1: \documentclass[prd,preprintnumbers,twocolumn,nofootinbib,showpacs]{revtex4}

2: %\documentclass[preprint,preprintnumbers,nofootinbib,showpacs]{revtex4}

3: 

4: \usepackage{amsfonts,amsmath,amssymb}

5: \usepackage{bm}

6: \usepackage{graphicx,epsfig}

7: \usepackage{dcolumn}

8: 

9: %%%%%

10: %%%%%

11: 

12: \newcommand{\gsim}{\mathrel{\lower2.5pt\vbox{\lineskip=0pt\baselineskip=0pt

13:                    \hbox{$>$}\hbox{$\sim$}}}}

14: \newcommand{\lsim}{\mathrel{\lower2.5pt\vbox{\lineskip=0pt\baselineskip=0pt

15:                    \hbox{$<$}\hbox{$\sim$}}}}

16: 

17: \newcommand{\too}{\longrightarrow}

18: \newcommand{\Too}{\Longrightarrow}

19: 

20: \newcommand{\abs}[1]{\left| #1 \right|}

21: \newcommand{\avg}[1]{\left\langle #1 \right\rangle}

22: \newcommand{\bra}[1]{\left\langle #1 \right|}

23: \newcommand{\ket}[1]{\left| #1 \right\rangle}

24: \newcommand{\braket}[2]{\left\langle #1 \vert #2 \right\rangle}

25: \newcommand{\sla}[1]{{\raise.15ex\hbox{$/$}\kern-.57em #1}}

26: \newcommand{\Sla}[1]{\kern0.12em{\raise.15ex\hbox{$/$}\kern-.74em #1}}

27: 

28: \newcommand{\mev}{{\rm meV}}

29: \newcommand{\ev}{{\rm eV}}

30: \newcommand{\kev}{{\rm keV}}

31: \newcommand{\Mev}{{\rm MeV}}

32: \newcommand{\gev}{{\rm GeV}}

33: \newcommand{\tev}{{\rm TeV}}

34: \newcommand{\Mpl}{M_{\rm P\ell}}

35: \newcommand{\Msun}{M_{\odot}}

36: \newcommand{\mW}{m_{\text{{\tiny $W$}}}}

37: 

38: \newcommand{\beq}{\begin{eqnarray}}

39: \newcommand{\eeq}{\end{eqnarray}}

40: \newcommand{\nn}{\nonumber}

41: \newcommand{\eql}[1]{\label{eq:#1}}

42: \newcommand{\eq}[1]{(\ref{eq:#1})}

43: 

44: \newcommand{\del}{\partial}

45: \newcommand{\dotprod}{\!\cdot\!}

46: \newcommand{\e}[1]{{\rm e}^{#1}}

47: \newcommand{\colmvec}[1]{\left( \begin{array}{c} #1 \end{array} \right)}

48: \newcommand{\order}[1]{{\mathcal O}\!\left( #1 \right)}

49: 

50: \newcommand{\ie}{\textit{i.e.}}

51: \newcommand{\eg}{\textit{e.g.}}

52: \newcommand{\hc}{{\rm H.c.}}

53: \newcommand{\ellg}{\ell_g}

54: \newcommand{\Lg}{\Lambda_g}

55: \newcommand{\Lgt}{\widetilde{\Lambda}_g}

56: \newcommand{\chit}{{\widetilde{\chi}}}

57: \newcommand{\Xt}{{\widetilde{X}}}

58: \newcommand{\micr}{{\rm \mu m}}

59: 

60: %%%%%

61: %%%%%

62: 

63: \begin{document}

64: 

65: \preprint{BUHEP-05-15}

66: \preprint{hep-ph/0511082}

67: 

68: \title{Probing Composite Gravity in Colliders}

69: 

70: \author{Takemichi Okui}

71: \affiliation{Physics Department, Boston University,\\

72:              590 Commonwealth Ave., Boston, MA 02215, U.S.A.}

73: 

74: \pacs{12.60.Rc, 04.90.+e}

75: 

76: \begin{abstract}

77: We explore scenarios in which the graviton is not a fundamental degree of 

78: freedom at short distances but merely emerges as an effective 

79: degree of freedom at long distances.  In general, the scale of such 

80: graviton `compositeness', $\Lg$, can only be probed by measuring gravitational 

81: forces at short distances, which becomes increasingly difficult and eventually

82: impossible as the distance is reduced.  Here, however, we point out that if 

83: supersymmetry is an underlying symmetry, the gravitino can be used as an 

84: alternative probe to place a limit on $\Lg$ in a collider environment, 

85: by demonstrating that there is a model-independent relation, 

86: $\Lg \gsim m_{3/2}$.  In other 

87: words, the gravitino knows that gravity is standard at least down to its 

88: Compton wavelength, so this can also be viewed as a test of general 

89: relativity possible at very short distances.  If composite gravity is found 

90: first at some $\Lg$, this would imply a model-independent upper bound on 

91: $m_{3/2}$.        

92: \end{abstract}

93: 

94: \maketitle

95: 

96: %%%%%

97: %%%%%

98: 

99: \section{Introduction}

100: \label{sec:intro}

101: %

102: Gravity at short distances is a vastly unexplored experimental frontier.

103: It is possible that a deviation or even a drastic departure from the 

104: standard gravitational law may be found in future experiments.  

105: On the theoretical side, we have string theory which replaces general 

106: relativity (GR) at distances shorter than the string scale $M_s^{-1}$.  

107: However, since string theory not only modifies gravity but also governs the 

108: matter sector, the fact that we have not observed any stringy phenomena in 

109: particle physics experiments requires $M_s$ to be higher than at least a 

110: few $\tev$.

111: 

112: In contrast, for a theory which modifies \textit{only} gravity, 

113: the bound on the scale of such new short-distance gravitational physics is 

114: significantly lowered to 

115: $(\order{100}\,\micr)^{-1} \approx \order{10^{-3}}\,\ev$ 

116: \cite{Hoyle:2004cw} \cite{Long:2003dx, Smullin:2005iv} (also see the review 

117: \cite{Adelberger:2003zx}), which is 15 orders of magnitude larger than 

118: $\tev^{-1} \approx 10^{-17}\,{\rm cm}$!  Therefore, there is {\it huge}

119: room for a theory of this kind.  This situation is quite intriguing, and this 

120: is the window that we will explore in this paper.

121:   

122: The striking fact about this range 

123: between $100\,\micr$ and $10^{-17}\,{\rm cm}$

124: is that we {\it know} that matter is 

125: described by the standard local relativistic quantum field theory there.  

126: The standard model (SM) has been tested including nontrivial loop

127: corrections with great precision \cite{:2005em}.  This point cannot be 

128: emphasized too much.  It means that a modification of gravity in this range

129: cannot be as radical as, for example, abandoning the notion of a continuum 

130: spacetime; when we say the Bohr radius is $0.509$~\AA, we know perfectly 

131: what we are talking about!  So, while we will boldly speak of modifying 

132: gravity in this paper, \textit{we will not mess around with matter;} 

133: we take it for granted that the matter sector is completely 

134: \textit{normal,} \ie, perfectly described by a local relativistic 

135: quantum field theory.

136: 

137: It should be also mentioned that, in general, changing the laws of gravity 

138: does not necessarily mean modifying or abandoning GR.  For example, if we 

139: add $n$ extra spatial dimensions with the size $L$ in which only gravitons may 

140: propagate, then the Newton's law changes from $1/r^2$ to $1/r^{2+n}$ for 

141: $r \ll L$ \cite{Arkani-Hamed:1998rs}.  But gravity in this example is 

142: perfectly governed by the conventional GR; it is just living in more 

143: dimensions than four.  

144: 

145: In this paper, however, we \textit{will} explore the possibility that GR 

146: is abandoned at short distances in the sense that the graviton is \textit{not} 

147: a fundamental propagating degree of freedom (d.o.f.)~in whatever underlying 

148: theory, but is merely an effective d.o.f.~appropriate at long distances.  

149: The scale, which we call $\Lg^{-1}$, corresponding to the boundary between 

150: `short-distances' and `long-distances' could be anywhere shorter than 

151: $\order{100}\,\micr$, but as we stated above, we will focus on the range

152: $10^{-17}\,{\rm cm} \lsim \Lg^{-1} \lsim 100\,\micr$ (or 

153: $10^{-3}\>\ev \lsim \Lg \lsim \tev$), so that we can exploit the fact that 

154: the matter sector is `normal'.

155: 

156: This includes various possibilities---the graviton may be a bound 

157: or solitonic state of the fundamental d.o.f.~\cite{Sundrum:1997js}, or an 

158: extended state in some intrinsically nonlocal theory 

159: \cite{Moffat:2001jf, Sundrum:2003tb}, or a sort of hydrodynamic state as in 

160: the scenarios often dubbed `emergent relativity' \cite{Unruh:1980cg}.

161: We will not distinguish these varieties but just focus on their common 

162: feature that the graviton is not an elementary propagating d.o.f.~in the 

163: fundamental theory but just appears as an effective d.o.f.~in the long-distance 

164: description for $d>\Lg^{-1}$.  Admittedly not an optimal name, we call it a 

165: {\it composite} graviton, {\it where by `composite' we simply mean `not 

166: elementary'.} 

167: 

168: %

169: \begin{figure*}

170: \includegraphics{fig.eps}

171: \caption{Diagrams representing (a) corrections to the cosmological constant,

172: (b) corrections to the gravitational mass, and (c) corrections to the inertial

173: mass.  A solid line represents a heavy matter particle, and a double wavy 

174: line represents a graviton.}

175: \label{fig:diagrams}

176: \end{figure*}

177: %

178: 

179: One may think such a composite graviton is excluded by the theorem by 

180: Weinberg and Witten \cite{Weinberg:1980kq}.  Actually, what the 

181: Weinberg-Witten (WW) theorem excludes is not just a composite graviton but 

182: \textit{any} massless spin-1 or -2 particle, {\it composite or not!}  

183: Therefore, we must be careful about the assumptions of the theorem; we all 

184: know QED and QCD which have a massless spin-1 particle, and GR which has a 

185: massless spin-2 particle.  Note that the WW theorem 

186: states that if a theory allows the existence of a Lorentz-covariant conserved 

187: vector (or symmetric 2nd-rank tensor) current, then the theory cannot contain 

188: any massless spin-1 (or spin-2) particle charged under this current.  QED 

189: evades the spin-1 part of the theorem because the photon is not charged under 

190: the current.  QCD evades the spin-1 part of the theorem because the current is 

191: not Lorentz covariant due to its dependence on the gluon field which is a 

192: 4-vector only up to a gauge transformation.  Similarly, GR evades the spin-2 

193: part of the theorem because the gravitational part of the energy-momentum 

194: `tensor' is not really a tensor in GR. 

195: 

196: Indeed, there is an {\it explicit} example of composite gauge bosons.  

197: Consider an $SU(N)$ supersymmetric QCD with $F$ flavors where 

198: $N+1 < F < 3N/2$, with no superpotentials.  In the far infrared (IR), this 

199: theory is described by an IR-free, weakly coupled $SU(F-N)$ gauge theory 

200: \cite{Seiberg:1994pq}.  However, these IR gauge bosons 

201: are {\it not}  a subset of the original ultraviolet (UV) d.o.f.; rather, 

202: they are new effective d.o.f.~appearing only in the IR description, which 

203: microscopically can be interpreted as solitonic states of the fundamental UV 

204: d.o.f.~\cite{Seiberg:1994pq}.  So this is indeed a \textit{concrete} example 

205: of composite massless gauge bosons, where the $SU(F-N)$ gauge symmetry emerges

206: at low energies, making it consistent with the WW theorem.  

207: 

208: Clearly, it is desirable to have a similar example for gravity.  To this 

209: goal, Gherghetta, Peloso and Poppitz recently presented a theory in a 

210: 5-dimensional Anti-de-Sitter (AdS) space which is dual to a 4-dimensional 

211: conformal field theory in which the conformal symmetry is dynamically broken

212: in the IR yielding a spectrum containing a massless spin-2 resonance 

213: \cite{Gherghetta:2005se}.  To complete their picture, analyses beyond the 

214: quadratic order in action must be performed, especially concerning effects 

215: of the stabilization of the AdS space, and if the delicate existence of the 

216: massless spin-2 state persists, this will be a solid, concrete example of 

217: a theory of a composite graviton.  (Note that the graviton in string theory 

218: is completely elementary.) 

219: 

220: Is there any reason or motivation to consider such drastic modification of 

221: gravity in this range?  Just near the edge of the range, there is a 

222: cosmologically interesting scale 

223: $\approx (20\,\micr)^{-1} \approx 10^{-2}\>\ev 

224: \approx (16\pi^2\rho_{\rm vac})^{1/4}$, where 

225: $\rho_{\rm vac}$ is the vacuum energy density corresponding to the observed 

226: acceleration of the expansion of the universe 

227: \cite{Perlmutter:1998np, Spergel:2003cb, Tegmark:2003ud}.  Kaplan and Sundrum 

228: also recently pointed out that the interesting scale in the context of the

229: cosmological constant problem (CCP) may instead be $\order{10}\,\Mev$ 

230: \cite{Kaplan:2005rr}.  

231: Therefore, it is quite interesting to ask if composite gravity can solve the

232: CCP by identifying $\Lg$ with, say, $10^{-2}\>\ev$.

233: However, it is not so hard to see the answer entirely depends on the nature of 

234: whatever underlying theory of composite gravity.

235: 

236: In particular, it appears that the underlying theory should not be a local

237: field theory if one wishes to suppress loop corrections to the cosmological 

238: constant by abandoning elementary gravitons \cite{Sundrum:2003jq}.  The 

239: argument goes as follows.  Consider three diagrams in FIG.~\ref{fig:diagrams}.

240: The diagram (\ref{fig:diagrams}-a) is a correction to the vacuum energy, 

241: (\ref{fig:diagrams}-b) is a correction to the gravitational mass, and 

242: (\ref{fig:diagrams}-c) is a correction to the inertial mass.  In a field 

243: theory, the loop integral in (\ref{fig:diagrams}-a) can be suppressed only if 

244: the vertex has a form factor that depends on the loop momentum.  Now, the 

245: problem is, once (\ref{fig:diagrams}-a) is suppressed by such a form factor, 

246: the correction (\ref{fig:diagrams}-b) also gets suppressed because it has the 

247: same form factor, while the correction (\ref{fig:diagrams}-c) does not get 

248: suppressed because there is no such form factor.  This violates the 

249: equivalence principle, and we need fine-tuning to restore it.  However, for a 

250: composite graviton which is not from a local field theory, there does not have

251: to be tension like this, 

252: and suppressing loop corrections to the cosmological constant may 

253: be consistent with the equivalence principle.  But even supposing we did

254: find such a nonlocal underlying theory, it would still be 

255: halfway to solving the cosmological constant problem, since there 

256: are also {\it tree-level} or classical contributions to the vacuum energy from 

257: phase transitions which must be somehow suppressed.  The door is not shut yet, 

258: and Ref.~\cite{Sundrum:2003tb} discusses a toy model for such a nonlocal theory 

259: without problems with the equivalence principle or the classical contributions.  

260: In the rest of the paper, we will not concern ourselves with the cosmological 

261: constant problem any further, and just focus on the physics of composite 

262: gravity. 

263: 

264: So, supposing that the graviton is not an elementary d.o.f.~in whatever 

265: fundamental theory, how do we see it?  Without having a concrete microscopic 

266: model of composite gravity, the scale $\Lg$ is the only quantity we can 

267: discuss.  So far, the lower bound on $\Lg$ has been placed by measuring 

268: gravitational forces between test masses, which has reached the scale of 

269: $\order{100}\,\micr$.  But it is clear that such direct measurement will be 

270: increasingly difficult and eventually impossible as the distance gets reduced.

271: Soon, some other methods must replace it to probe the scale of composite 

272: gravity.  

273: 

274: Such an alternative can arise if there is something that is related to the 

275: graviton 

276: but is more accessible than the graviton at short distances.  In general, 

277: there is nothing that is related to gravity except the graviton itself.  

278: However, if nature possesses (spontaneously broken) supersymmetry (SUSY), the 

279: {\it gravitino} precisely satisfies the criteria---it is related to the 

280: graviton and may be accessible even in colliders!  

281: The introduction of SUSY allows us to extract some informations 

282: relating the graviton and the gravitino without knowing what the underlying

283: theory is.  In fact, we will show that if a gravitino 

284: exists, it can indeed be used to probe gravity at very short distances where 

285: direct measurement of gravitational forces is impossible. 

286: 

287: To keep our discussions as model independent as possible, we would like to

288: have an effective field theory and ask questions that can be answered by

289: it.  This effective theory must have the following features:

290: \begin{itemize}

291:  \item{It must contain a {\it physical} scale $\Lg$ above which the graviton is

292:        no longer an elementary degree of freedom.  The scale $\Lg$ is not a 

293:        scale chosen for convenience but corresponds to a physical boundary 

294:        between two completely different phases of the theory, just like 

295:        $\Lambda_{\rm QCD}$ separates two different descriptions with totally 

296:        different degrees of freedom (\ie~partons versus hadrons).}

297:  \item[]{(Recall that $\Lg$ is a parameter anywhere from $\order{10^{-3}}\,\ev$ 

298:        to $\order{\tev}$ or whatever cutoff for the matter sector.)}

299:  \item{Nevertheless, to reproduce all the known gravitational physics, it 

300:        must include all the matter particles, \textit{even the ones heavier than 

301:        $\Lg$!}  And, as emphasized already, we know that the 

302:        matter sector is perfectly described by a local relativistic quantum 

303:        field theory with a cutoff higher than $\tev > \Lg$.}

304: \end{itemize}

305: Because of the second feature, we cannot use the usual effective field theory 

306: formalism in which all the particles heavier than $\Lg$ are simply integrated 

307: out; that would fail to capture all the {\it known} long-distance gravitational 

308: physics such as the $1/r^2$ law, the perihelion precession, the bending of 

309: light, etc.

310: 

311: Therefore, the first important question is whether or not there exists a 

312: sensible effective theory that can deal with this highly asymmetric situation 

313: in which gravity has a low cutoff and matter has a high cutoff.  This question 

314: was answered by R.~Sundrum, who developed a formalism, {\it soft graviton 

315: effective theory} (SGET) \cite{Sundrum:2003jq}, which assures that we can 

316: consistently analyze this asymmetric situation without referring to the 

317: underlying theory of composite gravity.  We will review the essential ideas 

318: of SGET in Sec.~\ref{sec:SGET} to keep our discussions self-contained.

319: 

320: Given that there is a consistent effective field theory to describe the 

321: low-cutoff gravity with the high-cutoff heavy matter, there seems nothing 

322: wrong to have a gravitino heavier than $\Lg$, since we should be able to 

323: treat it just as one of heavy matter particles.  After all, $\Lg$ is the 

324: scale of {\it graviton's} compositeness which does not have to be equal to

325: that of {\it gravitino's} once supersymmetry is broken.  Also, there is 

326: nothing wrong {\it a priori} for a composite particle to be heavier than the 

327: scale of its compositeness, like the $B$-mesons, the hydrogen atom, etc.  

328: 

329: Nevertheless, as we will show in Sec.~\ref{sec:SusySGET}, there is a nontrivial

330: remnant of the underlying supersymmetry which gives rise to the relation

331: \beq

332:    m_{3/2} \lsim \Lg ~.

333: \eeq 

334: Therefore, in fact a gravitino---if it exists---{\it knows} that gravity 

335: should be just GR (\ie~the graviton is an elementary d.o.f.)~at least down to 

336: its Compton wavelength!  In other words, the discovery of a gravitino and the 

337: measurement of its mass offers a short-distance test of GR and places a 

338: {\it model-independent} lower-bound on $\Lg$!  In particular, depending on the 

339: value of $m_{3/2}$, we may be able to completely exclude the possibility of 

340: composite gravity as a solution to the CCP.  

341: 

342: On the other hand, if we first 

343: discover composite gravity somehow and measure $\Lg$ before discovering a 

344: gravitino, then this inequality predicts that, once we see a gravitino, we 

345: will find its mass be lighter than $\Lg$.  

346: 

347: In Sec.~\ref{sec:existence} and \ref{sec:relation}, we will continue the

348: discussions to gain a further understanding of the inequality, followed 

349: by a brief comment in Sec.~\ref{sec:theo-tests} on the possibility of 

350: independent theoretical tests of the inequality. 

351: 

352: In order for our prediction to be useful, it is clearly crucial to 

353: experimentally convince ourselves that what we are observing is really a 

354: gravitino, not a random spin-$3/2$ resonance which may just happen to be 

355: there.  This issue will be discussed in Sec.~\ref{sec:detection}.  We will 

356: then conclude in Sec.~\ref{sec:conc}.

357: 

358: 

359: \section{Soft Graviton Effective Theory}

360: \label{sec:SGET}

361: %

362: As we have already mentioned, we need to describe all 

363: experimentally known gravitational physics occurring among heavy ($\gg \Lg$) 

364: matter particles, without extrapolating our knowledge of gravity beyond $\Lg$.

365: Soft graviton effective theory (SGET) \cite{Sundrum:2003jq} is designed 

366: precisely for this purpose.%  

367: %  

368: \footnote{Strictly speaking, to describe the typical observed gravitational 

369: phenomena involving gravitational bound states, we should switch to yet another 

370: effective field theory to have a transparent power-counting 

371: scheme appropriate for that purpose.  The interested reader should read 

372: Ref.~\cite{Goldberger:2004jt} which develops such an effective theory, 

373: dubbed `nonrelativistic general relativity' (NRGR).}

374: %

375: Here, we will review its central concepts to keep the discussions 

376: self-contained.

377:   

378: To start, let us consider gravity only.  In this case, the theory takes the 

379: form of a familiar effective field theory with the cutoff $\Lg$ imposed on 

380: the graviton field $h_{\mu\nu}$ defined via

381: \beq

382:    g_{\mu\nu} = \eta_{\mu\nu} + \frac{h_{\mu\nu}}{\Mpl} ~.

383: \eeq

384: Namely, the lagrangian is just the usual Ricci scalar term plus 

385: a whole series of higher-dimensional operators suppressed by powers of $\Lg$:

386: \beq

387:    {\cal L}_{\rm grav} 

388:       \sim \Mpl^2 \left( {\cal R} 

389:                          + \frac{{\cal R}^2}{\Lg^2}

390:                          + \frac{R_{\mu\nu} R^{\mu\nu}}{\Lg^2}

391:                          + \cdots

392:                          \right)  ~,

393: \eql{puregrav}

394: \eeq

395: where dimensionless $\order{1}$ coefficients are suppressed.%

396: %

397: \footnote{The operator $R_{\mu\nu\rho\sigma} R^{\mu\nu\rho\sigma}$ can be 

398: omitted in perturbation theory since it can be expressed as a linear 

399: combination of the two operators explicitly written in \eq{puregrav} plus

400: a total derivative.  Furthermore, in the absence of matter, even those two

401: operators could be removed by field redefinition, but we have kept 

402: them in \eq{puregrav} because we are interested in including matter.  

403: See Refs.~\cite{Donoghue:1995cz} for more discussions on these operators.}

404: %

405: As we mentioned earlier, $\Lg$ is a physical scale above which $h_{\mu\nu}$ 

406: is no longer an elementary degree of freedom.    

407: 

408: Note the scales and kinematic configurations to which this 

409: ${\cal L}_{\rm grav}$ is 

410: applicable.  It is appropriate only for processes where all momentum 

411: transfers among the gravitons are less than $\order{\Lg}$.  For example, it 

412: can {\it not} be used to calculate the cross-section 

413: for two highly energetic ($E \gg \Lg$) gravitons scattering with a large 

414: angle.  In fact, we do not even know if such a scattering occurs at 

415: all---maybe they would end up with `jets', like in hadron-hadron 

416: collisions---who knows?  No experiments so far have told us what would happen to 

417: such processes, and performing theoretical calculations requires specifying 

418: the full theory valid at distances shorter than $\Lg^{-1}$.  The moral here

419: is that a large momentum transfer should not be delivered to a 

420: graviton within our effective theory.

421: 

422: To also understand that a graviton should not be {\it exchanged} 

423: to mediate a large momentum transfer, imagine a theory with a fermion 

424: $\psi$ and a scalar $\phi$, and suppose we have verified that a Yukawa coupling 

425: $\phi\bar{\psi}\psi$ perfectly describes the $\psi\psi \to \psi\psi$ scattering

426: when both $\psi$'s get only very {\it low} recoils, \ie~the momentum transfer 

427: mediated by $\phi$ is very small.  But it may be completely wrong to use this 

428: Yukawa theory to describe the scattering of two very energetic $\psi$'s by a 

429: large angle, corresponding to a large momentum transfer mediated by $\phi$.  

430: For instance, suppose that the $\phi$ is actually a strongly bound state of 

431: two {\it new} fermions $\Psi$ interacting with $\psi$ via a 4-fermion coupling 

432: $\bar{\psi}\psi\,\overline{\Psi}\Psi$.  Then, when the $\psi$'s get recoils 

433: much larger than `$\Lambda_{\rm QCD}$' of this new strong interaction, we must 

434: use the 4-fermion theory with $\Psi$ rather than the Yukawa theory with 

435: $\phi$.  Here, $\phi$ is meant to be the analog of the graviton, and 

436: therefore, within our effective theory, a graviton should not be exchanged 

437: to mediate a large momentum transfer ($\gg \Lg$). 

438: 

439: Now, let us move on and include matter fields.  First, note that for a given

440: value of $\Lg$, some elementary particles in the standard model (SM) are too 

441: short-lived ($\tau \ll \Lg^{-1}$) to be included in SGET.  On the other hand, 

442: some composite particles in the SM live long enough ($\tau \gg \Lg^{-1}$) and 

443: also are too small in size ($\ll \Lg^{-1}$) for a soft graviton to recognize

444: that they are composite.  For example, if $\Lg$ is, say, $10^{-2}\>\ev$, then 

445: the proton, the hydrogen atoms in $1S$ and $2P$ states would be all 

446: {\it elementary} fields (fermion, scalar, and vector, respectively) in SGET.  

447: 

448: Secondly, there are many `hard processes' among those matter particles involving 

449: momentum transfers much larger than $\Lg$.  For example, if $\Lg$ is, say, 

450: $1\>\ev$, then the pair annihilation, $e^+ e^- \to \gamma \gamma$, would be a 

451: hard process.  Since a soft graviton in this case cannot resolve the $t$-channel 

452: electron propagator there, we should shrink it to a point and express the entire 

453: process by a {\it single} local operator.  Also, since soft gravitons in this

454: case cannot pair-produce an electron and a positron, they are completely 

455: unrelated particles from soft gravitons' viewpoint.  Whereas if $\Lg$ is, say,

456: $1\>\gev$, then there are soft gravitons who can see the $t$-channel propagator 

457: in $e^+ e^- \to \gamma\gamma$, and electrons and positrons must be described by 

458: a single field operator.

459: 

460: The general matching procedure for a SGET may be best explained by comparing 

461: it with the construction of a usual effective field 

462: theory in which heavy particles are simply integrated out.  In the derivation 

463: of a usual effective theory, we consider 

464: \textit{one-light-particle-irreducible} (1LPI) diagrams; in a 1LPI diagram, 

465: all external lines represent light particles to be kept in the effective 

466: theory, and the diagram would not split in two if any one of internal 

467: light-particle propagators were cut.  We then obtain effective vertices in the 

468: effective theory by shrinking every heavy propagator to a point.

469: 

470: Similarly, for a SGET, we consider 

471: \textit{one-nearly-on-shell-particle-irreducible} (1NOSPI) diagrams; in a 

472: 1NOSPI diagram, all external lines are nearly on-shell, \ie, its deviation 

473: from the mass shell is less than $\order{\Lg}$, and the diagram would not

474: split in two if any one of internal nearly-on-shell propagators were cut.  We 

475: then obtain effective vertices by shrinking every far-off-shell (\ie~not nearly 

476: on-shell) propagator to a point.  For the technical detail of `shrinking' 

477: or matching procedure, see Ref.~\cite{Sundrum:2003jq}.

478: 

479: Having matched all hard SM processes onto effective operators, 

480: we are now ready to couple to it the soft graviton described by 

481: \eq{puregrav}.  This step is straightforward---we just 

482: use general covariance as a guide, just as we do for the conventional general 

483: relativity.

484: 

485: By construction, SGET respects all fundamental requirements such as 

486: the equivalence principle, Lorentz invariance, and unitarity, as long as we 

487: stay within its applicability we have discussed above \cite{Sundrum:2003jq}.  

488: In particular, unitarity 

489: holds because all propagators that can be on-shell are included in SGET, so it 

490: correctly reproduces the imaginary part of any amplitude.  

491: 

492:  

493: \section{The Composite Gravitino}

494: \label{sec:SusySGET}

495: %

496: Now, let us consider putting a gravitino in the story, with the hope that

497: a gravitino may be more experimentally accessible than gravitons at short 

498: distances so we can learn something about gravity.  The new ingredient in this 

499: section is supersymmetry (SUSY) as a spontaneously broken exact underlying 

500: symmetry, not only in the matter sector {\it but also in the gravity sector.}  

501: As mentioned in Sec.~\ref{sec:intro}, the introduction of SUSY is a necessary 

502: and minimal additional ingredient if we wish to have an alternative probe for 

503: $\Lg$ which is as model-independent as possible, because without SUSY there is 

504: nothing that is necessarily related to gravity except the graviton itself.    

505: 

506: Since the graviton is not a fundamental degree of freedom at

507: short distances, neither is the gravitino.%

508: %

509: \footnote{Of course, there is also a possibility that a gravitino just does not 

510: exist.  Here, we assume that a gravitino exists and its lifetime is long

511: enough ($\gg \Lg^{-1}$) to be in the effective theory.  We will come back to 

512: this caveat in Sec.~\ref{sec:existence}.}

513: %

514: Let $\Lgt$ be the scale above which the gravitino ceases to be an elementary 

515: degree of freedom.  Because supersymmetry is broken, $\Lgt$ does not have to 

516: be equal to $\Lg$.  There is also another scale in the theory, the gravitino 

517: mass $m_{3/2}$.  {\it A priori, these three scales may come in any order.}  

518: SGET assures that there is a consistent framework to describe

519: particles which are much heavier than $\Lg$, so $m_{3/2}$ may be higher or 

520: lower than $\Lg$.  While $\Lgt$ is roughly the `size' of the gravitino, 

521: there is nothing wrong for a composite particle to be heavier than the 

522: inverse of its size, or the compositeness scale.  In fact, heavy quark 

523: effective theory (HQET) \cite{Georgi:1990um}, which describes a single 

524: $B$-meson system, takes advantage of the fact that the $B$-meson's 

525: compositeness scale $\Lambda_{\rm QCD}$ is much less than its mass 

526: $m_B \approx m_b$.        

527: 

528: In the case of HQET, the effective theory breaks down if a gluon delivers

529: a momentum transfer larger than $m_b$ to the $b$ quark.  But in general 

530: effective theories, the breakdown may happen at an energy much lower than 

531: any obvious mass scale in the theory.  For example, consider the effective 

532: field theory of a hydrogen atom in the ground state interacting with soft 

533: photons ($E_\gamma \ll \order{\ev}$).  This effective theory contains an 

534: elementary scalar field (the hydrogen atom in the $1S$ state) and the 

535: electromagnetic field, and it correctly accounts for the Rayleigh scattering, 

536: explaining why the sky is blue.%

537: %

538: \footnote{The reader not familiar with this cute application of

539: effective field theory may like to read Ref.~\cite{Kaplan:1995uv}.}

540: %

541: But this effective theory clearly goes wrong if a photon delivers an energy 

542: of $\order{\ev}$ or higher, where we should take into account the

543: fact that the scalar is actually not elementary.  But this breakdown scale 

544: is much less than the scalar mass, $\order{\gev}$.% 

545: %

546: \footnote{We get a different breakdown scale if we are interested in 

547: capturing a different physics, such as the pair-annihilation of a hydrogen

548: and an anti-hydrogen.}%

549: $^{,}$%

550: %     

551: \footnote{Interestingly, even if we take into account the internal structure,

552: the breakdown scale $\order{\ev}$ is still much smaller than the lightest mass

553: in the theory $m_e \approx .5~\Mev$.  See Ref.~\cite{Luke:1996hj} for an

554: illuminating formalism making this breakdown scale manifest.}

555: 

556: Therefore, {\it a priori} there seems no restriction on possible values

557: for $m_{3/2}$.  (For further discussions shedding different light on this 

558: matter, see Sec.~\ref{sec:existence}.)  Nevertheless, we will show

559: below that $m_{3/2}$ should be bounded from above by $\Lg$, which is a 

560: nontrivial constraint arising from the underlying supersymmetry. 

561: 

562: First, we must be clear about what we mean by `gravitino'.  For instance,

563: say, we have found a new spin-$3/2$ fermion which has no 

564: $SU(3)\times SU(2)\times U(1)$ interactions with the rest of the standard

565: model.  Does it mean we have seen a gravitino?  Not necessarily.  

566: In order for some spin-$3/2$ fermion to be a candidate for a gravitino, at 

567: least it must have---possibly among other things---a coupling to the 

568: supersymmetry current of the matter sector; in other words, it should be 

569: able to convert a matter particle to its superpartner.  Without this feature, 

570: it would be no different from a random spin-$3/2$ resonance.

571: 

572: So, we begin by supposing that we have seen a spin-$3/2$ 

573: fermion $X$ emitted in a process of the type $\widetilde{Y} \to Y + X$, 

574: where $\widetilde{Y}$ is the superpartner of 

575: a particle $Y$.  

576: 

577: For $m_{3/2} \ll \Lg$, it is clearly {\it consistent} to add the gravitino in

578: the pure gravity effective lagrangian \eq{puregrav}, treating it just 

579: like the graviton.  In other words, we can first forget about the graviton

580: and gravitino, construct the nearly-on-shell effective lagrangian for matter,

581: then couple the graviton and the gravitino using general covariance and 

582: local supersymmetry as a guide, where the effects of $m_{3/2}$ can be 

583: systematically included as perturbation. 

584: 

585: For $m_{3/2} \gg \Lg$, we clearly cannot include the gravitino in 

586: \eq{puregrav} together with the soft graviton, because whenever such a heavy 

587: gravitino is produced or exchanged, it is a hard process ($\gg \Lg$) by 

588: definition.  

589: But this simply suggests that we should treat it just as one of heavy matter 

590: fields instead.  The only difference seems that unlike all the other matter 

591: particles, we do not have a fundamental theory for the composite gravitino, 

592: so we cannot calculate the coefficients in SGET lagrangian---that is fine, 

593: we just leave them as parameters.  

594: 

595: However, we have to be careful, because this splitting of the graviton and 

596: gravitino into the soft and hard sectors may be incompatible with the 

597: underlying SUSY, which pairs them.

598: 

599: Let us build a gauge-theory analog of our problem.  First, recall our 

600: {\it global} symmetry structure: the underlying symmetry is the 

601: super-Poincar\'e group, which is spontaneously broken to its subgroup, the 

602: Poincar\'e group.  So, consider a global $SU(2)$ symmetry which spontaneously 

603: breaks down to a $U(1)$ by a triplet scalar $\phi = \phi^a \sigma^a$ getting 

604: a VEV $\avg{\phi^{1,2}}=0$ and $\avg{\phi^3}=v$.  (Here, $\phi$ is treated 

605: just as a spurion.)  The $SU(2)$ is the analog of the underlying supersymmetry, 

606: while the unbroken $U(1)$ is the analog of the unbroken Poincar\'e symmetry.  

607: 

608: Now, at long distances, the Poincar\'e group is gauged by the existence of the 

609: soft graviton which, however, is not a fundamental degree of freedom at 

610: short distances.  So, correspondingly, we gauge the $U(1)$ at long 

611: distances by introducing a soft massless vector field $W^3_\mu$, which we call 

612: `toy soft graviton'.  And just like the graviton, $W^3_\mu$ is not a 

613: propagating degree of freedom at short distances.  Finally, we also need a 

614: `toy gravitino', \ie, a massive vector 

615: $W^+_\mu \equiv (W^1_\mu - i W^2_\mu)/\sqrt{2}$.

616: 

617: Let us assume $\mW \gg \Lg$ which is the case of our interest. 

618: We want to write down `toy SGET' for the toy gravitino.  The only property 

619: of $W^+_\mu$ which possibly makes it different from other heavy 

620: particles is that it is the $SU(2)$-partner of the toy graviton $W^3_\mu$.

621: So, the question is whether there is any constraint on the structure of

622: the toy SGET from the underlying $SU(2)$, or the toy SUSY.

623: 

624: Let us forget $\Lg$ for a moment, and recall how a spontaneously broken 

625: symmetry leaves its trace in low-energy physics.  

626: To be concrete, consider couplings of $W^+_\mu$ and $W^3_\mu$ to a heavy 

627: Dirac fermion doublet $\psi$.  

628: ($\psi$ is of course the analog of the pair of a SM particle 

629: and its superpartner.)  If we limit to

630: only renormalizable operators, all three $W^a_\mu$ must couple to the three 

631: currents $J^{a\mu} = \bar{\psi}\sigma^a\gamma^\mu\psi$ with a 

632: single common coupling constant $g$.  This equality is a consequence of the 

633: underlying $SU(2)$, even though it is broken.  

634: 

635: However, once we take into account higher-dimensional operators, the coupling 

636: of $W^{1,2}_\mu$ to $J^{1,2}_\mu$ does not have to be equal to that of 

637: $W^3_\mu$ to $J^3_\mu$, because there are higher dimensional operators that 

638: reduce to these couplings after picking up the VEV.  Among such, the one with

639: the lowest dimension is the dimension-5 operator 

640: $\bar{\psi} \phi \,\Sla{D} \psi$.

641: We could go on and 

642: analyze this operator, but it turns out that we can learn the same lesson

643: with much less arithmetic from the following dimension-6 operator:

644: \beq

645:    {\cal L}_6 = -\frac{16\pi^2 c}{M^2} \bar{\psi} \phi \,i\Sla{D} \phi\psi ~,

646: \eql{dim-6op}

647: \eeq

648: where we take $c \sim 1$ so that $M$ corresponds to the scale obtained via

649: `naive dimensional analysis' (NDA), \ie,

650: the scale at which this operator would lead to strong coupling if the theory 

651: is not replaced with a more fundamental theory by then \cite{Manohar:1983md}.  

652: After substituting the VEV for $\phi$ and canonically normalizing the fields, 

653: we find that the coupling of $W^3_\mu$ stays equal to $g$ as expected from 

654: the unbroken $U(1)$ gauge invariance, but the coupling of $W^+_\mu$ does get 

655: modified as

656: \beq

657:    g \too g_+ &=& \frac{1+a}{1-a} \, g  ~,

658: \eeq

659: where 

660: \beq

661:    a \equiv \frac{16\pi^2 v^2 c}{M^2} \sim \left( \frac{4\pi v}{M} \right)^2 

662:    ~.

663: \eeq

664: Therefore, the equality of the $W^3_\mu$ and $W^+_\mu$ couplings no longer

665: holds.  Especially, if $v$ is $\order{M/4\pi}$, then $g_+/g$ could be 

666: anywhere between zero and infinity, and there would be no remnants of the 

667: underlying $SU(2)$ symmetry.   

668: 

669: This lesson can be generalized. In general $g_+$ differs from $g$ as

670: \beq

671:    g_+ = \alpha g ~,

672: \eql{g-alpha-rel}

673: \eeq

674: where the factor $\alpha$ includes contributions from all the 

675: operators that can mix with $W_\mu^a J^{a\mu}$.  The relation $\alpha \simeq 1$ 

676: holds as long as $v \ll M/4\pi$, but for $v \sim M/4\pi$,

677: all those operators would contribute to $\alpha$ equally in magnitude, and 

678: consequently $\alpha$ could be anywhere between zero and infinity.   

679: 

680: Now, let us go back to the case of our interest and take $\Lg$ into account.

681: Let us write the doublet $\psi$ as

682: \beq

683:    \psi = \colmvec{ \chit \cr \chi } ~,

684: \eeq

685: and, for definiteness, take $\chit$ to be heavier than $\chi$ 

686: with the mass difference larger than $\mW$ so that $\chit$ can decay into 

687: $\chi$ and $W^+$.  Clearly, this is the analog of a sparticle decaying

688: into its SM partner and a gravitino.  

689: 

690: Once we have seen a toy gravitino produced via this decay, $g_+$ must be 

691: nonzero.  In the rest frame of the decaying $\chit$, this decay is caused by

692: the operator

693: \beq

694:   {\cal H}_{\rm int} 

695:      \supset g_+ \, W^+_{-{\bf p}} \,\chi_{\bf p} \,\chit_{\bf 0} 

696:   ~,

697: \eeq

698: where the irrelevant indices, bars and daggers are suppressed, while the

699: important quantity here is $\abs{\bf p} = \sqrt{E_\chi^2 - m_\chi^2}$ where 

700: $E_\chi$ is the energy of the outgoing $\chi$ given by

701: \beq

702:   E_\chi = \frac{m_\chit^2 + m_\chi^2 - \mW^2}{2m_\chit}  ~.

703: \eql{E_chi}  

704: \eeq

705: 

706: %

707: \begin{figure}

708: \includegraphics{chichi-fig.eps}

709: \caption{The $\chit_{\bf p} \chit_{\bf 0} \to \chit_{\bf 0} \chit_{\bf p}$ 

710: scattering via the $t$-channel $W^3_\mu$ exchange.}

711: \label{fig:chichi}

712: \end{figure}

713: %

714: 

715: Now, if $g$ is also nonzero, there would also be a term  

716: \beq

717:   {\cal H}_{\rm int} 

718:      \supset g \, W^3_{-{\bf p}} \,\chit_{\bf p} \,\chit_{\bf 0}  ~,

719: \eql{bad-op}

720: \eeq

721: with the {\it same} ${\bf p}$.  The problem is that, at the second order in 

722: perturbation theory, this operator could cause the process 

723: $\chit_{\bf p} \,\chit_{\bf 0} \to \chit_{\bf 0} \,\chit_{\bf p}$ 

724: via the $t$-channel $W^3_\mu$ exchange (FIG.~\ref{fig:chichi}).  Note that

725: the momentum transfer $Q^2$ mediated by the $W^3_\mu$ is given 

726: by

727: \beq

728:   Q^2 &=& -\left( \sqrt{{\bf p}^2 + m_\chit^2} - m_\chit \right)^2 

729:            + {\bf p}^2  \nn\\

730:       &=& 2 m_\chit^2  \left( \sqrt{1+\frac{{\bf p}^2}{m_\chit^2}} 

731:            - 1 \right) ~.

732: \eeq

733: From $\abs{\bf p} = \sqrt{E_\chi^2 - m_\chi^2}$ and \eq{E_chi}, we see that

734: for generic $m_\chi$ and $\mW$, we have $\abs{Q} \sim m_\chit > \mW \gg \Lg$,

735: which is a {\it hard} momentum transfer.%

736: %

737: \footnote{The exception occurs in `highly degenerate' cases:~(a) either 

738: one of $\chi$ and $W^+$ is much heavier than the other and almost degenerate 

739: with $\chit$ so that $\abs{\bf p} \ll \Lg$ or (b) $m_\chi$ and $\mW$ are of 

740: the same order but they add up to nearly $m_\chit$ so that 

741: $\abs{\bf p} \ll \Lg$.  Although these case are logically possible, it looks 

742: too coincidental, so we will not pursue this caveat any further.}

743: %   

744: However, as discussed in 

745: Sec.~\ref{sec:SGET}, a graviton cannot be exchanged to mediate such a 

746: large momentum transfer within SGET.  Therefore, the operator \eq{bad-op}

747: should not be present in the effective theory.

748: 

749: Therefore, to decouple the operator \eq{bad-op}, we must take the limit 

750: $g \to 0$ while keeping $g_+$ fixed to a finite value.  Then the relation

751: \eq{g-alpha-rel} requires $\alpha \to \infty$, which, however, is possible 

752: only if $v \sim M/4\pi$, as noted before.  In this limit, all 

753: the higher-dimensional operators that can contribute to $g_+$ do contribute 

754: equally in magnitude, while all the other interactions have the full 

755: NDA strength.  Furthermore, having taken this limit, we have decoupled the 

756: {\it soft} $W^3_\mu$ as well, so we have to couple it back to the theory.  

757: This can be easily done by using the 

758: $U(1)$ invariance, but now the coupling of this $U(1)$---let us call it 

759: $g_{\rm soft}$---is completely arbitrary, with no relation to $g_+$!    

760: 

761: Therefore, although we cannot perform any quantitatively reliable analysis 

762: beyond estimates% 

763: %

764: \footnote{We have also neglected the effects of running.}

765: %

766: due to $v \sim M/4\pi$, this is good enough to give us the following 

767: qualitative understanding of what $W^+_\mu$ is like.  First, its coupling to the

768: $SU(2)$ current, $g_+$, is not related at all to the coupling of the soft 

769: $W^3_\mu$ to the $U(1)$ current, $g_{\rm soft}$.  Second, it has all kinds of

770: additional interactions, all with the full NDA strength.  Because of these

771: two features, $W^+_\mu$ should be viewed just as a random spin-1 resonance, 

772: rather than the `$SU(2)$-partner' of $W^3_\mu$.   

773: 

774: Recalling the dictionary of our analogy, translating this gauge-theory lesson 

775: back to gravity is straightforward.  (The only slight mismatch in the 

776: dictionary, which is not at all essential for us, appears in the $4\pi$ counting 

777: for broken SUSY, where the relation $v \sim M/4\pi$ should be translated as 

778: $F \sim M^2/4\pi$ where $F$ is the decay constant of the goldstino, or the 

779: square of the SUSY breaking scale \cite{Luty:1997fk}.)  Therefore, we have 

780: found  

781: %  

782: \begin{itemize}

783:  \item{If $m_{3/2} \ll \Lg$, it is {\it consistent} for the gravitino to be 

784:   just `canonical', with all the properties we expect from the standard 

785:   supergravity, except for the fact that the gravitino---like the 

786:   graviton---is not an elementary degree of freedom at short distances.  

787:   In other words, as long as we avoid processes where a gravitino receives or 

788:   mediates a large momentum transfer, the gravitino can behave normally.}

789:  \item{If $m_{3/2} \gg \Lg$, this `gravitino' is not really a gravitino,

790:   because the coupling of this `gravitino' to a SM particle and its 

791:   superpartner can have any value, with no relation to the `canonical' 

792:   strength, and we also expect this `gravitino' to have a whole series of other 

793:   couplings, all equally important with the full NDA strength.  In short, it 

794:   behaves just like a random spin-$3/2$ resonance with no relation to the 

795:   gravity sector.} 

796: \end{itemize} 

797: Hereafter, to distinguish these cases, we will use the term 

798: {\it gravitino} only to refer to the first case, while we will call the 

799: second case {\it pseudo-gravitino.}  

800: 

801: We postpone the issue of experimentally distinguishing a gravitino from a 

802: pseudo-gravitino until Sec.~\ref{sec:detection}.  At this point, let us just 

803: assume that the distinction can be made.  Then, we have found the 

804: {\it model-independent} relation between the gravitino mass and the composite

805: gravity scale:

806: \beq

807:   m_{3/2} \lsim \Lg ~.

808: \eql{theprediction}

809: \eeq

810: By definition, gravity is described by GR at distances longer than $\Lg^{-1}$,

811: because GR is the only consistent theory once we have a graviton coupled to

812: matter described by a local relativistic quantum field theory 

813: \cite{Weinberg:1965rz}.  (Note that we could not have said this if we had not 

814: restricted $\Lg$ below $\tev$ which assures the matter sector is `normal'.)  

815: Therefore, the relation \eq{theprediction} 

816: means that {\it the existence of a gravitino guarantees that GR is correct 

817: at least down to its Compton wavelength!}  Hence, this is a short-distance 

818: test of GR, which in turn places a lower bound on $\Lg$.      

819: On the other hand, the relation \eq{theprediction} implies that if we find 

820: composite gravity first at some $\Lg$, then we will not discover a gravitino

821: above the scale $\Lg$---at best we may just see a pseudo-gravitino which is 

822: nothing but a random spin-3/2 state.    

823: 

824: 

825: \section{Discussions}

826: \label{sec:discuss}

827: 

828: \subsection{Should a Gravitino Exist?}

829: \label{sec:existence}

830: %

831: The quick answer is, we don't know.  There is no strong argument indicating 

832: whether it should or shouldn't.  We will present below several arguments, 

833: not to answer this question but to shed different light and gain more 

834: insights on the result of Sec.~\ref{sec:SusySGET}.   

835: 

836: Imagine a huge hierarchy between the SUSY breaking scale $\sqrt{F}$ and $\Lg$, 

837: as $\sqrt{F} \gg \Lg$.  Above $\Lg$, the gravity sector is described by some 

838: exotic degrees of freedom---which may not even be field-theoretic---with no 

839: gravitons.  Here, there is no point of asking what the superpartner of the 

840: {\it graviton} is, because the graviton is not even in the theory.  

841: When we go below $\Lg$, the graviton emerges, but we do not expect that a 

842: gravitino appears there, because from the usual effective-field-theoretic 

843: viewpoint, the dynamics at $\Lg$ that generates the graviton should not 

844: `know' about SUSY which is broken \textit{way} above $\Lg$.  

845: 

846: This argument is too naive, however.  As we will argue below, not only is it 

847: possible that a 

848: pseudo-gravitino may exist, but also even an honest gravitino with all the 

849: (approximately)

850: canonical properties may exist!  Consider a supersymmetric $SU(3)$ gauge 

851: theory with two flavors with no superpotentials, and suppose that SUSY is 

852: broken with the soft masses much larger than $\Lambda_{\rm QCD}$.  For 

853: simplicity and definiteness, also assume that the squark masses are all 

854: degenerate, respecting the flavor symmetry, and that the gluino is much 

855: heavier than the squarks.  This theory possesses an $R$ parity under which 

856: all the quarks and gluon are even while all the squarks and gluino are odd.  

857: Hence, the squarks are stable, and there are stable fermionic meson-like bound 

858: states (`mesinos') with one quark and one anti-squark.%

859: %

860: \footnote{We are assuming that these mesinos are the lightest among the 

861: hadrons containing superparticles.}

862: %

863: 

864: So, apparently, the mesons have superpartners, the mesinos.  But look at other 

865: particles; for example, there is no `sproton' or `sneutron', because they 

866: would decay too quickly to form a bound state.  In fact, most particles lack 

867: their superpartner, so the interactions between the meson-mesino sector and 

868: the rest are completely non-supersymmetric.  Therefore, if these 

869: non-supersymmetric couplings are significant, there is no sense in which the

870: mesinos are the superpartners of the mesons, except for the quantum numbers.  

871: In other words, the mesinos in this case are just analogous to our 

872: pseudo-gravitino.

873: 

874: However, there is also a logical possibility that the couplings between the 

875: meson-mesino sector and the rest are sufficiently small for some reason. 

876: Then, it is at least {\it consistent} for the mesinos to retain the properties 

877: expected from supersymmetry.%

878: %

879: \footnote{Note that this is exactly what is happening in typical weak-scale 

880: SUSY models in which the visible-sector interactions at the weak scale are 

881: taken to be (approximately) supersymmetric, even though the actual SUSY 

882: breaking scale is often as high as $10^{11}\,\gev$. This is consistent because 

883: the interaction that transmits SUSY breaking to the visible sector is assumed 

884: to be sufficiently feeble.}

885: %

886: A similar situation could happen to a gravitino.  For example, if $\Lg$ is, say,

887: $10^{-2}~\ev$, then it could be perfectly consistent for a gravitino with, 

888: say, $m_{3/2} = 10^{-3}~\ev$ to carry all the (approximately) canonical 

889: couplings we expect from supergravity---as long as the gravitino does not 

890: receive or mediate a momentum transfer larger than $\Lg$---even though 

891: $\sqrt{F}$ here would be $\sim \tev$ which is way above $\Lg$.

892: To sum up, from the standard effective-field-theoretic view, there seems no 

893: preference among `nothing', `a pseudo-gravitino', and `a (real) gravitino'.     

894: 

895: To gain more insight, let us consider the limit in the opposite order.  This 

896: time we start with a finite $\Lg$ but no SUSY breaking ($F=0$).%

897: %

898: \footnote{An extreme but trivial limit of this case is to take 

899: $\Lg \gsim \Mpl$, \ie, the limit of an elementary graviton.  Note that for any 

900: $\sqrt{F} \lsim \Mpl$, the gravitino is a normal gravitino, and the inequality 

901: \eq{theprediction} is trivially satisfied since 

902: $m_{3/2} \sim F/\Mpl \lsim \Mpl \lsim \Lg$.}

903: %

904: So we start 

905: with a degenerate pair of massless graviton and gravitino.  This gravitino 

906: is of course exactly what we expect from supergravity, as long as we avoid 

907: momentum transfers larger than $\Lg$.  As we raise $F$, the gravitino mass 

908: goes up according to the usual relation $m_{3/2} \sim  F/\Mpl$, as long as 

909: $m_{3/2}  \ll \Lg$.  If we keep raising $F$, $m_{3/2}$ eventually hits

910: $\Lg$, beyond which the gravitino may start looking strange.  (The result of

911: Sec.~\ref{sec:SusySGET} says it {\it will} start looking strange, but here

912: let us pretend that we did not know Sec.~\ref{sec:SusySGET}.)  Then, in 

913: particular we no longer know how $m_{3/2}$ should vary as a function of $F$. 

914: (We will come back to this issue in detail in Sec.~\ref{sec:relation}.)  

915: Here, let us suppose that it still keeps going up, although not 

916: necessarily obeying the usual linear relation.  Will this `gravitino' 

917: eventually disappear?  Note that it will disappear from SGET if its 

918: lifetime becomes shorter than $\Lg^{-1}$.  Naively, we expect that the 

919: lifetime should be quite long because the coupling $1/\Mpl$ is extremely weak, 

920: so it would stay in the effective theory even if $m_{3/2}$ is as high as the 

921: weak scale.  But this `gravitino' may have unusual interactions, and there 

922: are probably many new states around $E \sim \Lg$ into which the `gravitino' 

923: could decay.  So the lifetime may or may not be quick enough for the 

924: `gravitino' to disappear from SGET.  We need the underlying theory to see 

925: which way it goes.   

926: 

927: Finally, it is also conceivable that $m_{3/2}$ `saturates' at $\Lg$ as we raise 

928: $F$.  We would expect this if there is an exotic state at $E\sim\Lg$ 

929: which can mix with the gravitino. Then, by the `no-level-crossing' theorem, 

930: $m_{3/2}$ cannot go up any further, and the `gravitino' becomes a mixture of 

931: the original gravitino and this exotic state.  Therefore, in this

932: case, we expect a pseudo-gravitino with $m_{3/2} \sim \Lg$.

933: 

934: To summarize, qualitative arguments seems completely inconclusive about 

935: the nature and fate of a gravitino.  The result of 

936: Sec.~\ref{sec:SusySGET} is therefore quite nontrivial. 

937: 

938: 

939: \subsection{Relation of $m_{3/2}$ to SUSY Breaking Scale}

940: \label{sec:relation}

941: %

942: Here, we comment on the validity of the famous relation between 

943: the gravitino mass and the SUSY breaking scale:

944: \beq

945:    m_{3/2} = \frac{F}{\sqrt{3} \Mpl}  ~.

946: \eql{famous-rel}

947: \eeq

948: In the pseudo-gravitino case ($m_{3/2} \gg \Lg$), this usual relation

949: has no reason to be true.  Clearly, we cannot use the supergravity formalism 

950: to derive it, because supergravity contains general relativity which is not 

951: applicable for $E \gg \Lg$ in our scenario.  But more fundamentally, recall 

952: that this relation is just a consequence of the equivalence between the 

953: goldstino and the longitudinal component of the gravitino at high energies 

954: ($E_{3/2} \gg m_{3/2}$).  Usually, we derive the relation by demanding that 

955: the amplitude of exchanging a gravitino between two supersymmetry currents 

956: be equal to that of exchanging a goldstino, in the global SUSY limit 

957: ($\Mpl \to \infty$) for $E_{3/2} \gg m_{3/2}$.  However, in the 

958: pseudo-gravitino case, it has a different coupling to the supersymmetry 

959: current as well as a host of additional interactions.  Hence, the formula

960: \eq{famous-rel} does not hold for a pseudo-gravitino.  In other words,

961: since the pseudo-gravitino does not eat the goldstino by exactly the right

962: amount, the SUSY currents must exchange something else to match the 

963: goldstino-exchange amplitude.  But this `something else' must be among the 

964: new exotic states in the full theory of gravity, which we have no idea about.  

965: (If we had the underlying theory, we could subtract the exotic contribution 

966: from the amplitude and figure out how the formula \eq{famous-rel} should get 

967: modified.)

968: 

969: On the other hand, for $m_{3/2} \ll \Lg$, we can apply the derivation 

970: for $m_{3/2} \ll E_{3/2} \ll \Lg$, and obtain the usual relation 

971: \eq{famous-rel}, assuming that the gravitino has the standard $1/\Mpl$ 

972: coupling to the SUSY current, which is at least a consistent thing to do as 

973: we discussed in Sec.~\ref{sec:SusySGET}. 

974: 

975: 

976: \subsection{Theoretical Tests}

977: \label{sec:theo-tests}

978: %

979: It is certainly desirable to confirm the result of Sec.~\ref{sec:SusySGET}

980: by a theoretical argument that has a firmer foundation.  Recall the 

981: concrete example of composite gauge bosons mentioned in Sec.~\ref{sec:intro}:

982: the $SU(N)$ supersymmetric QCD with $F$ flavors, where $N+1<F<3N/2$.  

983: Below the $\Lambda_{\rm QCD}$ of the $SU(N)$, this theory is described in 

984: terms of an IR-free $SU(F-N)$ gauge theory whose gauge bosons are composites 

985: of the original degrees of freedom \cite{Seiberg:1994pq}.  

986: 

987: Now let us deform the theory such that the low-energy gauge group $SU(F-N)$ 

988: gets spontaneously broken down to $SU(M)$ where $M<F-N$.  If we apply the

989: argument of Sec.~\ref{sec:SusySGET} to this theory, we predict that

990: the massive $W$ bosons, with all the `normal' couplings retained, cannot be 

991: heavier than $\Lambda_{\rm QCD}$.  $W$ bosons heavier than $\Lambda_{\rm QCD}$ 

992: may exist but they should behave like random spin-1 resonances, rather than 

993: as the `$SU(F-N)$-partners' of the $SU(M)$ gauge bosons.  While it sounds 

994: plausible, the currently available theoretical wisdoms are not powerful 

995: enough to definitively confirm the statement.

996: 

997: This SUSY QCD example also illustrates how extremely nontrivial it is 

998: to have a composite graviton coupled to elementary matter particles.  

999: In the case of the SUSY QCD model, this corresponds to the composite $SU(F-N)$ 

1000: gauge bosons coupled to elementary quarks that are point-like even far above

1001: $\Lambda_{\rm QCD}$!  This is clearly a very difficult, if possible, thing

1002: to do.  In the AdS composite graviton model of Ref.~\cite{Gherghetta:2005se}, 

1003: the graviton wavefunction is highly peaked toward the IR brane, but there is 

1004: an exponentially suppressed tail overlapping the UV brane where the SM fields 

1005: live, which can be thought of as an explanation for the weakness of gravity.  

1006: Adding supersymmetry to their setup to study the gravitino properties is

1007: saved for future work.       

1008: 

1009: 

1010: \section{Precision Gravitino Study and Probing $\Lg$ in Colliders}

1011: \label{sec:detection}

1012: %

1013: Clearly, the most important quantity in any composite graviton scenario is 

1014: the scale $\Lg$.  As we mentioned already in Sec.~\ref{sec:SusySGET}, 

1015: in order to probe the scale $\Lg$, it is crucial to experimentally 

1016: distinguish a gravitino from a pseudo-gravitino.  

1017: 

1018: Unfortunately, if the results of such `precision gravitino study' turn out 

1019: that what we have seen is actually a pseudo-gravitino, this will not be a 

1020: sufficient evidence that gravity is modified at short distances.  For 

1021: example, a pseudo-gravitino is also present in a scenario where supersymmetry 

1022: is not a fundamental symmetry at high energies but merely an (approximate) 

1023: {\it accidental} global symmetry of the matter sector at low energies

1024: \cite{Goh:2003yr}.  In this scenario, the gravity sector is just the 

1025: conventional GR (with no supersymmetry).  Therefore, for a pseudo-gravitino, 

1026: we need the underlying theory to derive more specific predictions to be 

1027: tested. 

1028: 

1029: On the other hand, if we can convince ourselves that it is not a 

1030: pseudo-gravitino, then we can put a {\it model-independent} lower bound 

1031: on $\Lg$, as $\Lg \gsim m_{3/2}$!  Interestingly, as we will see shortly,

1032: in precisely the regime 

1033: that the direct gravity measurement between test masses is impossible, the 

1034: measurement of $m_{3/2}$ becomes possible, so the precision gravitino study 

1035: can potentially exclude composite graviton scenarios dramatically at very

1036: short distances.   

1037: 

1038: Since it is impossible to see a gravitino $\psi_{3/2}$ directly, the only hope 

1039: to learn 

1040: something about it lies in the case where both $\Xt$ and $X$ can be precisely

1041: studied in the decay $\Xt \to X + \psi_{3/2}$.  This means that the decay 

1042: must be sufficiently slow and that $\Xt$ and $X$ both must be visible.  This 

1043: will indeed be realized if the $\Xt$ is the next-to-lightest supersymmetric 

1044: particle (NLSP) (the lightest (LSP) being the gravitino) and is electrically 

1045: charged and/or strongly-interacting.  In such a case, due to the very weak 

1046: coupling of $\Xt$ to the gravitino, there will be a long, highly visible 

1047: track of the NLSP inside a collider detector before it decays 

1048: \cite{Drees:1990yw}, unless $\psi_{3/2}$ is too light.  It is even possible 

1049: that the NLSP stops in the detector if it is strongly interacting or produced 

1050: sufficiently slow.  In such circumstances, the momenta and energies of the 

1051: NLSP and its SM partner as well as the NLSP lifetime should be measurable, 

1052: which in turn allows us to deduce the mass and the coupling of the gravitino 

1053: to see whether it is a pseudo-gravitino or not. 

1054: 

1055: This `gravitino LSP with charged NLSP' scenario has already been a 

1056: great interest in SUSY phenomenology, especially in the context of 

1057: gauge-mediated SUSY models where the gravitino is the LSP and $\Xt$ is often 

1058: a scalar tau lepton \cite{Dimopoulos:1996vz, Buchmuller:2004rq, 

1059: Hamaguchi:2004df}.  Note that once $X$ and $\Xt$ have been observed, the 

1060: gravitino mass can be simply determined from rewriting \eq{E_chi} :

1061: \beq

1062:    m_{3/2} = \left( m_X^2 + m_\Xt^2 - 2 m_\Xt E_X \right)^{1/2} ~,

1063: \eql{kinematics}

1064: \eeq

1065: where $E_X$ is the energy of the $X$ measured in the rest frame of the $\Xt$.

1066: If $\Xt$ stops inside a detector, $E_X$ can be directly measured.  

1067: Even if it does not, since both the $X$ and $\Xt$ are highly visible in the 

1068: detector, the measurement of their energies and the relative angle (the `kink'

1069: in the track) can determine $E_X$.    

1070: 

1071: On the other hand, the measurement of the $\Xt$ lifetime gives us the 

1072: gravitino's coupling.  If what we are seeing is not a pseudo-gravitino but

1073: is a real one, then the coupling should go as $1/\Mpl$ times the polarization

1074: factor $E_{3/2}/m_{3/2}$ for the helicity-$\pm 1/2$ components, so the rate 

1075: is given by

1076: \beq

1077: \eql {NLSP-rate}

1078:   \Gamma_\Xt

1079:         &=&    \frac{m_\Xt^5}{48\pi \Mpl^2 \, m_{3/2}^2}  \nn\\

1080:      &\approx& (20~\micr)^{-1} 

1081:                 \left( \frac{\ev}{m_{3/2}} \right)^2

1082:                 \left( \frac{m_\Xt}{100\,\gev} \right)^5  \\

1083:      &\approx& (20~{\rm hours})^{-1} 

1084:                 \left( \frac{\gev}{m_{3/2}} \right)^2

1085:                 \left( \frac{m_\Xt}{100\,\gev} \right)^5  \nn

1086: \eeq

1087: where we have dropped $m_X$ and $m_{3/2}$ for simplicity. (The 

1088: helicity-$\pm 3/2$ components have no $E_{3/2}/m_{3/2}$ enhancement and thus 

1089: have been neglected.) 

1090: The consistency of $m_{3/2}$ determined from this formula with the value 

1091: extracted from pure kinematics \eq{kinematics} will be an almost convincing 

1092: evidence that the gravitino is not a pseudo, because it would be such a 

1093: coincidence if the pseudo-gravitino coupling, which could be any size, just 

1094: happened to be $1/\Mpl$.%

1095: %

1096: \footnote{Note that this agreement between the two measurements

1097: of $m_{3/2}$ is equivalent to checking if the gravitino has really eaten the

1098: goldstino as it should if it is not a pseudo.}

1099: %

1100: Ref.~\cite{Buchmuller:2004rq} proposes to go even further, to test the 

1101: gravitino's {\it spin} by using the angular distribution in the 3-body decay 

1102: $\tilde{\tau} \to \tau + \gamma + \psi_{3/2}$.                           

1103: 

1104: Now, it is probably extremely hard to directly measure the gravitational force 

1105: between test masses for distances smaller than the micron scale which 

1106: would correspond to $\Lg \lsim 10^{-1}\,\ev$.  Let us see whether the 

1107: precision gravitino study can be used to place a bound on $\Lg$ beyond this 

1108: limitation.  Taking $m_\Xt = 100~\gev$, the rate \eq{NLSP-rate} tells us that

1109: for $m_{3/2} = 10^{-1}\,\ev$, the NLSP will decay within $\order{1}\,\micr$, 

1110: since the relativistic $\gamma$ factor for the NLSP cannot be larger than

1111: $\order{10}$ in a $\tev$-scale collider.  This is unfortunately too short 

1112: to be seen.  Demanding that the NLSP must fly at least a few 

1113: $100\,\micr$ to be clearly observed by a micro vertex detector, we need 

1114: $m_{3/2}$ to be at least a few $\ev$.  However, for such low values for 

1115: $m_{3/2}$, the formula \eq{kinematics} requires $m_X$, $m_\Xt$ and $E_X$ to be 

1116: measured with unrealistically high precision.  The problem is, to determine 

1117: a small $m_{3/2}$ from \eq{kinematics}, we have to nearly cancel two large

1118: terms and take the square-root.  Therefore, the lowest possible value for 

1119: $\Lg$ that can be probed is actually limited by the accuracy in measuring 

1120: these parameters rather than the minimal NLSP flight length that a detector 

1121: can resolve.  For example, if we are anticipating 

1122: $m_{3/2}$ of order $1~\gev$ and if we are content with determining 

1123: $m_{3/2}$ only up to a factor of a few, then for $m_\Xt \approx 100~\gev$ 

1124: (neglecting $m_X$ for simplicity), we would need to measure $m_\Xt$ and $E_X$ 

1125: with the accuracy of $\pm 10~\Mev$.  Therefore, measuring $m_{3/2}$ of 

1126: $\order{1}\,\gev$ event-by-event is unrealistic, so it must be done 

1127: statistically.  Taking the uncertainty in the individual $E_X$ 

1128: measurement to be $\order{1}\,\gev$, we need to observe $\order{10^4}$ NLSP 

1129: decays to have enough statistics for $m_\Xt \approx 100~\gev$ and 

1130: $m_{3/2} \sim \gev$.  

1131: 

1132: Also, note that for $m_{3/2} \sim \gev$, the $\Xt$ lifetime is about 

1133: a few hours to a week, so the NLSPs must be collected and stored to

1134: do the measurement.  Such a possibility for $\Xt = \tilde{\tau}$ has been 

1135: extensively studied in Refs.~\cite{Hamaguchi:2004df}, and the bottom line 

1136: is that collecting $\order{10^4}$ or even $\order{10^5}$ NLSPs and observing 

1137: their decays should be possible in the LHC and/or the ILC, although the

1138: prospect depends on other SUSY parameters.

1139: 

1140: Those analyses also conclude that we may be able to go up to $m_{3/2}$ 

1141: of $\order{100}\,\gev$.  Therefore, it is not too optimistic to expect that

1142: precision gravitino study may be able to probe the scale $\Lg$ between 

1143: $\gev$ and $100~\gev$.  While this is still quite challenging 

1144: (and we also have to be lucky with the SUSY spectrum), note that this is a 

1145: regime where direct measurement of gravitational forces is absolutely 

1146: impossible, so precision gravitino study is the only available probe for 

1147: composite gravity.   

1148: 

1149: 

1150: \section{Conclusions}

1151: \label{sec:conc}

1152: %

1153: In this paper, we have considered `composite gravity', namely, the 

1154: possibility that the graviton is not an elementary propagating degree of 

1155: freedom at distances shorter than $\Lg^{-1}$.  We pointed out that such

1156: a scenario is not necessarily forbidden by the Weinberg-Witten theorem.

1157: Another important assumption we made is that the matter sector is completely 

1158: described by a local quantum field theory, which is true for $\Lg$ between 

1159: the current experimental limit $\sim 10^{-3}\,\ev$, and $\sim \tev$ or whatever 

1160: cutoff for the matter sector.  To perform a model-independent, 

1161: effective-field-theoretic analysis, it is necessary to reconcile 

1162: `elementary matter with a high cutoff' and `composite gravity

1163: with a low cutoff', and for this purpose we have utilized soft graviton 

1164: effective theory (SGET) by Sundrum.

1165: 

1166: In general, the only way to place a lower limit on the scale $\Lg$ is by 

1167: a null result in experiments seeking a deviation from the standard $1/r^2$ 

1168: law between macroscopic test masses.  This method becomes increasingly 

1169: difficult as the distance gets reduced.  Therefore, it is desirable to have 

1170: an alternative probe.  The problem is, however, that in general there is 

1171: nothing related to gravity except the graviton itself, so there is no other 

1172: way to probe $\Lg$ without using gravity.

1173: 

1174: However, we noted that if there is an underlying supersymmetry, it may

1175: lead to the existence of a gravitino, which is related to gravity but easier

1176: to observe than the graviton.  Applying the SGET framework to the gravitino,

1177: we have shown the relation, $\Lg \gsim m_{3/2}$, \ie, the {\it graviton} 

1178: remains an elementary degree of freedom at least down to the {\it gravitino's} 

1179: Compton wavelength.  In other words, we can use a gravitino to test general

1180: relativity at short distances---once we see a gravitino, we know that GR is

1181: correct at least up to $m_{3/2}$, which in turn places a lower bound on $\Lg$!  

1182: This can have a significant impact on the possibility of composite graviton as

1183: a solution to the cosmological constant problem.  For example, if we find 

1184: $m_{3/2}$ to be, say, $1\>\gev$, the door will be completely shut. 

1185: 

1186: On the other hand, if we first find gravity compositeness and measure $\Lg$, 

1187: then our inequality says that we will not discover a gravitino above the scale 

1188: $\Lg$---at best we may just see some random spin-$3/2$ fermion with completely 

1189: random couplings, nothing to do with gravity.  

1190:     

1191: To utilize this inequality to place a limit on $\Lg$, it is crucial to

1192: experimentally convince ourselves that what we are looking at is really

1193: a gravitino, rather than a random spin-$3/2$ fermion.  In the future 

1194: colliders such as the LHC and ILC, the prospect of being able to do so 

1195: seems quite bright for the range $\gev \lsim  m_{3/2} \lsim 100~\gev$,

1196: corresponding to probing $\Lg$ in the range between 

1197: $10^{-14}~{\rm cm}$ and $10^{-16}~{\rm cm}$.  Therefore, precision 

1198: gravitino study can indeed be an alternative model-independent probe for 

1199: $\Lg$ or a test of general relativity, in a regime where direct measurement 

1200: of gravitational force is absolutely impossible.     

1201: 

1202: 

1203: 

1204: \begin{acknowledgments}

1205: I thank Zackaria Chacko, Markus Luty, Yasunori Nomura, Matt Schwartz, 

1206: Raman Sundrum, and Mithat \"Unsal for discussions and comments on the 

1207: manuscript.  I also thank Spencer Chang, Hitoshi Murayama, and Michael Peskin 

1208: for conversations.  In addition, I thank the Aspen Center for Physics where

1209: a portion of this work was conducted.  This work was supported by DOE grant 

1210: DE-FG02-91ER40676.   

1211: \end{acknowledgments}

1212: 

1213: 

1214: \begin{thebibliography}{99}

1215: 

1216: \bibitem{Hoyle:2004cw}

1217: C.~D.~Hoyle, D.~J.~Kapner, B.~R.~Heckel, E.~G.~Adelberger, J.~H.~Gundlach, U.~Schmidt and H.~E.~Swanson,

1218: %``Sub-millimeter tests of the gravitational inverse-square law,''

1219: Phys.\ Rev.\ D {\bf 70}, 042004 (2004)

1220: [arXiv:hep-ph/0405262].

1221: %%CITATION = HEP-PH 0405262;%%}

1222: 

1223: 

1224: \bibitem{Long:2003dx}

1225: J.~C.~Long, H.~W.~Chan, A.~B.~Churnside, E.~A.~Gulbis, M.~C.~M.~Varney and J.~C.~Price,

1226: %``Upper limits to submillimeter-range forces from extra space-time

1227: %dimensions,''

1228: Nature {\bf 421}, 922 (2003).

1229: %%CITATION = NATUA,421,922;%%

1230: 

1231: 

1232: \bibitem{Smullin:2005iv}

1233: S.~J.~Smullin, A.~A.~Geraci, D.~M.~Weld, J.~Chiaverini, S.~Holmes and A.~Kapitulnik,

1234: %``New constraints on Yukawa-type deviations from Newtonian gravity at

1235: %20-microns,''

1236: arXiv:hep-ph/0508204.

1237: %%CITATION = HEP-PH 0508204;%%

1238: 

1239: 

1240: \bibitem{Adelberger:2003zx}

1241: E.~G.~Adelberger, B.~R.~Heckel and A.~E.~Nelson,

1242: %``Tests of the gravitational inverse-square law,''

1243: Ann.\ Rev.\ Nucl.\ Part.\ Sci.\  {\bf 53}, 77 (2003)

1244: [arXiv:hep-ph/0307284].

1245: %%CITATION = HEP-PH 0307284;%%

1246: 

1247: 

1248: \bibitem{:2005em}

1249: The ALEPH, DELPHI, L3, OPAL, SLD Collaborations,

1250: the LEP Electroweak Working Group, and

1251: the SLD Electroweak and Heavy Flavour Groups

1252: %``Precision electroweak measurements on the Z resonance,''

1253: arXiv:hep-ex/0509008;\\

1254: %%CITATION = HEP-EX 0509008;%%

1255: %\bibitem{Eidelman:2004wy}

1256: S.~Eidelman {\it et al.}  [Particle Data Group],

1257: %``Review of particle physics,''

1258: Phys.\ Lett.\ B {\bf 592} (2004) 1;

1259: %%CITATION = PHLTA,B592,1;%%

1260: 

1261: 

1262: \bibitem{Arkani-Hamed:1998rs}

1263: N.~Arkani-Hamed, S.~Dimopoulos and G.~R.~Dvali,

1264: %``The hierarchy problem and new dimensions at a millimeter,''

1265: Phys.\ Lett.\ B {\bf 429}, 263 (1998)

1266: [arXiv:hep-ph/9803315].

1267: %%CITATION = HEP-PH 9803315;%%

1268: 

1269: 

1270: \bibitem{Sundrum:1997js}

1271: R.~Sundrum,

1272: %``Towards an effective particle-string resolution of the cosmological

1273: %constant problem,''

1274: JHEP {\bf 9907}, 001 (1999)

1275: [arXiv:hep-ph/9708329].

1276: %%CITATION = HEP-PH 9708329;%%

1277: 

1278: 

1279: \bibitem{Moffat:2001jf}

1280: J.~W.~Moffat,

1281: %``Quantum gravity resolution to the cosmological constant problem,''

1282: arXiv:hep-ph/0102088.

1283: %%CITATION = HEP-PH 0102088;%%

1284: 

1285: 

1286: \bibitem{Sundrum:2003tb}

1287: R.~Sundrum,

1288: %``Fat Euclidean gravity with small cosmological constant,''

1289: Nucl.\ Phys.\ B {\bf 690}, 302 (2004)

1290: [arXiv:hep-th/0310251].

1291: %%CITATION = HEP-TH 0310251;%%

1292: 

1293: 

1294: \bibitem{Unruh:1980cg}

1295: W.~G.~Unruh,

1296: %``Experimental Black Hole Evaporation,''

1297: Phys.\ Rev.\ Lett.\  {\bf 46}, 1351 (1981);\\

1298: %%CITATION = PRLTA,46,1351;%%

1299: %\bibitem{Visser:1993ub}

1300: M.~Visser,

1301: %``Acoustic propagation in fluids: An Unexpected example of Lorentzian

1302: %geometry,''

1303: arXiv:gr-qc/9311028;\\

1304: %%CITATION = GR-QC 9311028;%%

1305: %\bibitem{Garay:1999sk}

1306: L.~J.~Garay, J.~R.~Anglin, J.~I.~Cirac and P.~Zoller,

1307: %``Black holes in Bose-Einstein condensates,''

1308: Phys.\ Rev.\ Lett.\  {\bf 85}, 4643 (2000)

1309: [arXiv:gr-qc/0002015];\\

1310: %%CITATION = GR-QC 0002015;%%

1311: %\bibitem{Chapline:2000en}

1312: G.~Chapline, E.~Hohlfeld, R.~B.~Laughlin and D.~I.~Santiago,

1313: %``Quantum phase transitions and the breakdown of classical general

1314: %relativity,''

1315: Int.\ J.\ Mod.\ Phys.\ A {\bf 18}, 3587 (2003)

1316: [arXiv:gr-qc/0012094].

1317: %%CITATION = GR-QC 0012094;%%

1318: 

1319: 

1320: \bibitem{Weinberg:1980kq}

1321: S.~Weinberg and E.~Witten,

1322: %``Limits On Massless Particles,''

1323: Phys.\ Lett.\ B {\bf 96}, 59 (1980).

1324: %%CITATION = PHLTA,B96,59;%%

1325: 

1326: 

1327: \bibitem{Seiberg:1994pq}

1328: N.~Seiberg,

1329: %``Electric - magnetic duality in supersymmetric nonAbelian gauge theories,''

1330: Nucl.\ Phys.\ B {\bf 435}, 129 (1995)

1331: [arXiv:hep-th/9411149].

1332: %%CITATION = HEP-TH 9411149;%%

1333: 

1334: 

1335: \bibitem{Gherghetta:2005se}

1336: T.~Gherghetta, M.~Peloso and E.~Poppitz,

1337: %``Emergent gravity from a mass deformation in warped spacetime,''

1338: arXiv:hep-th/0507245.

1339: %%CITATION = HEP-TH 0507245;%%

1340: 

1341: 

1342: \bibitem{Perlmutter:1998np}

1343: S.~Perlmutter {\it et al.}  [Supernova Cosmology Project Collaboration],

1344: %``Measurements of Omega and Lambda from 42 High-Redshift Supernovae,''

1345: Astrophys.\ J.\  {\bf 517}, 565 (1999)

1346: [arXiv:astro-ph/9812133];\\

1347: %%CITATION = ASTRO-PH 9812133;%%

1348: %\bibitem{Riess:2004nr}

1349: A.~G.~Riess {\it et al.}  [Supernova Search Team Collaboration],

1350: %``Type Ia Supernova Discoveries at z>1 From the Hubble Space Telescope:

1351: %Evidence for Past Deceleration and Constraints on Dark Energy Evolution,''

1352: Astrophys.\ J.\  {\bf 607}, 665 (2004)

1353: [arXiv:astro-ph/0402512].

1354: %%CITATION = ASTRO-PH 0402512;%%

1355: 

1356: 

1357: \bibitem{Spergel:2003cb}

1358: D.~N.~Spergel {\it et al.}  [WMAP Collaboration],

1359: %``First Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations:

1360: %Determination of Cosmological Parameters,''

1361: Astrophys.\ J.\ Suppl.\  {\bf 148}, 175 (2003)

1362: [arXiv:astro-ph/0302209];\\

1363: %%CITATION = ASTRO-PH 0302209;%%

1364: %\bibitem{deBernardis:2000gy}

1365: P.~de Bernardis {\it et al.}  [Boomerang Collaboration],

1366: %``A Flat Universe from High-Resolution Maps of the Cosmic Microwave

1367: %Background Radiation,''

1368: Nature {\bf 404}, 955 (2000)

1369: [arXiv:astro-ph/0004404];\\

1370: %%CITATION = ASTRO-PH 0004404;%%

1371: %\bibitem{Balbi:2000tg}

1372: A.~Balbi {\it et al.},

1373: %``Constraints on cosmological parameters from MAXIMA-1,''

1374: Astrophys.\ J.\  {\bf 545}, L1 (2000)

1375: [Erratum-ibid.\  {\bf 558}, L145 (2001)]

1376: [arXiv:astro-ph/0005124].

1377: %%CITATION = ASTRO-PH 0005124;%%

1378: 

1379: 

1380: \bibitem{Tegmark:2003ud}

1381: M.~Tegmark {\it et al.}  [SDSS Collaboration],

1382: %``Cosmological parameters from SDSS and WMAP,''

1383: Phys.\ Rev.\ D {\bf 69}, 103501 (2004)

1384: [arXiv:astro-ph/0310723];\\

1385: %%CITATION = ASTRO-PH 0310723;%%

1386: %\bibitem{Peacock:2001gs}

1387: J.~A.~Peacock {\it et al.},

1388: %``A measurement of the cosmological mass density from clustering in the 2dF

1389: %Galaxy Redshift Survey,''

1390: Nature {\bf 410} (2001) 169

1391: [arXiv:astro-ph/0103143].

1392: %%CITATION = ASTRO-PH 0103143;%%

1393: 

1394: 

1395: \bibitem{Kaplan:2005rr}

1396: D.~E.~Kaplan and R.~Sundrum,

1397: %``A symmetry for the cosmological constant,''

1398: arXiv:hep-th/0505265.

1399: %%CITATION = HEP-TH 0505265;%%

1400: 

1401: 

1402: \bibitem{Sundrum:2003jq}

1403: R.~Sundrum,

1404: %``Fat gravitons, the cosmological constant and sub-millimeter tests,''

1405: Phys.\ Rev.\ D {\bf 69}, 044014 (2004)

1406: [arXiv:hep-th/0306106].

1407: %%CITATION = HEP-TH 0306106;%%

1408: 

1409: 

1410: \bibitem{Goldberger:2004jt}

1411: W.~D.~Goldberger and I.~Z.~Rothstein,

1412: %``An effective field theory of gravity for extended objects,''

1413: arXiv:hep-th/0409156.

1414: %%CITATION = HEP-TH 0409156;%%

1415: 

1416: 

1417: \bibitem{Donoghue:1995cz}

1418: J.~F.~Donoghue,

1419: %``Introduction to the Effective Field Theory Description of Gravity,''

1420: arXiv:gr-qc/9512024;\\

1421: %%CITATION = GR-QC 9512024;%%

1422: %\bibitem{Burgess:2003jk}

1423: C.~P.~Burgess,

1424: %``Quantum gravity in everyday life: General relativity as an effective  field

1425: %theory,''

1426: Living Rev.\ Rel.\  {\bf 7}, 5 (2004)

1427: [arXiv:gr-qc/0311082].

1428: %%CITATION = GR-QC 0311082;%%

1429: 

1430: 

1431: \bibitem{Georgi:1990um}

1432: H.~Georgi,

1433: %``An Effective Field Theory For Heavy Quarks At Low-Energies,''

1434: Phys.\ Lett.\ B {\bf 240}, 447 (1990).

1435: %%CITATION = PHLTA,B240,447;%%

1436: 

1437: 

1438: \bibitem{Kaplan:1995uv}

1439: D.~B.~Kaplan,

1440: %``Effective field theories,''

1441: arXiv:nucl-th/9506035.

1442: %%CITATION = NUCL-TH 9506035;%%

1443: 

1444: 

1445: \bibitem{Luke:1996hj}

1446: M.~Luke and A.~V.~Manohar,

1447: %``Bound states and power counting in effective field theories,''

1448: Phys.\ Rev.\ D {\bf 55}, 4129 (1997)

1449: [arXiv:hep-ph/9610534].

1450: %%CITATION = HEP-PH 9610534;%%

1451: 

1452: 

1453: \bibitem{Manohar:1983md}

1454: A.~Manohar and H.~Georgi,

1455: %``Chiral Quarks And The Nonrelativistic Quark Model,''

1456: Nucl.\ Phys.\ B {\bf 234}, 189 (1984);\\

1457: %%CITATION = NUPHA,B234,189;%%

1458: %\bibitem{Georgi:1985kw}

1459: H.~Georgi,

1460: ``Weak Interactions And Modern Particle Theory,'';\\

1461: %\bibitem{Georgi:1986kr}

1462: H.~Georgi and L.~Randall,

1463: %``Flavor Conserving CP Violation In Invisible Axion Models,''

1464: Nucl.\ Phys.\ B {\bf 276}, 241 (1986).

1465: %%CITATION = NUPHA,B276,241;%%

1466: 

1467: 

1468: \bibitem{Luty:1997fk}

1469: M.~A.~Luty,

1470: %``Naive dimensional analysis and supersymmetry,''

1471: Phys.\ Rev.\ D {\bf 57}, 1531 (1998)

1472: [arXiv:hep-ph/9706235];\\

1473: %%CITATION = HEP-PH 9706235;%%

1474: %\bibitem{Cohen:1997rt}

1475: A.~G.~Cohen, D.~B.~Kaplan and A.~E.~Nelson,

1476: %``Counting 4pi's in strongly coupled supersymmetry,''

1477: Phys.\ Lett.\ B {\bf 412}, 301 (1997)

1478: [arXiv:hep-ph/9706275].

1479: %%CITATION = HEP-PH 9706275;%%

1480: 

1481: 

1482: \bibitem{Weinberg:1965rz}

1483: S.~Weinberg,

1484: %``Photons And Gravitons In Perturbation Theory: Derivation Of Maxwell's And

1485: %Einstein's Equations,''

1486: Phys.\ Rev.\  {\bf 138}, B988 (1965).

1487: %%CITATION = PHRVA,138,B988;%%

1488: 

1489: 

1490: \bibitem{Goh:2003yr}

1491: H.~S.~Goh, M.~A.~Luty and S.~P.~Ng,

1492: %``Supersymmetry without supersymmetry,''

1493: JHEP {\bf 0501}, 040 (2005)

1494: [arXiv:hep-th/0309103].

1495: %%CITATION = HEP-TH 0309103;%%

1496: 

1497: 

1498: \bibitem{Drees:1990yw}

1499: M.~Drees and X.~Tata,

1500: %``Signals For Heavy Exotics At Hadron Colliders And Supercolliders,''

1501: Phys.\ Lett.\ B {\bf 252}, 695 (1990).

1502: %%CITATION = PHLTA,B252,695;%%

1503: 

1504: 

1505: \bibitem{Dimopoulos:1996vz}

1506: S.~Dimopoulos, M.~Dine, S.~Raby and S.~Thomas,

1507: %``Experimental Signatures of Low Energy Gauge Mediated Supersymmetry

1508: %Breaking,''

1509: Phys.\ Rev.\ Lett.\  {\bf 76}, 3494 (1996)

1510: [arXiv:hep-ph/9601367].

1511: %%CITATION = HEP-PH 9601367;%%

1512: 

1513: 

1514: \bibitem{Buchmuller:2004rq}

1515: W.~Buchmuller, K.~Hamaguchi, M.~Ratz and T.~Yanagida,

1516: %``Supergravity at colliders,''

1517: Phys.\ Lett.\ B {\bf 588}, 90 (2004)

1518: [arXiv:hep-ph/0402179].

1519: %%CITATION = HEP-PH 0402179;%%

1520: 

1521: 

1522: \bibitem{Hamaguchi:2004df}

1523: K.~Hamaguchi, Y.~Kuno, T.~Nakaya and M.~M.~Nojiri,

1524: %``A study of late decaying charged particles at future colliders,''

1525: Phys.\ Rev.\ D {\bf 70}, 115007 (2004)

1526: [arXiv:hep-ph/0409248];

1527: %%CITATION = HEP-PH 0409248;%%

1528: %\bibitem{Feng:2004yi}

1529: J.~L.~Feng and B.~T.~Smith,

1530: %``Slepton trapping at the Large Hadron and International Linear Colliders,''

1531: Phys.\ Rev.\ D {\bf 71}, 015004 (2005)

1532: [Erratum-ibid.\ D {\bf 71}, 0109904 (2005)]

1533: [arXiv:hep-ph/0409278].

1534: %%CITATION = HEP-PH 0409278;%%

1535: 

1536: \end{thebibliography}

1537: 

1538: \end{document}

1539: 

1540: