momenta to be quantized $p_1=\frac{2\pi\,

1106:   k}{L}$ with $k\in\Zint$.

1107: 

1108: Next consider a general two-magnon state of the form

1109: \be

1110: |\psi(p_1,p_2) \rangle =\sum_{1\leq x_1<x_2\leq L} \psi(x_1,x_2)\, |x_1,x_2\rangle\, .

1111: \ee

1112: with a two-particle wave-function $\psi(x_1,x_2)$. 

1113: The ``position space'' Schr\"odinger equation following from 

1114: $\sum_{i=1}^L(1-P_{i,i+1})\,|\psi(p_1,p_2)\rangle = {E_2}\, |\psi(p_1,p_2) \rangle$

1115: then leads to two sets of equations, depending on whether the particles lie next to each other

1116: or not:

1117: \begin{align}

1118: x_2&>x_1+1 \quad  &{E_2}\, \psi(x_1,x_2) =&\, 2\,\psi(x_1,x_2) -\psi(x_1-1,x_2)-\psi(x_1+1,x_2)\nn\\

1119:               &   &                            +&\,2\,\psi(x_1,x_2) -\psi(x_1,x_2-1)-\psi(x_1,x_2+1)\nn\\

1120: & &&\label{one}\\

1121: x_2&=x_1+1   &{E_2}\, \psi(x_1,x_2) =& \,2\,\psi(x_1,x_2) -\psi(x_1-1,x_2)-\psi(x_1,x_2-1)\, .

1122: \label{two}

1123: \end{align}

1124: $E_2$ is the eigenvalue of $Q_2$ and related to the gauge theory scaling dimensions

1125: as $\Delta= L +\frac{\lambda}{8\pi^2}\, E_2 + {\cal O}(\lambda^2)$.

1126: The above equations can be fulfilled by a superposition ansatz with an incoming and outgoing plane

1127: wave (Bethe's ansatz)

1128: \be

1129: \label{Bethesansatz}

1130: \psi(x_1,x_2)= e^{i({p_1}\, x_1 + {p_2}\, x_2)} + {S({p_2},{p_1})}\,

1131: e^{i({p_2}\, x_1 + {p_1}\, x_2)} \,  ,

1132: \ee

1133: where $S(p_1,p_2)$ denotes the S-matrix of the scattered particles. Note that in the second

1134: term describing the scattered contribution the two particles have simply exchanged their momenta.

1135: One easily sees that

1136: \eqn{one} is fulfilled for an arbitrary $S(p_2,p_1)$ yielding the energy as a sum of one-particle

1137: energies

1138: \be

1139: E_2=4\, \sin^2 (\frac{p_1}{2}) + 4\, \sin^2 (\frac{p_2}{2}) \, .

1140: \label{2ME}

1141: \ee

1142: Eq.~\eqn{two} then determines the S-matrix to be of the form

1143: \be

1144: S(p_1,p_2)=\frac{\varphi(p_1)-\varphi(p_2)+i}

1145: {\varphi(p_1)-\varphi(p_2)-i}\qquad \mbox{with}\quad  \varphi(p)=\ft 1 2 \cot(\ft p 2)\, .

1146: \label{S-matrix}

1147: \ee

1148: Note that $S(p_1,p_2)^{-1}=S(p_2,p_2)$.

1149: This solves the infinite length chain. For a finite chain the momenta $p_i$ are no longer

1150: arbitrary continuous quantities, but become discrete through the periodic boundary condition

1151: \be

1152: \psi(x_1,x_2)=\psi(x_2,x_1+L)\, .

1153: \ee

1154: This in turn then leads to the Bethe equations for the two magnon problem

1155: \be

1156: e^{ip_1 L}= S(p_1,p_2) \quad \mbox{and} \quad e^{ip_2 L}= S(p_2,p_1)

1157: \label{2MBE}

1158: \ee

1159: implying $p_1+p_2=2\pi m$ with an arbitrary integer $m$. The solutions of the algebraic equations 

1160: \eqn{2MBE} for $p_1$ and $p_2$

1161: then determine the corresponding energies by plugging the resulting quasi-momenta $p_i$

1162: into \eqn{2ME}.

1163: 

1164: The magic of integrability now is that this information is all that is needed to solve the general

1165: $M$-body problem! This phenomenon is known as factorized scattering: The multi-body

1166: scattering process factorizes into a sequence of two-body interactions under which

1167: two incoming particles of momenta $p_i$ and $p_j$ scatter off each other {\it elastically}

1168: with the S-matrix $S(p_j,p_i)$, thereby simply exchanging their momenta. 

1169: That is the $M$-body wavefunction takes the form \cite{Bethe:1931hc,muller}

1170: \be

1171: \psi(x_1,\ldots ,x_M) = \sum_{P\in{\rm Perm}(M)} 

1172: \exp\Bigl [ i\sum_{i=1}^M p_{P(i)}\, x_i  +\frac i 2 \sum_{i<j} \theta_{{P(i)P(j)}}\Bigr ]

1173: \label{bethewf}

1174: \ee

1175: where the sum is over all $M!$ permutations of the labels

1176: $\{1,2,\ldots, M\}$ and the phase shifts $\theta_{ij}=-\theta_{ji}$ are related to

1177: the S-matrix \eqn{S-matrix} by

1178: \be

1179: S(p_i,p_j)=\exp[i\theta_{ij}]\, .

1180: \ee

1181: The $M$-magnon Bethe ansatz then yields the set of $M$ Bethe equations

1182: \be

1183: e^{ip_k L}=\prod_{i=1,i\neq k}^M S(p_k,p_i)

1184: \label{BE}

1185: \ee

1186: with the two-body S-matrix of \eqn{S-matrix} and the additive energy expression 

1187: \be

1188: \label{E21loop}

1189: E_2= \sum_{i=1}^M 4\, \sin^2 (\frac{p_i}{2})\, .

1190: \ee

1191: In order to reinstate the cyclicity of the trace condition one needs to further impose the 

1192: constraint of a total vanishing momentum

1193: \be

1194: \sum_{i=1}^M p_i = 0\, .

1195: \label{zerop}

1196: \ee

1197: 

1198: As an example let us diagonalize the two magnon problem exactly. Due to \eqn{zerop} we have

1199: $p:=p_1=-p_2$ and the Bethe equations \eqn{2MBE} reduce to the single equation

1200: \be

1201: e^{ipL} = \frac{\cot\frac{p}{2}+i}{\cot\frac{p}{2}-i} = e^{ip} \quad

1202: \Rightarrow \quad e^{ip(L-1)}=1 \quad\Rightarrow \quad

1203: p=\frac{2\pi\, n}{L-1} \, .

1204: \ee

1205: The energy eigenvalue then reads 

1206: \be

1207: E_2= 8 \sin^2\left (\frac{\pi\, n}{L-1}\right)\quad  \stackrel{L\to\infty}{\to}\quad

1208:  8\pi^2\, \frac{n^2}{L^2} \, ,

1209: \ee

1210: which upon reinserting the dropped prefactor of $\frac{\lambda}{8\pi^2}$ yields

1211: the one-loop scaling dimension $\Delta^{(1)}=\frac{\lambda}{\pi^2}\, \sin^2(\frac{\pi\, n^2\,}{L^2})$

1212: of the two-magnon operators \cite{hep-th/0211032,hep-th/0212208}

1213: \be

1214: {\cal O}_n^{(J,2)} = \sum_{p=0}^J \cos\left( \pi\, n\, \frac{2p+1}{J+1}\right )\,

1215: \Tr(W\, Z^p\, W\, Z^{J-p}) \, .

1216: \ee

1217: In the BMN limit $N,J\to\infty$ with $\lambda/J^2$ fixed the scaling dimension takes the

1218: famous value $\Delta^{(1)}=n^2\,\lambda/J^2$, corresponding to the first term in the 

1219: expansion of the level-two energy spectrum of the plane-wave superstring

1220: $E_{\rm light-cone}=\sqrt{1+n^2\, \lambda/J^2}$ \cite{hep-th/0202021}.

1221: 

1222: Hence from the viewpoint of the spin chain the plane-wave limit corresponds to a 

1223: chain of diverging length $L>>1$ carrying a finite number of magnons $M$, which are nothing

1224: but the gauge duals of the oscillator excitations of the plane-wave superstring.

1225: 

1226: \subsection{The thermodynamic limit of the spin-chain}

1227: \label{subsect:thermo}

1228: 

1229: In order to make contact with the spinning string solution discussed in section 

1230: \ref{section:spinning_string_solutions} we will now consider the thermodynamic limit

1231: of the spin chain in which the length $L$ {\it and} the number of magnons $M$

1232: become large. This is necessary as the classical string solutions 

1233: only limit to the true quantum result 

1234: in the BMN type limit $J_2,J\to\infty$ with the filling fraction $J_2/J$ held fixed (here

1235: $J_2=M$ and $J=L$). This thermodynamic groundstate solution of the gauge theory Bethe equations  

1236: was found in \cite{hep-th/0306139,hep-th/0308117,hep-th/0310182,hep-th/0310188}

1237:  which we closely follow.

1238: 

1239: For this it is useful to reexpress the Bethe equations \eqn{BE} in terms of the Bethe

1240: roots $u_k$ related to the momenta via

1241: \be

1242: u_k=\frac{1}{2}\, \cot\frac{p_k}{2}

1243: \ee

1244: for which the Bethe equations \eqn{BE} and the momentum constraint \eqn{zerop} become

1245: \be

1246: \label{TBE}

1247: \left ( \frac{u_i+i/2}{u_i-i/2}\right )^L = \prod_{k\neq i}^M 

1248: \frac{u_i-u_k+i}{u_i-u_k-i}\, ,

1249: \qquad \prod_{i=1}^M \frac{u_i+i/2}{u_i-i/2}= 1\, .

1250: \ee

1251: The energy then is

1252: \be

1253: \label{TE}

1254: Q_2=\sum_{i=1}^M \frac{1}{{u_i}^2 + \ft 1 4}\, .

1255: \ee

1256: The momentum constraint can be satisfied by considering symmetric root distributions

1257: of the form $(u_i,-u_i,u_i^\ast,-u_i^\ast)$.

1258: The thermodynamic limit is now obtained by first taking the logarithm of \eqn{TBE}

1259: \be

1260: L\,\ln\left ( \frac{u_i+i/2}{u_i-i/2}\right ) 

1261: = \sum_{k=1\, (k\neq i)}^M \ln \left ( \frac{u_i-u_k+i}{u_i-u_k-i} \right )

1262: - 2\, \pi\, i\, n_i \, ,

1263: \ee

1264: where $n_i$ is an arbitrary integer associated to every root $u_i$. One

1265: self-consistently assumes that the momenta scale as $p_i\sim 1/L$ for $L\to \infty$ 

1266: implying that the Bethe roots scale as $u_i\sim L$.  Therefore in the $L\to\infty$

1267: limit the above equation reduces to

1268: \be

1269: \label{BATL}

1270: \frac{1}{u_i} = 2\pi\, n_j +\frac{2}{L} \sum_{k=1\, (k\neq i)}^M

1271: \frac{1}{u_j-u_k} \, .

1272: \ee

1273: In the thermodynamic limit the roots $u_i$ accumulate on smooth contours in the

1274: complex plane known as ``Bethe strings'' which turn the set of algebraic Bethe equations into

1275: an integral equation. To see this introduce the Bethe root density 

1276: \be

1277: \label{RhoDef}

1278: \rho(u):= \frac{1}{M} \sum_{j=1}^M \delta(u-u_j)

1279: \qquad \mbox{with}\quad \int_C du\, \rho(u) =1 \, ,

1280: \ee

1281: where $C$ is the support of the density. i.e.~the union of all Bethe string contours. 

1282: Multiplying \eqn{BATL} with

1283: $u_i$ and introducing $\rho(u)$ one arrives at the singular integral 

1284: equation \footnote{$\fpint{}{} dv\,

1285:   \frac{(\ldots)}{v-u}$ denotes the principle part prescription.} 

1286:  \be

1287: \label{contBE}

1288: \fpint{C}{} dv \, \frac{\rho(v)\, u}{v-u} = -\frac{1}{2\,\alpha} +

1289: \frac{\pi\, n_{C(u)}\, u}{\alpha} \qquad \mbox{where}\quad  u\in C\,   \quad 

1290: \mbox{and} \quad \alpha:=

1291: \frac{M}{L}\, .

1292: \ee

1293: The mode numbers $n_{C(u)}$ are integers which are assumed to be

1294: constant on each smooth component $C_n$ of the density support $C=\cup \, C_n$ in the complex

1295: plane. These integers and the distribution of components $C_n$ select

1296: the numerous solutions to the continuum Bethe equations \eqn{contBE}. Furthermore the 

1297: continuum energy now becomes

1298: \be

1299: Q_2=M\, \int_C\, \frac{\rho(u)}{u^2} \, .

1300: \ee

1301: 

1302: As was shown in \cite{hep-th/0306139} the gauge dual to the folded string 

1303: solution of section~\ref{subsect:3.2} corresponds to a two cut 

1304: support $C=C_1\cup C_1^\ast$ with $n_{C_1}=-1$ and

1305: $n_{C_1^\ast}=1$ sketched in figure~\ref{roots}.

1306: \begin{figure}[t]

1307: \begin{center}

1308: \includegraphics[width=9cm]{roots.eps}

1309: \end{center}

1310: \caption{Bethe root distribution for the gauge dual of the folded

1311:   string. For large $L$ the roots condense into two cuts in the

1312:   complex plane.}

1313:  \label{roots}

1314: \end{figure}

1315: The key trick to obtain analytical expressions for $\rho(u)$ is to consider the 

1316: analytic continuation to negative filling fraction $\beta:=-\alpha$: Then the two cuts

1317: $C_1\cup C_1^\ast$ are mapped to intervals on the real line

1318: ($C^\ast_1\to [-b,-a]$ and $C_1\to [a,b]$) \cite{hep-th/0306139}.

1319: Then \eqn{contBE} may be brought into the compact form 

1320: \be

1321: \label{FcontBE}

1322: \fpint{a}{b} dv \, \frac{\tilde\rho(v)\, u^2}{v^2-u^2} = \frac{1}{4}-

1323: \frac{\pi\, u}{2}

1324: \qquad \mbox{with}\quad \int_a^b dv\, \tilde\rho(v)=\frac{\beta}{2}\, ,

1325: \ee

1326: using $\rho(-v)=\rho(v)$ and defining $\tilde\rho(v):=\beta\, \rho(v)$.

1327: In order to proceed one introduces the resolvent

1328: \be

1329: H(u):=\int_a^b dv\, \tilde\rho(v)\, \frac{v^2}{v^2-u^2}=-\frac{\alpha}{2}+\sum_{k=1}^\infty

1330: Q_{2k}\, u^{2k}

1331: \ee

1332: which gives rise to the infinite tower of conserved even charges $Q_{2k}$ 

1333: with the energy $E_2=\ft{1}{8\pi^2}\,Q_2$ \cite{hep-th/0310182}. 

1334: Across the cut $u\in[a,b]$ the resolvent $H(u)$ behaves as

1335: \be

1336: \label{Heqn}

1337: H(u\pm i\, \epsilon) =-\frac{\alpha}{2} +\frac{1}{4} -\frac{\pi}{2}\, u

1338: \pm i\,\pi \,\frac{u}{2}\, \tilde\rho(u)\, ,\qquad u\in[a,b]\,,

1339: \ee

1340: which one shows using the distributional identity $\frac{1}{x\pm i\, \epsilon}=

1341: P(\frac{1}{x}) \mp i\,\pi \, \delta(x)$  and eq.~\eqn{FcontBE}.  From this

1342: one obtains an integral expression for the resolvent

1343: \be

1344: \label{Hres}

1345: H(u)=-\frac{\alpha}{2} +\frac{1}{4}-\int_a^b dv\, \frac{v^2}{v^2-u^2}\,

1346: \sqrt{\frac{(b^2-u^2)\, (a^2-u^2)}{(b^2-v^2)\, (v^2-a^2)}}\, ,

1347: \ee

1348: which in turn self-consistently yields the density

1349: \be

1350: \tilde\rho(u)=\frac{2}{\pi\, u}\, \fpint{a}{b}  dv\, \frac{v^2}{v^2-u^2}\,

1351: \sqrt{\frac{(b^2-u^2)\, (u^2-a^2)}{(b^2-v^2)\, (v^2-a^2)}}\, ,

1352: \ee

1353: Finally the interval boundaries $a$ and $b$ are implicitly determined through the normalization and

1354: positivity conditions on $\tilde\rho(u)$ by the relations \cite{hep-th/0308117,hep-th/0310182}

1355: \be

1356: \frac{1}{a}=4\, K(q)\, ,\qquad \frac{1}{b}= 4\, \sqrt{1-q}\, K(q)\, ,

1357: \qquad q:=\frac{b^2-a^2}{b^2}\, .

1358: \ee

1359: The resolvent and the density may be expressed in closed forms using the elliptic

1360: integral of the third kind\footnote{Our convention is $\Pi(m^2,q):=\int_0^{\pi/2}

1361: \frac{d\phi}{(1-m^2\,\sin^2\phi)\sqrt{1-q\,\sin^2\phi}}$\,.}

1362: \begin{align}

1363: H(u) &=-\frac{\alpha}{2}+\frac{1}{4}-\frac{\pi}{2}\, u -\frac{1}{4b}\,

1364: \sqrt{\frac{a^2-u^2}{b^2-u^2}}\, \Bigl [ \frac{b^2}{a} - 4\, u^2\, \Pi

1365: \Bigl(\frac{b^2-u^2}{b^2},

1366: q\Bigr)\, \Bigr ]\, ,\nn\\

1367: \tilde\rho(u) &=\frac{1}{2\,\pi\, b\, u}\,

1368: \sqrt{\frac{u^2-a^2}{b^2-u^2}}\, \Bigl [ \frac{b^2}{a} - 4\, u^2\, \Pi

1369: \Bigl(\frac{b^2-u^2}{b^2}, q\Bigr)\, \Bigr ]\, .

1370: \end{align}

1371: From this it is straightforward to (finally) extract the energy eigenvalue $Q_2$ of the

1372: two cut solution in the parametric form

1373: \be

1374: \label{GTF1}

1375: E_2= \frac{1}{2\pi^2}\, K(q)\, \Bigl [2\, E(q)-(2-q)\, K(q)\, \Bigr ]\,

1376: \ee

1377: with

1378: \be

1379: \label{GTF2}

1380: \alpha=\frac{J_2}{J}= \frac{1}{2}-\frac{1}{2\,\sqrt{1-q}}\, \frac{E(q)}{K(q)}\, .

1381: \ee

1382: This final result for the one loop gauge theory anomalous scaling dimension can

1383: now be compared to the folded string energies of section~\ref{subsect:3.2}

1384:  eqs.~\eqn{FE1} and \eqn{FEprediction}.  They do not manifestly agree, however, if

1385:  one relates the auxiliary parameters $q_0$ and $q$ through \cite{hep-th/0308117}

1386:  \be

1387:  q_0=-\frac{(1-\sqrt{1-q})^2}{4\, \sqrt{1-q}}

1388:  \ee

1389:  one may show that

1390:  \be

1391:  K(q_0)=(1-q)^{1/4}\, K(q) \, , \qquad E(q_0)=\ft 12 (1-q)^{-1/4}\, E(q) + \ft 12 (1-q)^{1/4}

1392:  \, K(q)

1393:  \ee

1394:  using elliptic integral modular transformations. Using these relations the gauge theory

1395:  result \eqn{GTF1} and \eqn{GTF2} may be transformed into the string result of

1396:  \eqn{FE1} and \eqn{FEprediction}.  Hence the one-loop gauge theory scaling dimensions 

1397:  indeed agree with the string prediction!

1398: 

1399: The analysis of the circular string configuration goes along the same lines. Here the Bethe roots 

1400: turn out to condense on the imaginary axis. The root density is then symmetric along 

1401: the imaginary axis, $\rho(u)=\rho(- u)$

1402: and remains constant along a segment $[-c,c]$. For $u>c$ and $u<-c$ it falls off

1403: towards zero. We shall not go through the detailed construction of  the density for this

1404: configuration but refer the reader to the original papers of \cite{hep-th/0306139,hep-th/0308117}

1405: being best explained in \cite{hep-th/0310182}. The outcome of this analysis is again a

1406: perfect matching of the energy eigenvalue of the spin chain with 

1407: the circular string energy of eqs.~\eqn{CS1} and \eqn{CS2}.

1408: 

1409: As a matter of fact one can go beyond this and match all the higher charges of gauge and string

1410: theory, as was shown for the first time in \cite{hep-th/0310182} by using an approach based on the

1411: B\"acklund transform. 

1412: 

1413: \subsection{Higher Loops in the SU(2) sector and discrepancies}

1414: 

1415: The discussed connection to an integrable spin chain at one-loop raises the question whether

1416: integrability is merely an artifact of the one-loop approximation or a genuine property of planar

1417: \Nfour gauge theory. Remarkably all present gauge theory data  points towards the 

1418: latter being the case.

1419: 

1420: Higher loop contributions to the planar dilatation operator in the $SU(2)$ subsector

1421: are by now firmly established to

1422: the two-loop \cite{hep-th/0303060} and three-loop level \cite{hep-th/0310252,hep-th/0409009}. 

1423: In  $s=1/2$ quantum spin chain language they take the explicit forms

1424: \begin{align*}

1425:  {\widehat{{\cal D}}_{\rm 2-loop}}  &= \sum_{i=1}^L

1426: -\vec\sigma_l\cdot\vec\sigma_{l+2} + 4\,

1427: \vec\sigma_l\cdot\vec\sigma_{l+1} - 3\cdot \eins \\

1428:  {\widehat{{\cal D}}_{\rm 3-loop}}  &= \sum_{i=1}^L

1429: -\vec\sigma_l\cdot\vec\sigma_{l+3} +

1430: (\vec\sigma_l\cdot\vec\sigma_{l+2})\, (\vec\sigma_{l+1}\cdot\vec\sigma_{l+3})-

1431: (\vec\sigma_l\cdot\vec\sigma_{l+3})\,

1432: (\vec\sigma_{l+1}\cdot\vec\sigma_{l+2})\\& \qquad

1433: +10 \,\vec\sigma_l\cdot\vec\sigma_{l+2} -29\,

1434: \vec\sigma_l\cdot\vec\sigma_{l+1} +20\cdot \eins \, .

1435: \end{align*}

1436: In general the $k$-loop contribution to the dilatation operator involves interactions of $k+1$

1437: neighboring spins, i.e.~the full dilatation operator ${\cal D}=\sum_{k=1}^\infty

1438: {\cal D}_{\rm k-loop}$ will correspond to a long-range interacting spin-chain

1439: hamiltonian. Note also the appearance of novel quartic spin interactions $(\vec\sigma_i\cdot\vec\sigma_j)

1440: (\vec\sigma_k\cdot\vec\sigma_l)$ at the three-loop level. Generically even higher interactions

1441: of the form $(\vec\sigma_\bullet\cdot\vec\sigma_\bullet)^k$ are expected at the $[\frac k 2]+1$ loop

1442: levels.

1443:  

1444: Integrability remains stable up to the three-loop order and acts in a 

1445: perturbative sense:

1446: The conserved charges of the Heisenberg XXX${}_{1/2}$ chain receive higher order corrections  in 

1447: $\lambda$ of the form

1448: $

1449: {\cal Q}_k= Q_k^{(1)}+\lambda\, Q_k^{(2)} +\lambda^2\, Q_k^{(3)}+\ldots

1450: $ as one would expect.

1451: The full charges ${\cal O}_k$ commute with each other ($[{\cal O}_k,{\cal O}_l]=0$) 

1452: which translates into commutation relations for the various loop contributions

1453: ${\cal O}^{(r)}_k$ upon expansion in $\lambda$,

1454: i.e.

1455: \begin{align}

1456: [Q^{(1)}_k,Q^{(1)}_l]=0\, ,\qquad [Q^{(1)}_k,Q^{(2)}_l]+[Q^{(2)}_k,Q^{(1)}_l]&=0\, \nonumber\\

1457: [Q^{(1)}_k,Q^{(3)}_l]+[{Q}^{(2)}_k,{Q}^{(2)}_l]+[Q^{(3)}_k,Q^{(1)}_l]&=0

1458: \end{align}

1459: and so on. However, opposed to the situation for the Heisenberg chain

1460: \cite{hep-th/9605187}, 

1461: there does not yet exist

1462: an algebraic construction of the gauge theory charges at higher loops. Nevertheless the

1463: first few $Q_k$ have been constructed manually to higher loop-orders \cite{hep-th/0409054}. 

1464: 

1465: An additional key property of these higher-loop corrections is that they obey BMN

1466: scaling: The emergence of the effective loop-counting parameter $\lambda':=\lambda/J^2$

1467: in the $J\to\infty$ limit  leading to the scaling dimensions $\Delta\sim\sqrt{1+\lambda'\, n^2}$

1468: for two magnon states in quantitative agreement with plane-wave superstrings.

1469: 

1470: Motivated by these findings Beisert, Dippel and Staudacher \cite{hep-th/0405001} 

1471: turned the logic around and simply {\sl assumed} integrability, BMN scaling and a Feynman

1472: diagrammatic origin of the $k$-loop $SU(2)$ dilatation operator. Interestingly these assumptions

1473: constrain the possible structures of the planar dilatation operator completely up to

1474: the five-loop level (and possibly beyond).

1475: 

1476: How can one now diagonalize the higher-loop corrected dilatation operator? For this the

1477: ansatz for the Bethe wave-functions \eqn{Bethesansatz} needs to be adjusted in a perturbative 

1478: sense in order to accommodate the long-range interactions, leading to a 

1479: ``perturbative asymptotic Bethe ansatz'' for the two magnon wave function \cite{hep-th/0412188} 

1480: \begin{align}

1481: \psi(x_1,x_2)&= e^{i({p_1}\, x_1 + {p_2}\, x_2)}\, {f(x_2-x_1,p_1,p_2)}\nn\\

1482: & \quad + {S(p_2,p_1)}\,

1483: e^{i({p_2}\, x_1 + {p_1}\, x_2)}\, {f(L-x_2+x_1,p_1,p_2)} \, .

1484: \label{PABA}

1485: \end{align}

1486: Here one needs to introduce a perturbative deformation of the S-matrix

1487: \be

1488: \label{PABAS}

1489: S(p_1,p_2)=S_0(p_1,p_2)+\sum_{n=1}^\infty {\lambda^n}\, S_n(p_1,p_2)

1490: \ee

1491: which is determined by the eigenvalue problem. Moreover suitable  ``fudge functions'' 

1492: enter the ansatz

1493: \be

1494: {f(x,p_1,p_2)}=1+ \sum_{n=0}^\infty {\lambda^{n+|x|}}\, f_n(x,p_1,p_2)

1495: \stackrel{x\gg 1}{\longrightarrow} 1 \,

1496: \ee

1497: which account for a deformation of the plane-wave form of the eigenfunction when two magnons

1498: approach each other within the interaction range of the spin-chain hamiltonian\footnote{We are here actually

1499: using a slightly modified definition of these functions to the one presented in \cite{hep-th/0412188}

1500: which was considered in \cite{hep-th/0412331}.}. By construction they are invisible in the 

1501: asymptotic regime $x\gg 1$ (or rather $x$ larger than the highest loop order considered) of well separated

1502: magnons. The detailed form of these functions is completely irrelevant for the physical spectrum as

1503: a consequence of the factorized scattering property of the integrable system.

1504: 

1505: With this perturbative asymptotic Bethe ansatz \eqn{PABA} one shows that the form of the

1506: Bethe equations remains unchanged, i.e.~the perturbative S-matrix \eqn{PABAS} simply

1507: appears on the right hand side of the equations

1508: \be

1509: e^{ip_k L}=\prod_{i=1,i\neq k}^M S(p_k,p_i) \, ,

1510: \ee

1511: and is determined by demanding $\psi(x_1,x_2)$ to be an eigenfunction of the dilatation operator

1512: just as we did in section~\ref{sect:coordBethe}.

1513: Based on the constructed five-loop form of the dilatation operator the S-matrix is then determined up

1514: to ${\cal O}(\lambda^4)$. The obtained series in $\lambda$ turns out to be of a remarkably simple

1515: structure, which enabled the authors of \cite{hep-th/0405001} to conjecture 

1516: an asymptotic {\sl all} loop  expression for the perturbative S-matrix 

1517:  \be

1518:  \label{BDS1}

1519: S(p_1,p_2)=\frac{\varphi(p_1)-\varphi(p_2)+i}{\varphi(p_1)-\varphi(p_2)-i}

1520: \qquad \mbox{with} \quad

1521: \varphi(p)= \ft 1 2\, \cot (\ft p 2)\, \sqrt{1+{\lambda}\, \sin^2(\ft p 2)} \, ,

1522: \ee

1523: to be compared to the one-loop form of \eqn{S-matrix}.  The conjectured asymptotic all loop form of the

1524: energy density generalizing the one-loop expression \eqn{E21loop} reads\footnote{The full

1525: conjecture for all the higher charge densities $q_k$ may be found in \cite{hep-th/0405001}.}

1526: \be

1527: \label{BDS2}

1528: \lambda\, q_2(p)=  \sqrt{1+8\,\lambda\,\sin^2(\ft p2)}-1 

1529: \ee

1530: with the total energy being given by $E_2=\sum_{i=1}^M q_2(p_i)$. Note that both expressions

1531: manifestly obey BMN scaling as the quasi-momenta scale like $p\sim L^{-1}$ in the thermodynamic limit 

1532: as we discussed in section~\ref{subsect:thermo}. It is important to stress that these Bethe equations

1533: only make sense asymptotically: For a chain (or gauge theory operator) of

1534: length $L$ the eqs.~\eqn{BDS1} and \eqn{BDS2} yield a prediction for the energy (or

1535: scaling dimension) up to $L-1$ loops. This is so, as the interaction range of the Hamiltonian will

1536: reach the length of the spin chain beyond this point, and the multi-magnon wavefunctions of \eqn{PABA} 

1537: can never enter the asymptotic regime. What happens beyond the $L$ loop level is still a mystery.

1538: At this point the ``wrapping" interactions 

1539: start to set in: The interaction range of the spin chain

1540: Hamiltonian cannot spread any further and starts to ``wrap" around the chain. In the dimensional

1541: reduced model of plane-wave matrix theory 

1542: \cite{hep-th/0202021,hep-th/0205185,hep-th/0207034,hep-th/0207050,hep-th/0306054,hep-th/0310232}

1543: these effects have been studied

1544: explicitly at the four-loop level in \cite{hep-th/0412331} where the wrapping effects set

1545: in for the first time in the $SU(2)$ subsector. No ``natural" way of transforming the generic 

1546: dilatation operator to the wrapping situation was found. Finally let us restate that it has not been shown 

1547: so far that a microscopic long-range spin-chain Hamiltonian truly exists, which has a 

1548: spectrum determined by the conjectured perturbative asymptotic Bethe equations \eqn{BDS1} 

1549: and \eqn{BDS2} of Beisert, Dippel and Staudacher.

1550: 

1551: \medskip

1552: 

1553: In any case, the proposed all-loop asymptotic Bethe equations \eqn{BDS1} and \eqn{BDS2} may now

1554: be studied in the thermodynamic limit just as we did above for the one-loop case. This

1555: was done in \cite{hep-th/0401057} and \cite{hep-th/0405001} and allows

1556: for a comparison to the results obtained in section~\ref{section:spinning_string_solutions}

1557: for the energies of the spinning folded and closed string solutions. Recall that these yield

1558: predictions to all-loops in $\lambda'$. While the two loop gauge theory result is in perfect

1559: agreement, the three-loop scaling dimensions {\sl fail} to match with the expected dual 

1560: string theory result! This three-loop disagreement also arises in the comparison to the 

1561: near plane-wave string spectrum computed in 

1562: \cite{hep-th/0307032,hep-th/0404007,hep-th/0405153,hep-th/0407096}, i.e.~the first

1563: $1/J$ corrections to the Penrose limit of \AdSS to the plane-wave background.

1564: 

1565: Does this mean that the AdS/CFT correspondence does not hold in its

1566: strong sense? While this logical possibility certainly

1567: exists,  an alternative explanation is that one is dealing with an

1568: order of 

1569: limits problem as pointed out 

1570: initially in \cite{hep-th/0405001}. While in string theory one works

1571: in 

1572: a limit of $\lambda\to\infty$

1573: with $J^2/\lambda$ held fixed, in gauge theory one stays in the perturbative regime $\lambda\ll 1$

1574: and thereafter takes the $J\to\infty$ limit, keeping only terms which scale as $\lambda/J^2$.

1575: These two limits need not commute. Most likely the above-mentioned ``wrapping" interactions

1576: must be included into the gauge theory constructions in order to match the string theory energies.

1577: On the other hand, the firm finite $L$ results at order $\lambda^3$ of

1578: the gauge theory are still free of ``wrapping"

1579: interactions: These only start to set in at the four-loop level (in the considered minimal $SU(2)$

1580: subsector). Moreover to what extent the integrability

1581: is preserved in the presence of these ``wrapping" interactions is unclear at the moment.

1582: Certainly the resolution of this discrepancy remains a pressing open problem in the field.

1583: 

1584: \subsection{Further developments}

1585: 

1586: The gauge theory analysis of the planar dilatation operator and its relation to

1587: integrable spin chain models has been extended in two directions: Larger sectors within \Nfour super

1588: Yang-Mills and conformal deformations of the original theory. 

1589: 

1590: In their seminal paper uncovering the integrable spin chain structure Minahan and

1591: Zarembo \cite{hep-th/0212208} actually considered the full scalar sector of 

1592: the gauge theory at one-loop order. This gives rise to an integrable $SO(6)$ magnetic quantum spin chain

1593: of which the discussed $SU(2)$ Heisenberg model arises in a subsector.

1594: This work was generalized in \cite{hep-th/0307015} 

1595: to all local operators of the planar one-loop \Nfour theory, leading to an integrable super-spin

1596: chain with $SU(2,2|4)$ symmetry discussed in \cite{hep-th/0307042}. 

1597: The excitations of this super-spin chain consist of scalars, field strengths, fermions and an

1598: arbitrary number of covariant derivatives of these three leading to an infinite number of spin

1599: degrees of freedom on a single lattice site. 

1600: The thermodynamic limit of this 

1601: super spin chain was later on constructed in larger supersymmetric subsectors in \cite{hep-th/0412254} and 

1602: in \cite{hep-th/0503200} for the full system leading 

1603: to spectral curves, which reproduce the results of the classical string at one-loop order.

1604: 

1605: The conjectured form of the asymptotic higher-loop Bethe ansatz for the $PSU(2)$ subsector was

1606: generalized to the full theory recently  in \cite{hep-th/0504190} in

1607: form of a long-range $SU(2,2|4)$

1608: Bethe ansatz. In this paper the corresponding generalization for the

1609: quantum string Bethe equations,

1610: generalizing \cite{hep-th/0406256} relevant for the $SU(2)$ sector, was also provided. 

1611: A novel feature of leaving the minimal $SU(2)$ sector at higher loop orders

1612: is that the spin chain begins to fluctuate in length \cite{hep-th/0310252}: The hamiltonian 

1613: (or dilatation operator) preserves

1614: the classical scaling dimensions but not the length of the chain,

1615: e.g.~two fermions have the same classical

1616: scaling dimension as three scalars and can mix if they carry identical

1617: charges.

1618: 

1619: An alternative route for comparing string energies to gauge theory scaling dimensions lies in the

1620: coherent-state effective action approach pioneered by Kruczenski \cite{hep-th/0311203}.

1621: Here one establishes an effective action for the string whose center of mass moves along a big

1622: circle of the $S^3$ with large angular momentum in the ``weak

1623: coupling" limit $\lambda/J^2\ll 1$.

1624: This action is then shown to agree with the long-wave length

1625: approximation of the gauge theory spin chain at one-loop. In this

1626: approach there is no need to compare explicit solutions any longer,

1627: however, considering higher-loop effects and fermions becomes more challenging in this language.

1628: For details see

1629: \cite{hep-th/0403120,hep-th/0403139,hep-th/0404133,hep-th/0406189,hep-th/0410022,hep-th/0503185,hep-th/0409086,hep-th/0503159}

1630: and also Tseytlin's review \cite{hep-th/0409296}.

1631: 

1632: In view of the reviewed insights an obvious next question to address is what can 

1633: be said about

1634: the non-planar sector of the gauge theory dual to string interactions. The gauge theory dilatation

1635: operator is indeed known for the first two loop orders in the $SU(2)$ sector exactly, that is

1636: including all non-planar contributions. However, extracting

1637: physical data from it, such as amplitudes for the decay of single trace operators to double 

1638: trace ones is nontrivial. This has been performed with great success for the case of two or three

1639: magnon excitations in the BMN limit being dual to the interacting plane-wave superstring  

1640: (reviewed in \cite{hep-th/0307027}). Performing the same computation for a macroscopic

1641: number of magnons, thus describing the quantum decay of the discussed spinning strings,

1642: is complicated enormously by the complexity of the Bethe

1643: wavefunction \eqn{bethewf} for large $M$. The analysis on the string side, however, can be

1644: performed by considering a semiclassical decay process \cite{hep-th/0410275}. 

1645: For a related discussion on the non-planar gauge theory aspects see \cite{hep-th/0404066}.

1646: 

1647: An interesting toy model for \Nfour Super Yang-Mills is its dimensional reduction on

1648: a three sphere to a quantum mechanical system  \cite{hep-th/0306054}, which turns out to be the 

1649: plane-wave matrix theory of

1650: \cite{hep-th/0202021,hep-th/0205185,hep-th/0207034,hep-th/0207050} related to M-theory on

1651: the plane-wave background. 

1652: The Hamiltonian of this matrix quantum mechanics

1653: reduces to an integrable spin chain in the large $N$ limit, which remarkably is identical to 

1654: the full \Nfour system up to the three loop level in the overlapping $SU(2|3)$ sector 

1655: \cite{hep-th/0310232} (via a perturbative 

1656: redefinition of the coupling constant). However, a recent four loop study displays a breakdown

1657: of BMN scaling at this level of perturbation theory while integrability is stable 

1658: \cite{hep-th/0412331}. 

1659: What this finding implies for the full \Nfour model remains to be seen.

1660: 

1661: A recent line of research concerns the study of deformations of ${\cal

1662:   N}=4$ Super Yang-Mills

1663: which maintain the quantum conformal structure known as the Leigh-Strassler or $\beta$

1664: deformations \cite{hep-th/9503121}. The one-loop dilatation operator was constructed in

1665: subsectors of the theory in \cite{hep-th/0312218,hep-th/0405215}. Moreover the explicit

1666: construction of the supergravity background dual to the $\beta$ deformed gauge theory

1667: was achieved by Lunin and Maldacena \cite{hep-th/0502086}. Again the classical bosonic

1668: string theory in this background is integrable and exhibits a Lax pair \cite{hep-th/0503192}

1669: yielding a string Bethe equation which agrees with the thermodynamic limit of the one-loop

1670: Bethe equation for the gauge theory dilatation operator \cite{hep-th/0503201}. So the complete

1671: discussion of this review lifts to the $\beta$ deformed case. As a matter of fact even larger

1672: (three-parameter families) of generically non-supersymmetric deformations of $\Nfour$ 

1673: Super Yang-Mills have been considered with known supergravity duals \cite{hep-th/0503201}.

1674: The corresponding ``twisted" gauge theory spin chain and Bethe ansatz was constructed in

1675: \cite{hep-th/0505187}. 

1676: 

1677: Open integrable spin chains have also appeared in the AdS/CFT setting where the

1678: boundaries of the spin chain correspond to fields in

1679: the fundamental representation, see 

1680: \cite{hep-th/0312091,hep-th/0401016,hep-th/0403004,hep-th/0401041} for

1681: such constructions. In \cite{hep-th/0501078} open

1682: integrable spin chains emerged within subdeterminant operators in \Nfour

1683: super Yang-Mills dual to so-called ``giant gravitons''.

1684: 

1685: First investigations on the role of integrability for the three-point functions in the gauge

1686: theory were performed in \cite{hep-th/0404190,hep-th/0407140,hep-th/0502186}. 

1687: 

1688: Finally let us mention that integrable structures are known to also

1689: appear in QCD, such as in high-energy scattering processes and other

1690: instances \cite{hep-th/9311037,hep-th/9404173,hep-ph/9805225,hep-ph/9907420},

1691: see \cite{hep-th/0407232} for a recent review.

1692: 

1693: In conclusion the emergence of integrable spin chains in the AdS/CFT correspondence has led

1694: to great insights into dynamical aspects of the duality and might hold the key to a complete

1695: determination of the spectrum of both theories. Recent developments point towards 

1696: integrability being a generic 

1697: property of conformal gauge theories in the planar limit not necessarily connected to supersymmetry. 

1698: Finally a great challenge for the future is to 

1699: understand the integrable spin chain nature of the {\sl quantum} string in \AdSS and related backgrounds.

1700: 

1701: 

1702: \newpage

1703: 

1704: %==============================================================================

1705: 

1706: \section{Acknowledgements}

1707: 

1708: I wish to thank Gleb Arutyunov, Niklas Beisert, Sergey Frolov,

1709: Matthias Staudacher, Arkady Tseytlin and Marija Zamaklar for helpful discussions and

1710: important comments on the manuscript. This review grew out of a lecture

1711: delivered at the Post-Strings 2004 Meeting at Durham. I 

1712: thank the organizers of the meeting for hospitality and an inspiring workshop.

1713: 

1714: \newpage

1715: 

1716: %==============================================================================

1717: 

1718: % To use the bibtex bibliography in 'LivRevRel.bib' do:

1719: % 'latex LivRevRel'

1720: % 'bibtex LivRevRel'

1721: % 'latex LivRevRel'

1722: % 'latex LivRevRel'

1723: 

1724: \bibliography{lrl}

1725: 

1726: \end{document}

1727: 

1728: