1: \documentclass{ws-p8-50x6-00}

2: \def\rightnote{Statistics of Vacua\hfill

3: String Phenomenology 2003, Durham}

4: 

5: \begin{document}

6: 

7: \title{Statistics of String vacua}

8: 

9: \author{Michael R. Douglas}

10: \address{Department of Physics and Astronomy, Rutgers University\\

11: Piscataway, NJ 08855-0849 USA\\

12: {\it and}\\

13: I.H.E.S., Le Bois-Marie, Bures-sur-Yvette, 91440 France}

14: 

15: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

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

17: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

18: 

19: \maketitle

20: 

21: \newcommand{\adref}{0307049}

22: \abstracts{

23: We give an introduction to the statistical approach to studying

24: vacua of string/M theory,

25: and discuss recent results of 

26: Ashok and Douglas on

27: counting supersymmetric flux vacua in type IIb

28: Calabi-Yau compactification.\\

29: \noindent

30: To appear in the proceedings of the 2003 String Phenomenology

31: workshop in Durham, UK.}

32: 

33: % statistics talk

34: \newcommand{\page}{}

35: \newcommand{\p}{}

36: \newcommand{\NN}{{\mathcal \bf N}}

37: %

38: \hyphenation{super-symmetric super-symmetry}

39: \hyphenation{non-super-symm-etric di-men-sion-al}

40: \def\blue{}

41: \def\red{}

42: %

43: \newcommand{\sfont}[1]{{\tiny #1}}

44: \newcommand{\sref}[1]{{\blue (#1)}}

45: \newcommand{\heading}[1]{\begin{center}\large\bf #1\end{center}}

46: %\newcommand{\comment}

47: %

48: \newfam\black

49: \font\blackboard=msbm10 %scaled \magstep1

50: \font\blackboards=msbm7

51: \font\blackboardss=msbm5

52: \textfont\black=\blackboard

53: \scriptfont\black=\blackboards

54: \scriptscriptfont\black=\blackboardss

55: \def\Bbb#1{{\fam\black\relax#1}}

56: %

57: \def\sgn{{\rm sgn\ }}

58: \def\etal{{\it et.al.}}

59: \def\slashslash{{/\hskip-0.2em/}}

60: \def\CA{{\cal A}}

61: \def\CC{{\cal C}}

62: \def\CD{{\cal D}}

63: \def\CF{{\cal F}}

64: \def\CH{{\cal H}}

65: \def\CI{{\cal I}}

66: \def\CJ{{\cal J}}

67: \def\CL{{\cal L}}

68: \def\CM{{\cal M}}

69: \def\CN{{\cal N}}

70: \def\CO{{\cal O}}

71: \def\CP{{\cal P}}

72: \def\CT{{\cal T}}

73: \def\CW{{\cal W}}

74: \def\BC{\Bbb{C}}

75: \def\BH{\Bbb{H}}

76: \def\BM{\Bbb{M}}

77: \def\BP{\Bbb{P}}

78: %\def\P{\Bbb{P}}

79: \def\BR{\Bbb{R}}

80: \def\BX{\Bbb{X}}

81: \def\BZ{\Bbb{Z}}

82: \def\mapr{\mathop{\longrightarrow}\limits}

83: \def\half{{1\over 2}}

84: \def\GeV{~{\rm GeV}}

85: \def\TeV{~{\rm TeV}}

86: \def\Coh{{\rm Coh}~}

87: \def\Cohc{{\rm Coh}_c~}

88: \def\Mod{{\rm Mod}~}

89: \def\ind{{\rm ind}~}

90: \def\Tr{{\rm Tr}~}

91: \def\tr{{\rm tr}~}

92: \def\grad{{\rm grad}~}

93: \def\CY#1{CY$_#1$}

94: \def\rk{{\rm rk}~}

95: \def\Im{{\rm Im}~}

96: \def\Hom{{\rm Hom}}

97: \def\Ext{{\rm Ext}}

98: \def\Vol{{\rm Vol~}}

99: \def\Stab{{\rm Stab}}

100: \def\End{{\rm End~}}

101: \def\ib{{\bar i}}

102: \def\jb{{\bar j}}

103: \def\zb{{\bar z}}

104: \def\pp{\partial}

105: \def\pb{{\bar\partial}}

106: \def\I{{I}}

107: \def\II{{II}}

108: \def\IIa{{IIa}}

109: \def\IIb{{IIb}}

110: \def\vev#1{{\langle#1\rangle}}

111: \def\vvev#1{{\langle\langle#1\rangle\rangle}}

112: \def\ket#1{{|#1\rangle}}

113: \def\Dslash{\rlap{\hskip0.2em/}D}

114: \def\dual{{v}} % or check mark, or whatever

115: \def\intersect{\cdot}

116: \def\Nbar{{\bar N}}

117: \def\NRR{{N_{RR}}}

118: \def\NNS{{N_{NS}}}

119: \def\bigvev#1{\bigg\langle{#1}\bigg\rangle}

120: 

121: \section{Introduction}

122: 

123: What is the most important problem of string phenomenology?

124: 

125: Most would probably say that it is to find a compactification which

126: reproduces the Standard Model in all its details, and makes new

127: predictions.  More realistically, one could settle for qualitative

128: agreement with the Standard Model, as long as one gets new

129: predictions.  

130: 

131: If one interprets ``qualitative'' loosely enough, this has been done,

132: but the standard which has been met is not very high.  While the

133: general structure of supersymmetric grand unification was already realized

134: in the early works, many of the important features, especially

135: the hierarchy of scales and the small cosmological constant, 

136: not to mention specific parameter values,

137: were controlled by non-perturbative effects, which were not at all understood

138: at that time.  Besides these classic problems,

139: string and M theory constructions come with new problems

140: of their own, which require solution or explanation: why our universe

141: appears four-dimensional, what chooses a particular compactification

142: or other extra dimensional structure, and so forth.

143: 

144: Building on the understanding of non-perturbative string/M theory 

145: achieved in the mid-late 1990's, it is now possible to study

146: compactifications which address these goals.  Still, no

147: model has been proposed which meets all of the requirements.

148: Should we be worried?\p

149: 

150: One can argue that, to the contrary, this is {\bf good} news for any

151: hope for making predictions from string theory.  The argument is

152: simply that we have only explored a tiny fraction of the

153: compactifications now believed to exist, and these seem to be a fairly

154: random cross section chosen not because they are more likely to work,

155: but more because of ease of analysis and historical accident.  \p

156: 

157: If we had already found satisfactory candidates in this small sample,

158: it is likely that the full set would contain large numbers of theories

159: which agree with the Standard Model {\bf in every detail}.  Even worse,

160: different theories out of this set might 

161: lead to very different predictions.  In a worst-case scenario, 

162: string theory might not be testable at all.

163: 

164: To bring home this point, let us look at the variety of string/M

165: theory constructions, and ask how many might be candidates. 

166: Focus on Calabi-Yau compactification, where some numbers

167: are known.

168: 

169: The number of distinct CY threefolds is believed to be $10^5$--$10^6$.

170: All these constructions involve additional choices -- a choice of

171: bundle or brane, with comparable multiplicity, which directly affect

172: the spectrum.  Choices of flux or nonperturbative gauge theory vacuum

173: probably bring in much larger multiplicities -- numbers 

174: $\sim 10^{100}$ are often cited (more  below). 

175: 

176: In a weakly coupled construction, it is not hard to exclude models with

177: exotic matter of various types.  But most constructions contain sectors

178: with $O(1)$ couplings.  Many proposals exist for strongly coupled

179: sectors (composite models, supersymmetric technicolor, the supersymmetry

180: breaking sector, or just hidden sectors) and our current understanding 

181: of nonperturbative gauge theory tends to support these claims.

182: Thus, one cannot throw out such models from the start.  On the positive

183: side, one is better able to check if a specific model works.

184: 

185: Even leaving out considerations of moduli fixing and couplings,

186: numbers like $10^{10}$ qualitatively distinct models seem very

187: plausible.  So far, the number which have been considered in any

188: detail is more like $100$.

189: 

190: If one agrees with this argument, then one is led to the belief that

191: constructing a model which agrees with the Standard Model in

192: detail is {\bf not} the primary goal we should be pursuing

193: at this point:

194: \begin{itemize}

195: \item If it is possible within our present limitations,

196: this is a negative result.

197: \item If it is not possible within our present limitations,

198: one cannot get a result.

199: \end{itemize}

200: While this conundrum oversimplifies the situation, it deserves

201: consideration.  \p

202: 

203: Of course there are ways around it.  Before focusing on one, 

204: let us at least mention the two others we know of.\p

205: 

206: One is to look for predictions which cannot be matched by four

207: dimensional effective field theory, for example short range

208: modifications to gravity.  This is important, and many speakers here

209: are discussing it.  On the other hand, we have no clear reason to

210: think such effects must exist. Indeed, the ``traditional''

211: Kaluza-Klein scenario still seems to fit evidence such as unification

212: of coupling constants better than any of the more recently proposed

213: variations, and must be taken seriously.

214: 

215: The other is to look for some {\it a priori} principle which selects

216: among the possible vacuum configurations.  Given such a principle, it

217: is of course much more interesting to know if the vacuum it selects

218: can reproduce the observations. 

219: 

220: As an example to illustrate ``Vacuum Selection Principles,'' let me

221: give you my best idea along these lines.

222: \p

223: One might imagine  (see for example \cite{dns})

224: that our vacuum is in some sense ``the most symmetric''

225: among the various possibilities.

226: Besides esthetics, many other candidate

227: principles -- for example, maximizing some natural wave function of the

228: universe -- seem likely

229: to prefer such vacua.

230: \p

231: 

232: On the face of it, this proposal is absurd.  Higher dimensional models

233: with more supersymmetry are obviously more symmetric.  

234: By rights, the ``most symmetric'' four dimensional theory, is the one with

235: the most gauge symmetry.  F theory examples are known with

236: rank $10^5$ gauge groups.

237: \cite{Canska}

238: The Standard Model is not even in the running.

239: \p

240: 

241: However, one might imagine that there exists some unstable

242: nonsupersymmetric vacuum with a huge gauge group, which is preferred

243: by Planck scale cosmology.  It would then roll down to our physical

244: vacuum.  Then, the ``right vacuum'' would be near this preferred

245: symmetric point.

246: 

247: This is as close to a plausible {\it a priori} principle as I have

248: come, but using it still requires some fairly detailed knowledge about

249: the set of possible vacua, and the configuration space which contains

250: the vacua.  Most such ideas require even more information.

251: It does not seem reasonable to hope for a principle which will tell

252: one in advance which string theory, Calabi-Yau, brane configuration

253: etc. to look at.  \p

254: 

255: Of course, it could be that the tests we already know of (or will know

256: in ten years time) are already stringent enough to select out zero or one

257: vacua.  Maybe if we could work with them, an {\it a priori} principle

258: would not be necessary. \p

259: 

260: And, we should keep in mind, that there is no guarantee that any

261: {\it a priori} vacuum selection principle exists.  We only have one

262: sample, and the question of why we observe this one need not have any

263: better answer than ``because we are here.'' \p

264: 

265: This is not

266: an anthropic argument, which is a much more specific and 

267: predictive type of argument.  It is just a conservative interpretation

268: of the goal of science: to explain what we see, not why

269: we see it.

270: 

271: \section{Statistics of vacua}

272: 

273: So what to do?  There is a third approach, which has been advocated for

274: some time by Dine \cite{dine}; it is to look for ``generic'' predictions of

275: string theory.  Can we make this idea precise?

276: 

277: Any attempt to make generic predictions has to be founded on the idea

278: that we know ``all'' string vacua or at least some representative set,

279: in the sense that the distribution of a property of interest is the

280: same in the representative set as in the whole.  At present we can

281: only hypothesize that the sets of vacua we can study concretely could

282: be representative, and eventually check this by the results.  For

283: example, one can argue that the set of all type \II\ compactifications

284: on Calabi-Yau, to the extent that it can be studied at arbitrary

285: volume and string coupling, could be a large fraction of the vacua,

286: because the other known large classes of vacua are believed to be 

287: dual to these.

288: 

289: This at least gets us started, but it is clear that any representative

290: class of vacua is far too complicated to study in any detail at

291: present.  We need to ask simpler questions, which might give us some

292: picture of this huge ``landscape'' of theories.

293: 

294: To do this, we proposed in \cite{mrdstat} to work as follows.  The

295: idea is to make a precise hypothesis for an approximate description of

296: this set: the vacua are vacua of a specific {\blue ensemble of

297: effective field theories}, something like a list or set of theories

298: which we believe can come out of string/M theory.  While the ensemble

299: should be precisely specified, we need not claim that it exactly

300: represents the set of string/M theory vacua, only that it represents

301: it well enough for our purposes.\p

302: 

303: We then can proceed in two directions:

304: \begin{itemize}

305: \item We can test whether our hypothesized ensemble is accurate, by

306: comparing with actual string/M theory constructions.

307: \item We can find out what fraction of vacua out of our ensemble

308: meet a specified phenomenological test.

309: \end{itemize}

310: 

311: \page

312: 

313: Let us give a very simple example to illustrate the point, by asking

314: the question:

315: Out of {\bf all} the four dimensional vacua obtained by string/M

316: theory compactification, {\bf how many} of them are effective field

317: theories with $SU(3)\times SU(2)\times U(1)$ gauge symmetry unbroken

318: at low energy?  \p

319: If we define our terms, and if string/M theory has a precise

320: definition, {\bf and} if there are finitely many physically distinct

321: vacua, then this question has a definite answer.  \p

322: 

323: One can just as easily generalize the question to, out of all vacua,

324: how many have low energy gauge group $G$ ?  Let us denote this number

325: by the function

326: $$

327: d\mu[G] .

328: $$

329: While finding this function exactly is hard, perhaps

330: it can be approximated in some simple and useful way.

331: \p

332: 

333: For example, could it be that the rank $r={\rm rk}\ G$

334: of the unbroken gauge group, roughly satisfies a power law distribution,

335: $$

336: d\mu[r] \sim N \times r^{-\alpha} .

337: $$

338: If so, and if we could estimate $N$ and $\alpha$,

339: we could get a rough estimate for how many vacua have a rank

340: 4 gauge group, without much effort.  One could go on to study

341: the distribution $d\mu[N_1,N_2,\ldots]$ of the ranks of the simple

342: factors, etc.

343: Although this may sound ambitious, given that the function $d\mu[r]$ is

344: well defined, why shouldn't it have a simple approximate description ?

345: 

346: \subsection{Distribution of quiver gauge theories}

347: 

348: Let me turn to another problem along these lines, for which I can even

349: suggest a simple approximate description (details can be found in

350: \cite{mrdstat}).  We consider $U(N)$ quiver gauge theories, {\it i.e.}

351: with gauge group $\prod_i U(N_i)$, and a spectrum of purely

352: bifundamental matter.  Such theories arise on the world-volumes of

353: D-branes in type \II\ string theory, embedded in the $3+1$ observable

354: dimensions. and constructing these is the first step in making ``brane

355: world'' realizations of the Standard Model, as discussed in many talks

356: here.

357: 

358: It would be quite interesting to know the distribution of gauge groups

359: and matter content for the theories which come out of string

360: compactification.  Let us focus on part of this information: the

361: difference between the number of multiplets $(\bar N_i,N_j)$, and the

362: number $(\bar N_j,N_i)$.  For a theory with $K$ factors in the gauge

363: group, which would arise by wrapping branes on $K$ distinct cycles,

364: these are $K(K-1)/2$ {\it a priori} independent numbers;

365: we can summarize them in an

366: antisymmetric matrix, the ``intersection matrix'' $I_{ij}$.

367: 

368: These numbers counts chiral matter multiplets which cannot be lifted

369: by mass terms, and thus generalize the ``number of generations'' in

370: the Standard Model.  In explicit brane constructions, they are

371: entirely determined by the topology of the branes $B_i$ and $B_j$

372: carrying the $U(N_i)$ and $U(N_j)$ gauge groups.  The simplest example

373: of this is to consider D$6$-branes in Calabi-Yau compactification of

374: \IIa\ string theory: in this case, the intersection numbers are

375: literally topological intersection numbers between the

376: three-dimensional world-volumes of the branes in the CY.  Similar

377: formulas are known for the other types of branes, in all cases

378: topological.

379: 

380: Let us grant that the totality of type \II\ compactifications on CY,

381: with subsequent choices, leads to a finite set of 

382: vacua, each with a quiver gauge theory realized on

383: D-brane world-volumes.  If this set is finite, it defines a

384: distribution $d\mu[N_i; I_{ij}]$, the number of theories realizing

385: each possible choice of the $N_i$ and $I_{ij}$.

386: \p

387: 

388: In \cite{mrdstat},

389: we give arguments that as a matrix element $I_{ij}$ becomes large

390: (but not too large), this distribution goes as

391: \begin{equation}\label{eq:scaling}

392: d\mu[N_i; I_{ij}] \sim {dI_{ij}\over|I_{ij}|} ,

393: \end{equation}

394: (with a cutoff $I_{max} \sim \min(N_1,N_2)$, determined in string

395: theory compactification to be $I_{max} \sim 100$ by tadpole cancellation.)\p

396: 

397: The basic argument for this is that any given $U(N_1)\times U(N_2)$ theory

398: leads to a distribution $I\sim k,k^2,k^3,\ldots$

399: with this power-like falloff, and the ``total'' distribution obtained

400: by adding such distributions will also have this power-like falloff.

401: 

402: We then generalize to many gauge groups, by taking the

403: distributions of bifundamentals for each pair of gauge groups to be 

404: independent.  This ensemble has the great virtue of simplicity, and

405: should not be dismissed out of hand.  However, it is probably too simple,

406: as it ignores the fact that branes tend to wrap groups of cycles 

407: which intersect among themselves, and do not intersect between groups.

408: 

409: A better candidate ensemble of theories can be obtained by considering

410: $K\times K$ matrices which can be decomposed into blocks of size

411: $K_1\times K_1$, $K_2\times K_2$, and so on, and using (\ref{eq:scaling})

412: to describe the expected distribution of matter content in each block.

413: This is also very simple, and not obviously wrong.  

414: 

415: Of course, we are not claiming that this is an exact description of

416: the list of gauge theories coming out of string theory

417: compactification, only that it models some features of the true list.

418: One could go further and assert that, at the moment, this is the best

419: simple candidate description of the range of matter spectra coming

420: from brane constructions; it would be interesting to test it by

421: checking it against the families of models which have already been

422: concretely developed.  In any case, it is a precise ansatz which one

423: can use to study the fraction of models with a specified matter

424: content.

425: 

426: For example, one can

427: obtain the Standard Model by taking the gauge group

428: $U(3)\times U(2)\times U(1)\times U(1)$, and the

429: intersection matrix

430: $$

431: \left(\matrix{

432: 0& -3& 3& 3\cr

433: 3& 0& -1& 2\cr

434: -3& 1& 0& 0\cr

435: -3& -2& 0& 0}\right) ,

436: $$

437: and applying a subsequent orientifold projection.

438: 

439: In the ensemble (\ref{eq:scaling}) with $K=4$, the fraction of brane

440: models which realizes this spectrum is

441: $$

442: d\mu(-3)d\mu(3)d\mu(3)d\mu(-1)d\mu(2)d\mu(0) \sim 10^{-6} .

443: $$

444: 

445: If a given compactification has more than $4$ gauge groups, we need to

446: enumerate subsets of $4$ and compare them with the SM.  The fraction which

447: work depends on whether we allow exotic matter charged under the SM

448: gauge groups, which would live in the off-diagonal terms in the following

449: block decomposition:

450: $$

451: I = \left(

452: \matrix{ I_{SM}& I_{exotic}\cr -I_{exotic}^t& \ddots}\right)

453: $$

454: If we do not allow exotic matter, since most of the distribution has

455: $I_{ij}\ne 0$,

456: models in which the SM is realized in

457: a block with $K_i=4$ are very much favored.

458: 

459: While these ensembles are rather oversimplified, we believe that a

460: description of the true ensemble of brane gauge theories, which suffices

461: for this purpose, need not be too much more complicated.  One can refine

462: our estimate

463: by formulating more detailed ensembles, and comparing them with actual

464: string theory constructions.  We suspect this will

465: lead to similar results, say

466: $$

467: 10^{-16} < {N_{SM}\over N_{{\rm all}\ G,R}} < 1 .

468: $$

469: 

470: In any case, we have formulated a quantitative sense in

471: which the Standard Model matter content is ``generic.''  It is not

472: to say that most models have this spectrum; indeed the fraction which

473: do is small.  But it might be large, when compared with other numbers.

474: 

475: \section{Flux vacua}

476: 

477: Perhaps the most straightforward class of vacua to which to apply these

478: ideas is the set of ``flux vacua'' obtained by turning on gauge field

479: strengths in the compact dimensions.  These have been the focus of

480: much recent work, in which it has been shown that their contributions

481: to the vacuum energy can stabilize moduli.

482: 

483: Flux compactifications can even realize the

484: observed small positive vacuum energy, and in this sense solve the

485: cosmological constant problem.

486: A particularly simple proposal of this type was made by Bousso and

487: Polchinski \cite{boupol} (see also Feng {\it et al} \cite{feng}).

488: They argued that, in compactification on a Calabi-Yau with $K$

489: cycles, the number of vacua with small cosmological constant could

490: go as $c^K$, growing exponentially with the number of cycles.

491: Since typical Calabi-Yau's have $K\sim 100-500$ cycles, this suggests

492: that string theory could have a huge multiplicity of vacua, $N_{vac}

493: \sim 10^{100}-10^{500}$.  Furthermore, these vacua realize a spectrum or

494: ``discretuum'' of values for the cosmological constant, which is 

495: roughly uniform near zero.  Even if the distribution has no special

496: properties near zero, this makes it quite likely that vacua with

497: the small observed cosmological constant could exist just on statistical

498: grounds.

499: 

500: In more detail, let $F$ be a gauge field strength.

501: The equations of motion $\nabla F=0$ force it to be harmonic, so

502: determined by its integral over non-trivial homology cycles $\Sigma_\alpha$.

503: Let

504: $$

505:  N^\alpha  = \int_{\Sigma_\alpha} F

506: $$

507: be the quantized number of $F$ fluxes on the cycle $\Sigma_\alpha$,

508: and $K$ be the number of cycles.

509: \p

510: 

511: A qualitative description of the total energy is

512: is

513: $$

514: E = E_0 + {1\over l^4} \sum_{i=1}^K q_i(z)^2 N_i^2

515: $$

516: where $E_0$ is a flux-independent contribution, $q_i$ is a ``charge''

517: (determined by kinetic terms) and $l$ is the length scale

518: of the internal space.

519: 

520: \page

521: Suppose $E_0 < 0$ and $q\sim 1$, then the number of flux vacua with 

522: given $\Lambda = E(N)$ is roughly

523: \begin{eqnarray}

524: d\mu_{vac}(\Lambda) &\sim \int d^KN\ \delta(\Lambda-E) \\

525: &\sim \left(\Lambda - (E_0 l^4)\right)^{K/2-1} .

526: \end{eqnarray}

527: \p

528: 

529: Thus, the number of vacua with $|\Lambda| < \epsilon/l^4$ is roughly

530: $$

531: d\mu_{vac}(\Lambda\sim 0) \sim \epsilon L^{K/2-1}

532: $$

533: with $L=E_0 l^4$, substantiating the claims.

534: 

535: While we see the exponential emerge, and the large number of vacua,

536: this argument raises many questions: for example,

537: what determines the crucial parameters $E_0$ and $l$.

538: 

539: More to the point, this is only a heuristic argument, which ignored

540: the fact the fluxes {\bf back react} on the metric.  In the above

541: formula, this was expressed in the dependence $q_i(z)$ of the ``charges''

542: on moduli.  The actual energy is found by minimizing $z$, and this is

543: what determines the distribution of vacua in the moduli space.

544: 

545: Could it be that for many fluxes, there is no minimum apart from the

546: infinite volume limit?  Then, back reaction would eliminate most of

547: this supposed large number of vacua ?  Or, could most of them be dual

548: realizations of the same vacua ?

549: 

550: \subsection{Counting flux vacua}

551: 

552: In \cite{ad}, with Sujay Ashok, we answer this question by giving the

553: first precise estimate for the number of vacua in a family of

554: compactifications.  Without going into excessive detail, we work with

555: the \IIb\ compactifications developed by Giddings, Kachru and

556: Polchinski,\cite{gkp} in which the problem of finding vacua in the

557: full ten dimensional theory can be shown (in the large volume limit)

558: to precisely reduce to a problem in an $\CN=1$ effective supergravity

559: theory.

560: 

561: This effective theory has the following

562: chiral superfields:

563: $\tau=C^{(0)}+ie^{-D}$ the axion-dilaton, 

564: $z^i$ the complex structure moduli of $M$, 

565: and $\rho^i$ the K\"ahler moduli of $M$.

566: Their

567: K\"ahler potential is the same (up to truncating fields) as in

568: the related $\CN=2$ supersymmetric compactification with no flux,

569: $$

570: K(z,\bar z) = -\log \Im \bar z^i {\pp\CF(z)\over\pp z^i}

571:  -\log\Im\tau - 3\log\Im\rho .

572: $$

573: 

574: Besides a choice of CY and orientifolding, a flux compactification

575: sector is characterized by a quantized flux.  In \IIb\ theory these

576: are described by two integers for each three-cycle of $M$,

577: $$

578: N^\alpha \equiv N_{RR}^\alpha + \tau N_{NS}^\alpha =

579: \int_{\Sigma_\alpha} F^{(3)}_{RR} + \tau H^{(3)}_{NS} .

580: $$

581: The potential can then be computed exactly at large volume, using special

582: geometry and the

583: superpotential \cite{gvw}

584: \begin{eqnarray*} \label{eq:gvw}

585: W(z) &= &\int_M (F^{(3)}_{RR} + \tau H^{(3)}_{NS}) \wedge \Omega(z)

586: \equiv \int_M G \wedge \Omega(z) ; \\

587: \end{eqnarray*}

588: This $W$ and $K$ can be computed explicitly, using techniques 

589: developed in the study of mirror symmetry.

590: %\sref{Candelas, de la Ossa, Morrison, Katz, etc.}.

591: 

592: This superpotential depends on dilaton and complex structure moduli in

593: a fairly complicated way, and indeed stabilizes all of these moduli.

594: On the other hand, it does not depend on K\"ahler moduli, and as is

595: well known this $K$ exhibits ``no scale'' structure and does not

596: stabilize these moduli at all.  Let us completely neglect the K\"ahler

597: moduli for the time being and treat the others, returning to this

598: point below.

599: 

600: In this problem, it turns out that the allowed ``amount of flux,''

601: the number $L$ of the previous argument, is 

602: constrained by tadpole cancellation:

603: \begin{equation}\label{tadpole}

604: \int F \wedge H =

605: N(O3\ {\rm planes}) -

606: N(D3\ {\rm branes}) \equiv L

607: \end{equation}

608: 

609: Thus, the precise question we ask, is to count the number of

610: supersymmetric vacua, {\it i.e.} solutions of

611: $$

612: D_iW(z,\tau)= 0 ,

613: $$

614: satisfying this bound, as a function of $L$, up

615: to duality equivalences.  The duality group in this context (large

616: volume Calabi-Yau) is $SL(2,\BZ)$ acting on the dilaton-axion, times a

617: subgroup of $Sp(b_3,\BZ)$, which acts simultaneously on the choice of

618: flux, and on the complex structure.  A simple way to get one

619: representative of each duality class is to sum over all choices of

620: flux, but only count vacua which are stabilized at $(z,\tau)$

621: in a fundamental region (of both duality groups) in moduli space.

622: 

623: In \cite{ad}, we compute the large $L$ asymptotics for

624: an ``index'' which counts supersymmetric vacua

625: with signs, and is thus a lower bound for the total number of supersymmetric

626: vacua.  The general formula for this is

627: \begin{eqnarray} \label{eq:dist}

628: I_{vac}(L \le L_{max})

629:  &=&  {(2\pi L)^{b_3}\over \pi^{n+1} b_3!}\int_{\CF\times\CH} 

630: \det(-R-\omega\cdot 1) ,

631: \end{eqnarray}

632: where $\CF$ is a fundamental region in the complex structure moduli space,

633: and $\CH$ is the fundamental region of $SL(2,\BZ)$ in dilaton-axion moduli

634: space.  This formula will have corrections in a series in $1/L$, but these

635: should be small if $L >> K$.

636: 

637: An important point about the result is that it gives not just the

638: number, but the actual distribution of points where 

639: the flux vacua stabilize the moduli.  The term $\det(-R-\omega\cdot 1)$ which

640: appears under the integral sign, is a $2n+2$ form, derived from

641: the K\"ahler form $\omega$, and the matrix of curvature two-forms $R$.

642: If we integrate this form over a subregion in moduli space, we obtain

643: the number of vacua which sit in this region.

644: These results can be

645: used to compare the numbers of vacua in different regimes 

646: of coupling and field space, as we will discuss in \cite{dd}.

647: 

648: While mathematical techniques exist to do this type of integral, 

649: the actual result is known at present only for

650: compactification on $T^6/\BZ_2$, for which

651: $K=b_3=20$ and $L=32$: one finds 

652: $$

653: I_{vac} = {7\cdot (2\pi L)^{20}\over 4\cdot 181440\cdot 12\cdot 20!} 

654: \sim 4\cdot 10^{21}.

655: $$

656: 

657: To complete the discussion, we need to discuss

658: stabilization of K\"ahler moduli.

659: Now no-scale structure is generically

660: spoiled by $\alpha'$ and non-perturbative corrections, say

661: $$

662: W_{NP} = e^{i N\rho} + \ldots .

663: $$

664: which, it has been suggested \cite{KKLT},

665: can be arranged to arise from some brane world-volume theory.

666: Very generally,

667: a solution of $DW(z)=0$ for the complex structure moduli,

668: with $e^K|W|^2 << M_{pl}^4$, will become a stable supersymmetric AdS

669: vacuum once these are taken into account.  \cite{KKLT,mrdstat}

670: For example,

671: $$

672: 0 = D_\rho W = iNe^{iN\rho} - {3\over\rho-\bar\rho}W_{rest} 

673: $$

674: has a solution for

675: $$

676: {2N\over 3}(\Im\rho) e^{iN\rho} = W_{rest} .

677: $$

678: The function on the l.h.s. can take any value up to $2/3e \sim 1$, and

679: one expects an exact nonperturbative $W(\rho)$ to behave similarly.

680: Thus, any vacuum with $W_{rest}$ not too large, can be stabilized.\p

681: 

682: Thus, we need to know the distribution of AdS cosmological constants.

683: This distribution can be computed in the same way as above; details

684: will appear in \cite{dd}.  One finds that the distribution is ``uniform near

685: zero,'' in other words the number of vacua with small

686: $\Lambda=\epsilon << L$ goes as $N_{vac} \times \epsilon b_3/L$. 

687: 

688: There is a simple intuitive argument for this result.\cite{argument}

689: The AdS

690: cosmological constant can be regarded as the length squared of a

691: ``vector in flux space'' defined by the values of the various periods

692: of cycles entering (\ref{eq:gvw}).  If we assume that this vector is

693: totally uncorrelated with the specific choice of flux, all vacua with

694: flux roughly orthogonal to this vector will have small cosmological

695: constant, leading to this result.  Now this assumption is not innocent

696: and is false in very similar examples (such as flux in the heterotic

697: string); however the actual computation in type \II\ does produce this

698: result.

699: 

700: Thus, taking $\epsilon\sim 10^{-3}$, we obtain a lower bound

701: $4\cdot 10^{18}$ on the number of flux vacua on $T^6/\BZ_2$.  

702: One might be even more

703: conservative and cut out the strong coupling regime, where the analysis

704: is not presently under control.  If we do this by

705: insisting on $\Im\tau > 40$, we find that $10^{17}$ 

706: vacua satisfy these constraints.\p

707: 

708: \section{Conclusions}

709: 

710: We discussed the statistical approach to string phenomenology, and

711: gave the first computation of the number of flux vacua in a specific

712: model.  This result confirmed the suggestion of Bousso and Polchinski

713: that this number should grow as $L^K$,

714: exponentially in the number of cycles $K$,

715: and is a first step in getting a solid idea of just how

716: many string theory vacua there are.

717: 

718: CY's are known with $K\sim 500$ and $L\sim 10^4$, so

719: there is some danger that the number of vacua is large enough to spoil

720: predictivity.  

721: We can roughly quantify the number at which we should start to worry,

722: as follows \cite{mrdstat}.  Consider the twenty dimensional parameter

723: space of the Standard Model couplings, along with the cosmological

724: constant.  Now, compute the volume in this space consistent with

725: present-day observations, where we measure each coupling in Planck

726: units (so, a coupling $\lambda$ of dimension $n$ has measure 

727: $M_{pl}^{-n} \int d\lambda$.  For the Standard Model, this comes out

728: to about $10^{-240}$, where $10^{-120}$ of this comes from matching

729: the cosmological constant, and the rest from the Higgs mass and other

730: couplings.

731: 

732: Now, suppose we consider string models which match the spectrum of the

733: Standard Model at low energies.

734: As we discussed, the fraction of four

735: dimensional models which do this is surely much greater than

736: $10^{-100}$, and $10^{-10}$ would seem to be a reasonable guess.

737: 

738: Take these models, and plot their couplings in this twenty dimensional

739: space.  The computation we discussed is also a first step towards

740: doing this, as it gives the distribution of vacua in moduli space, and

741: these moduli will control the couplings in the resulting low energy

742: theory.  To some approximation, the result (\ref{eq:dist}) says

743: that models are roughly uniformly distributed in moduli space; this

744: suggests that the dimensionless couplings could come out roughly

745: uniformly distributed as well.  This is less clear for dimensionful

746: couplings, and discussing these requires taking supersymmetry and its

747: breaking into account in more detail.  Let us continue however to

748: illustrate the idea, and return to this point.

749: 

750: Then, we can distinguish various cases.  If the number of models is

751: less than $10^{120}$, we do not expect to match the cosmological

752: constant, and need to find mechanisms which produce the observed small

753: value, to reasonably expect string theory to reproduce the data.  If

754: it is between $10^{120}$ and $10^{240}$, we expect models which

755: reproduce the cosmological constant to exist, and can in principle

756: test string theory by checking that the observed Standard Model

757: couplings are possible in some compactification -- just on statistical

758: grounds, this would be unlikely.

759: 

760: On the other hand, suppose there were many more than $10^{240}$

761: models, say $10^{1000}$ for definiteness.  In this case, one would

762: expect a vast number of models to reproduce the Standard Model in

763: every detail, just on statistical grounds.  If it furthermore turned

764: out that they led to many different predictions at higher energies,

765: one would clearly have grounds to worry that string theory was not

766: testable.

767: 

768: Of course supersymmetry changes these numbers.  With low scale breaking

769: (say $10 \TeV$), the corresponding ``volume in coupling space'' becomes

770: about $10^{-120}$.  Thus, unless the ratio of nonsupersymmetric

771: candidates to these candidates is greater than $10^{120}$, these

772: considerations might be regarded as favoring supersymmetry, as discussed

773: at more length in \cite{BDG}.  

774: 

775: There are some important points to be made here.  First, it is

776: essential to have some estimate of the {\it a priori} numbers, here

777: the ratio between the numbers of nonsupersymmetric and

778: supersymmetric vacua, to make this claim.  It is true that one does not

779: need a very accurate estimate of this ratio (here any accuracy better than

780: $10^{100}$ would suffice), but one does need some estimate, for this

781: argument to have any content beyond the standard argument

782: for supersymmetry from naturalness.

783: 

784: Second, it should be realized that ``favoring'' one mechanism over

785: another is interesting but not in itself decisive, if in fact both

786: types of vacuum exist.  We only observe one vacuum, and it might be of

787: either type.  On the other hand, suppose we estimated the number of

788: supersymmetric vacua which pass all tests as $10^{40}$, while the

789: number of non-supersymmetric vacua was estimated as $10^{-40}$.  This

790: could come out of the type of approximate estimates we are discussing,

791: and would mean that we need a coincidence which tunes parameters at

792: the $10^{40}$ level to get a candidate nonsupersymmetric vacuum.  If

793: our estimates were reliable, then the most reasonable interpretation

794: of such evidence would be that there are in fact no candidate

795: nonsupersymmetric vacua, and we would get a strong prediction from the

796: statistics of vacua.  It is this goal which motivates trying to get

797: controlled estimates for these numbers.

798: 

799: As a final comment, it is interesting that the numbers which are

800: coming out, of order $10^{100}$, are of the general order we need to

801: solve the cosmological constant problem, while not obviously spoiling

802: predictivity.  Thus, the picture we just outlined need not be

803: depressing -- there are many further conditions the correct vacuum

804: must satisfy, and it may be that further work along these lines will

805: demonstrate that only one or a few vacua work.  In any case, our main

806: point is that these are questions about string theory which, if the

807: theory has a precise definition, have definite answers, and that are

808: now becoming accessible to investigation.

809: 

810: \smallskip

811: 

812: I would like to thank S. Ashok, T. Banks,

813: F. Denef, S. Dimopoulos, M. Dine, S. Kachru, G. Kane, J. Maldacena,

814: G. Moore, B. Shiffman, S. Thomas, S. Trivedi, E. Witten and S. Zelditch

815: for collaboration and valuable discussions.

816: 

817: \begin{thebibliography}{99}

818: \bibitem{ad}

819: S. Ashok and M. R. Douglas, hep-th/\adref.

820: 

821: \bibitem{BDG}

822: T. Banks, M. Dine and E. Gorbatov,

823: hep-th/0309170.

824: 

825: \bibitem{boupol}

826: R. Bousso and J. Polchinski,

827: JHEP 0006 (2000) 006, hep-th/0004134.

828: 

829: \bibitem{Canska}

830: P. Candelas and H. Skarke, hep-th/9706226.

831: 

832: \bibitem{dd}

833: F. Denef and M. R. Douglas, to appear.

834: 

835: \bibitem{dns}

836: M. Dine, Y. Nir and Y. Shadmi,

837: Phys.Lett. B438 (1998) 61-68, hep-th/9806124.

838: 

839: \bibitem{dine}

840: M. Dine, ``TASI Lectures on M Theory Phenomenology,''

841: hep-th/0003175.

842: 

843: \bibitem{mrdstat}

844: M.~R.~Douglas,

845: JHEP {\bf 0305}, 046 (2003)

846: hep-th/0303194.

847: 

848: \bibitem{feng}

849: J. L. Feng, J. March-Russell, S. Sethi and F. Wilczek,

850: Nucl.Phys. B602 (2001) 307; hep-th/0005276.

851: 

852: \bibitem{gkp}

853: S. B. Giddings, S. Kachru and J. Polchinski,

854: Phys.Rev. D66 (2002) 106006; hep-th/0105097.

855: 

856: \bibitem{gvw}

857: S.~Gukov, C.~Vafa and E.~Witten,

858: Nucl.\ Phys.\ B {\bf 584}, 69 (2000) [Erratum-ibid.\ B {\bf 608}, 477

859: (2001)] hep-th/9906070.

860: 

861: \bibitem{KKLT}

862: S.~Kachru, R.~Kallosh, A.~Linde and S.~P.~Trivedi,

863: hep-th/0301240.

864: 

865: \bibitem{argument}

866: M.~R.~Douglas, unpublished; S.~Kachru, unpublished; also G.~Moore 

867: (private communication) has made related

868: observations about the closely related attractor problem.

869: 

870: \end{thebibliography}

871: 

872: \end{document}

873: 

874: 

875: