Arxiv:Astro-Ph/0701357V3 21 Feb 2007

Provided byCERNDocumentServer arXiv:astro-ph/0701357v3 21 Feb 2007 CORE atog etil o nalo hm,teeeg scale energy model the tions them), inflationary of typical all a E in inac- not In otherwise certainly at (although scales. physics energy on window cessible unique a 16]. 15, provide 14, 13, mi- 12, cosmic 11, 10, obser- the 9, with [8, agreement in vation forma- excellent anisotropies in galaxy 7]. background, to temperature crowave 6, rise [5, and re- give discussed tion that perturbations widely Models resulting been have The scenario 4]. this im- [3, of alize radius, perturbations spectrum Hubble density scale-invariant the classical approximately scalar outside an light far a printing in- redshifted gen- in During are for fluctuations field mechanism vacuum perturbations. quantum successful flation, density very problems primordial a monopole provides erating and 2], flatness, horizon, [1, the to lution ic eobserve we Since M utplso eylresae 2,2,2,2,26, 25, 24, 23, [22, the scales of large alignment the very apparent an in on Evil,” detected, multipoles of been CMB “Axis already the has possi- of effect is form an It such anisotropies. that the CMB ble on study breaking and such an of disappeared, by effects subsequently inflation are has during we that broken Here effect been 21]. have rotationa may that 20, invariance possibility 19, the 18, with concerned [17, specifically years recent extensively in (and been theoretical universe studied have current and thereon) the constraints invariance, in experimental violation Lorentz of Lorentz is of of invariance models subset Rotational a – inflation. course invariance during rotational violated – is physics low-energy of property the during to sense happen makes might it what era. inflationary energies, about such con- open-minded few at relevant. be very physics is provide on gravity probes not straints quantum and experimental where unification, direct scales grand Since from of far scale the too near occurs tion fdniyprubtosd rs rmiflto,they inflation, from arise do perturbations density If so- a as proposed originally cosmology, Inflationary nti ae eakwa apn hnacherished a when happens what ask we paper this In = V δ 1 mrnso rmrilPeerdDrcino h Microwa the on Direction Preferred Primordial a of Imprints n h eue lnkmass Planck reduced the and / 4 srltdt h mltd fdniyfluctua- density of amplitude the to related is ntecsi irwv akrudaiorp,adprovide and anisotropy, era background inflationary microwave amplitudes the cosmic during generated the direction rotationa were on preferred of fluctuations a breaking density out larg small primordial picked been a the have of when could consequences era but the today, examine small we extremely be must metry cl-nain,adeaieasml oe nwihti is this which in prefer a model of simple spectrum a power examine primordial and the scale-invariant, on imprint the that oainlivrac sawl-salse etr flow of feature well-established a is invariance Rotational .INTRODUCTION I. δ ∼ 10 h − a 5 lm ti eypasbeta infla- that plausible very is it , a l ∗ ′ m ot cemn enM arl n akB Wise B. Mark and Carroll M. Sean Ackerman, Lotty 1 ′ aionaIsiueo ehooy aaea A91125 CA Pasadena, Technology, of Institute California i fteshrclhroi offiins esgetta tis it that suggest We coefficients. spherical-harmonic the of M P via E Dtd eray2,2007) 22, February (Dated: ∼ √ δM P . l ain optcntanso t magnitude. its obser- on allowing constraints inflation, put during to vations direction preferred consequences observable to a the of rather the for explain but predictions Our to robust anomalies, 52]. contrived make 51, model large-scale 50, a observed construct 49, currently to 48, this not 47, explain is 46, to aim 45, forward 44, put 42]. [43, been 41, phenomenon a 40, have quantify, 39, models to 38, hard 37, of is 36, variety 35, significance 34, statistical 33, its 32, Although 31, 30, 29, 28, 27, Here r r.O h te ad h atta h power the that fact the hand, other the On spectrum inflation- era. the con- modes during ary invariance a that translational is implies of and (1) sequence uncoupled, are Eq. wavenumbers in different function with delta Dirac The sboe ytepeec fasalvco htpoints vector that unit vector a small of a direction of presence the the in by invariance broken rotational era is inflationary the during that pose vector perturbations oainlivrac hne h otgnrlfr of form from general spectrum most power the the changes invariance rotational k eaengetn ihrpwr of powers higher neglecting are we nain arsnZloihsetu ie,primordial scale- the (i.e., to spectrum lead Harrison-Zeldovich that invariant arguments naturalness same inflation. during The relevant particle- be dimensionful might that the quantities of physics con- any Hubble or inverse astrophysical inflation the during of with stant compared scales large are to interest correspond that wavelengths [53].) studied been (Effects the also of invariance. magnitude rotational a the breaks of of that powers vector more small by suppressed be − → h oe spectrum power The oad h n fteifltoayea h physical the era, inflationary the of end the Towards timelike k ˆ eeg hsc.Voain fti sym- this of Violations physics. -energy k k eepoeteipiti ol leave would it imprint the explore we , steui etraogtedrcinof direction the along vector unit the is true. suigta xdnr vector fixed-norm a that Assuming . ymty h edn ffc ftevoainof violation the of effect leading the symmetry, sacneuneo oainlivrac.Sup- invariance. rotational of consequence a is e pta ieto sapproximately is direction spatial red P nainedrn h inflationary the during invariance l P xlctfrua o h expected the for formulas explicit ri ale pcs nti paper this In epochs. earlier in er ( k ′ h ( etro nainr etrain have perturbations inflationary on vector nydpnso h antd fthe of magnitude the on depends only ) δ δ k ( ( = ) k k sdfie by defined is ) ) δ ( P q ( ) k i Metadata, citationandsimilarpapersatcore.ac.uk ) P = 1 ( P + 1 k P ( aua oexpect to natural o h rmrildensity primordial the for ) k ( eBackground ve to ) k g ) ( δ k 3 n )( ( P k suigaparity a Assuming . k ˆ k ˆ ′ ( − k · · n ,where ), n q ) ) 2 . ic hywill they since CALT-68-2626 . k and (2) (1) 2

P (k) 1/k3) imply that g(k) in Eq. (2) should be in- It is useful to introduce the “spherical” components of dependent∝ of k. Assuming that g(k) is a k-independent the unit vector n that deﬁnes the preferred direction for constant g∗ over the scales of astrophysical interest, we rotational non-invariance, arrive at

′ ˆ 2 P (k)= P (k) 1+ g∗(k n) . (3) nx + iny nx iny · n+ = , n− = − , n0 = nz. − √2 √2 This is the form of the primordial power spectrum that (8) takes into account the leading effects of the violation of In terms of these components the unit norm condition 2 rotational invariance by a small vector in the inflationary becomes n0 2n+n− = 1. Note that we do not assume era that points in the direction n. In the next section we that the preferred− direction n coincides with the ˆz axis discuss the implications of the power spectrum in Eqs. (2) of the coordinate system used to parameterize the mi- and (3) for the anisotropy of the microwave background crowave sky (i.e., that n+ = n− = 0). Expressions anal- radiation. The breaking of rotational invariance gives rise ogous to ours have been derived by Gümrük¸cüo˘glu et al. to correlations between multipole moments that would [52] under the assumption that these two directions are normally vanish and also alters the predictions for the coincident; see also [37]. usual multipole moment correlations. In section three we discuss a simple model that realizes the form of the Using the identity primordial power spectrum in Eq. (3). Concluding remarks are given in section four. l ˆ 4π m ∗ m ˆ Pl(k e)= (Yl (e)) Yl (k), (9) · 2l +1 − II. MICROWAVE BACKGROUND mX= l

We are interested in a quantitative understanding of ′ it is straightforward to express the sought-after pertur- how the substitution, P (k) P (k), changes the pre- bation as diction for the microwave background→ anisotropy ∆T/T . The multipole moments are defined by ∞ ′ ′ ′ ′ 2 ′ m ∆T ∆(lm; l m )= ξlm;l m dkk P (k)g(k)Θl(k)Θl (k), a = dΩeY (e) (e). (4) lm l T Z0 Z (10) The anisotropy of the microwave background tempera- where ture T along the direction of the unit vector e is related to the primordial fluctuations by ′ 4π m m ∗ ξ ′ ′ = dΩkY (kˆ)(Y ′ (kˆ)) (11) lm;l m 3 l l ∆T 2l +1 Z (e)= dk Pl(kˆ e)δ(k)Θl(k), (5) 2 T 4π · 1 ˆ −1 ˆ 0 ˆ n+Y1 (k)+ n−Y (k)+ n0Y1 (k) . Z Xl × 1 where Pl is the Legendre polynomial of order l and Θl(k) k is a function of the magnitude of the wave-vector that The integral in (10) encodes information about the power includes, for example, the effects of the transfer function. spectrum and the transfer function, as well as the scale- It can only depend on the magnitude of the wave-vector dependence of the preferred-direction effect, while the since the dynamics after the inflationary era is assumed constants ξlm;l′m′ are purely geometric. The integration to be rotationally invariant. over solid angles is straightforward to perform. It is con- We would like to compute the expectation values ∗ venient to decompose the ξlm;l′m′ into coefficients of the almal′m′ to first order in the small quantity g(k) that quadratic quantities n n , via characterizesh i the primordial violation of rotational invari- i j ance. We write 2 ++ 2 −− +− ∗ ∗ ′ ′ ′ ′ ′ ′ ′ ′ − ′ ′ ′ ′ ′ ′ ξlm;l m = n+ξ + n−ξ +2n+n ξ (12) almal m = almal m 0 + ∆(lm; l m ), (6) lm;l m lm;l m lm;l m h i h i +0 −0 2 00 +2n+n0ξ ′ ′ +2n−n0ξ ′ ′ + n ξ ′ ′ . where the subscript 0 denotes the usual rotationally in- lm;l m lm;l m 0 lm;l m variant piece, ∞ These coefficients are then given by the following expres- ∗ 2 2 a a ′ ′ = δ ′ δ ′ dkk P (k)Θ (k) . (7) sions: h lm l m i0 ll mm l Z0 3

2 2 ++ (l (m + 1) )(l + m + 2)(l m) 1 (l + m + 1)(l + m + 2)(l + m + 3)(l + m + 4) ξ ′ ′ = δ ′ δ ′ − − δ ′ lm;l m − m ,m+2 l ,l (2l + 3)(2l 1) − 2 l ,l+2 (2l + 1)(2l + 3)2(2l + 5) " p − s 1 (l m)(l m 1)(l m 2)(l m 3) δ ′ − − − − − − − − , −2 l ,l 2 (2l + 1)(2l 1)2(2l 3) s − − # −− ++ ξlm;l′m′ = ξl′m′;lm,

2 2 2 2 2 2 +− 1 ( 1+ l + l + m ) ((l + 1) m )((l + 2) m ) ξ ′ ′ = δ ′ 2 δ ′ − + δ ′ − − lm;l m 2 m ,m − l ,l (2l 1)(2l + 3) l ,l+2 (2l + 1)(2l + 3)2(2l + 5) − s (l2 m2)((l 1)2 m2) + δ ′ − − − − , l ,l 2 (2l 3)(2l 1)2(2l + 1) s − − # 2 2 +0 1 (2m + 1) (l + m + 1)(l m) ((l + 1) m )(l + m + 2)(l + m + 3) ξ ′ ′ = δm′,m+1 δl′,l − + δl′,l+2 − lm;l m √2 (2l 1)(2l + 3) (2l + 1)(2l + 3)2(2l + 5) " p− s (l2 m2)(l m 1)(l m 2) δ ′ − − − − − − , − l ,l 2 (2l 3)(2l 1)2(2l + 1) s − − # −0 +0 ξ ′ ′ = ξ ′ ′ , lm;l m − l m ;lm 2 2 2 2 2 2 00 (2l +2l 2m 1) ((l + 1) m )((l + 2) m ) ξ ′ ′ = δ ′ δ ′ − − + δ ′ − − lm;l m m,m l,l (2l 1)(2l + 3) l ,l+2 (2l + 1)(2l + 3)2(2l + 5) " − s (l2 m2)((l 1)2 m2) +δ ′ − − − − . (13) l ,l 2 (2l 3)(2l 1)2(2l + 1)) s − − #

The formulas (12,13) are explicit expressions for the ge- of the order of l, this expression simpliﬁes to ometrical part of the perturbation (10). As we mentioned 2 in the introduction, it is natural to imagine that the vio- ∆(lm; lm) g∗ 2 2 m = 1+cos θ∗ (3cos θ∗ 1) . (16) lation of rotational invariance is approximately scale in- ∗ 2 almalm 0 4 " − − l # variant, which implies that it is a good approximation to h i set g(k)= g∗, a constant. If we deﬁne polar coordinates θ∗, φ∗ for the preferred direction, III. INFLATION MODEL WITH A PREFERRED DIRECTION

nx = sinθ∗cosφ∗ , ny = sinθ∗sinφ∗ , nz = cosθ∗ , (14) It is interesting to see how the rotationally non- these expressions can be compared directly with obser- invariant power spectrum in Eq. (3) can arise in an ex- vations to constrain the three parameters (g∗,θ∗, φ∗). plicit model of anisotropic inflation. We will assume that, ′ during most of the inflationary era, rotational invariance When g(k)= g∗, a simplification occurs for l = l and ′ is broken by a spacelike four-vector uµ with invariant m = m , as the dependence on the power spectrum for length the terms that violate rotational invariance ∆(lm; lm) ∗ is the same as the rotationally-invariant part alma 0. µ ν 2 lm gµν u u = m . (17) We can then find a simple expression for theirh ratio, i We will consider the effect of the energy-momentum tensor associated with this vector on the expansion of the ∆(lm; lm) g∗ 2 = sin θ∗ + (15) universe during inflation, ignoring direct couplings of a a∗ 2 h lm lmi0 " uµ to other fields. Gravitational effects of dynamical 2 2 Lorentz-violating vector fields have been considered pre- 2 2l +2l 2m 1 (3cos θ∗ 1) − − . viously in the literature [55, 56, 57, 58, 59, 60]. − (2l 1)(2l + 3) − # We assume that the four-vector uµ is non-zero only during the time interval 0

1 µν a˙ 2 χ = g ∂µχ∂ν χ (19) H = = H (1+16πGβ m ), L −2 a a b 1 and b˙ 8πGρ H = = Λ .(27) b b (1+8πGβ m2)(3 + 32πGβ m2) = β µuσ u β ( uµ)2 (20) s 1 1 Lu − 1∇ ∇µ σ − 2 ∇µ β µuσ u + λ(uµu m2) . − 3∇ ∇σ µ µ − According to the cosmic no-hair theorem, initially ex- Here λ is a Lagrange multiplier that enforces the con- panding homogeneous cosmological models in the pres- straint (17). Quantum fluctuations in the massless scalar ence of a positive cosmological constant will rapidly ap- field χ are assumed to dominate the density perturba- proach a de Sitter solution, if the other matter fields obey tions via the DGZK mechanism [54]. In that case we the dominant and strong energy conditions [61]. Our spe- need simply calculate the fluctuations in χ, without wor- cific model violates these conditions. Nevertheless, for β = β and β = 0 the kinetic term for fluctuations rying about the behavior of the inflationary potential. 3 − 1 2 We approximate the inflaton energy density as a con- about our background has the form of a field strength stant, modeling the effects of the inflaton field by a vac- tensor squared and so is ghost free. We therefore expect the configuration to be stable with respect to small uum energy ρΛ in Eq. (18). The inflationary spacetime is taken of the form fluctuations. It will turn out to be convenient to refer to a fictitious 2 2 2 2 2 2 ds = dt + a(t) dx⊥ + b(t) dz (21) isotropic metric, − since we have chosen the four-vector to be aligned along d¯s2 = dt2 +¯a(t)2[dx2 + dy2 + dz2], (28) the z-axis direction, − m in which the scale factor expands exponentially u0 =0 ,ux =0 ,uy =0 ,uz = . (22) b(t) ¯ a¯(t)= eHt (29) The energy-momentum tensor for uµ derived from (20) is [58] with an “average” Hubble parameter, (u) ρ ρ Tµν = 2β1( µu ν uρ uµ ρuν) ∇ ∇ − ∇ ∇ ¯ 1 2[ (u J ρ )+ (uρJ ) (u J ρ)] H = (2Ha + Hb). (30) − ∇ρ (µ ν) ∇ρ (µν) − ∇ρ (µ ν) 3 −2 ρσ +2m uσ ρJ uµuν + gµν u, (23) ∇ L Deviations from isotropy can be parameterized by µ where J σ is the current tensor, 2 Hb Ha µ µ µ ρ µ ǫH = − , (31) J σ = β1 uσ β2 δσ ρu β3 σu . (24) 3 H¯ − ∇ − ∇ − ∇ Given (22) and (23), the nonvanishing components of the stress tensor are where the 2/3 will become useful later. We work in the limit N∗ ǫH << 1, where N∗ = Ht¯ ∗ is the number of 2 | | µ ˙ e-foldings during the time when the four-vector u is (u) 2 b T = β1m non-zero. This assures that the violation of rotational 00 b ! invariance due to the anisotropic expansion is always a ˙ 2 small perturbation. (u) 2 2 b T = Tyy = β1m a We need to compute the correlation function xx b ! χ(x,t)χ(y,t) . Treating ǫ as a small perturbation, we h i H ˙ find that to first order in this quantity we obtain (see for (u) 2 ˙2 ¨ a˙ bb Tzz = β1m b 2bb 4 . (25) example ref. [62]) − − a ! Note that the components of the energy momentum ten- χ(x,t)χ(y,t) χI (x,t)χI (y,t) (32) h i ≃ h t i sor in our chosen background are independent of β2 and ′ ′ +i dt [HI (t ),χI (x,t)χI (y,t)] . β3. h i Z0 5

Here the interaction-picture Hamiltonian HI (t) is given The canonical commutation relations imply that χk sat- by isﬁes the normalization condition,

2 dχ (τ) dχ (τ)∗ 3 1 dχI k ∗ k ¯ 2 HI (t) = d x (b(t) a¯(t)) χk(τ) χk(τ)= i(Hτ) . x⊥ dτ − dτ − 2 " − d Z (41) 2 2 a(t) dχI We find that the properly normalized solution to Eq. (40) + a¯(t) . (33) b(t) − dx3 is # ¯ 3/2 ¯ −ǫH 1+ǫH The interaction-picture (i.e. free) field obeys the H√πτ (2) (k/H) (kτ) χk(τ)= H , (42) 2√1+ ǫ ν 1+ ǫ rotationally-invariant equation of motion, H H d2χ dχ 1 d2χ (2) I +3H¯ I I =0. (34) where Hν is a Hankel function, and dt2 dt − a¯(t)2 dx2 3 We can write the two-point correlation function (33) ν = . (43) 2+2ǫH in terms of Fourier transforms as The contribution to the Fourier transform of the two d3k x y −ik·(x−y) point χ correlation for a mode along the zˆ direction is χ( ,t)χ( ,t) = 3 e [P (k) h i (2π) χ (τ) 2. For small ǫ and kτ and large k/H¯ , this be- Z | k | H | | + (kˆ n)2∆P (k) . (35) comes · i H¯ 2 Converting to the conformal time of the isotropic metric, χ (τ) 2 1+3ǫ log(k/H¯ ) . (44) | k | ≃ 2k3 H 1 − ¯ τ = e Ht, (36) Here we have neglected terms linear in ǫ that are not −H¯ H enhanced by the large logarithm. Combining this result (0) 2 with a similar analysis for modes perpendicular to the ˆz and expanding in ǫH , we find that P (k) χ (τ) , and ≃ | k | direction reproduces the result in Eq. (39). τ 2 Finally we note that the density perturbation power 2 ′ 1 ∆P (k) 3ik ǫH dτ ′ (37) spectrum is defined by a Fourier transform with respect ≃ − ¯ −Hτ¯ Z 1/H to coordinates where physical laws have manifest rota- ′ (0) ′ (0) ∗ (0) ′ ∗ (0) ¯ 2 2 ∗ log( Hτ ) (χk (τ )χk (τ) ) (χk (τ ) χk (τ)) , tional invariance. However at time t = t the coordi- × − − nates in Eq. (21) do not exhibit manifest rotational in- h i where variance, due to the difference between a(t∗) and b(t∗). Rescaling coordinates, z z(¯a(t∗)/b(t∗)) and x⊥ H¯ − i → → χ(0)(τ)= e ikτ τ . (38) x⊥(¯a(t∗)/a(t∗)), we find that the function g(k) charac- k √ − k 2k terizing the rotationally non-invariant part of the power We assume that the modes k of astrophysical interest spectrum for the primordial density perturbations is have wavelenghts much smaller than the Hubble radius 9 ¯ at the beginning of inflation, which in our normalization g(k) = ǫH (log(k/H) N∗) ¯ 2 − implies k >> H. They cross the horizon around sixty 9 e-foldings before the end of inflation (which we take to = ǫ log(q(t∗)/H¯ ), (45) 2 H occur at about t∗). Taking kτ << 1, we find that | | where the term proportional to N∗ comes from the rescal- ¯ 2 9 H ∗ ∗ ∆P (k) ǫH log(k/H¯ ), (39) ing of coordinates and q(t ) = k/a¯(t ) is the physical ≃ 4 k3 wavelength of the mode of interest at the end of infla- where we have neglected contributions not enhanced by tion. the large logarithm. The logarithm in (45) is actually nearly constant over There is another way to derive Eq. (39). For modes values of q(t∗) of astrophysical interest. The range of with wavenumbers along the zˆ direction or perpendicular q(t∗) probed by CMB measurements is about a factor 3 to this direction, the Fourier transform of the two point of 10 , so log(q(t∗)/H¯ ) changes by roughly 7. But the function χ(x,t)χ(y,t) can be found exactly without re- modes of cosmological interest cross the deSitter hori- h i sorting to perturbation theory. For example, modes χk zon around 60 e-foldings before the end of inflation. So k z log(q(t∗)/H¯ ) is approximately 60. Hence, in this model with = kˆ (wavevectors parallel to the preferred direc- | | tion) obey the differential equation g(k) varies by about 10% over the range of modes of cosmological interest and our general expectation that 2 2 d χk dχk k setting g(k)= g∗ is a reasonable approximation has been +3H¯ + χ =0. (40) dt2 dt b(t)2 k confirmed. 6

For simplicity in this analysis we neglected terms that in a scale-invariant way, just as the amplitude of the per- directly couple uµ to χ. For example we could have added turbations themselves are approximately scale-invariant. µ ν 2 the term u u ∂µχ∂ν χ/M to the Lagrange density. It is Under that assumption, we derive a powerful set of pre- ∗ easy to see that this gives an additional scale invariant dictions for the expectation values a a ′ ′ that de- h lm l m i contribution, 3m2/M 2, to g(k). pend on only three parameters: a single amplitude g∗, and a direction on the sky defined by a unit vector n. In- vestigation of a simple model confirms the approximate IV. CONCLUDING REMARKS scale-independence of this effect. The resulting expressions (10,12,13) can be directly compared with observa- We have investigated the possibility that rotational in- tions to probe the existence of small Lorentz-violating variance may have been explicitly broken during inflation effects in the very early universe. by an effect that has disappeared in the later universe. The observed CMB temperature anisotropies provide a direct window onto the physics of the inflationary era, Acknowledgments and therefore offer a unique opportunity for constraining (and discovering) new phenomena at high scales. Our We are thankful for interesting discussions with aim has been to investigate the generic predictions we Jonathan R. Pritchard at the beginning of this work, expect from the presence of a preferred direction during Suz Tolwinski before it even began, and Chris Gordon inflation. and Dragan Huterer near its completion. This research If rotational invariance is violated during inflation, it was supported in part by the U.S. Department of Energy is natural for the effects of such a violation to show up and by the David and Lucile Packard Foundation.

[1] A. H. Guth, Phys. Rev. D 23, 347 (1981); [15] A. T. Lee et al., Astrophys. J. 561, L1 (2001). [2] A. D. Linde, Phys. Lett. B 108, 389 (1982); A. Albrecht [16] D. N. Spergel et al., arXiv:astro-ph/0603449; G. Hinshaw and P. J. Steinhardt, Phys. Rev. Lett. 48, 1220 (1982). et al., arXiv:astro-ph/0603451. [3] V. F. Mukhanov and G. V. Chibisov, JETP Lett. [17] A. Zee, Phys. Rev. D 25, 1864 (1982). 33 (1981) 532 [Pisma Zh. Eksp. Teor. Fiz. 33 (1981) [18] S. M. Carroll, G. B. Field and R. Jackiw, Phys. Rev. D 549]; A. A. Starobinsky, Phys. Lett. B 117 (1982) 175; 41, 1231 (1990). A. H. Guth and S. Y. Pi, Phys. Rev. Lett. 49 (1982) [19] S. R. Coleman and S. L. Glashow, Phys. Lett. B 405, 1110; S. W. Hawking, Phys. Lett. B 115, 295 (1982). 249 (1997) [arXiv:hep-ph/9703240]. [4] For a review of inﬂation see e.g. S. Dodelson, Modern [20] D. Colladay and V. A. Kostelecky, Phys. Rev. D 58, Cosmology, Academic Press, San Diego (2003); D. Lan- 116002 (1998) [arXiv:hep-ph/9809521]. glois, arXiv:hep-th/0405053. [21] J. Soda and S. Kanno, arXiv:gr-qc/0612069. [5] A. D. Linde, Phys. Lett. B 129 (1983) 177. [22] A. de Oliveira-Costa, M. Tegmark, M. Zaldarriaga [6] A. D. Linde, Phys. Rev. D 49, 748 (1994) and A. Hamilton, Phys. Rev. D 69, 063516 (2004) [arXiv:astro-ph/9307002]. [arXiv:astro-ph/0307282]. [7] A review of various models can be found in [23] H. K. Eriksen, F. K. Hansen, A. J. Banday, K. M. Gorski D. H. Lyth and A. Riotto, Phys. Rept. 314, 1 (1999) and P. B. Lilje, Astrophys. J. 605, 14 (2004) [Erratum- [arXiv:hep-ph/9807278]. ibid. 609, 1198 (2004)] [arXiv:astro-ph/0307507]. [8] G. F. Smoot et al., Astrophys. J. 396, L1 (1992); [24] C. J. Copi, D. Huterer and G. D. Starkman, Phys. Rev. C. L. Bennett et al., Astrophys. J. 464, L1 (1996) D 70, 043515 (2004) [arXiv:astro-ph/0310511]. [9] J. E. Ruhl et al., Astrophys. J. 599, 786 (2003); S. Masi [25] D. J. Schwarz, G. D. Starkman, D. Huterer and et al., arXiv:astro-ph/0507509; W. C. Jones et al., As- C. J. Copi, Phys. Rev. Lett. 93, 221301 (2004) trophys. J. 647, 823 (2006); [arXiv:astro-ph/0403353]. [10] M. C. Runyan et al., Astrophys. J. Suppl. 149, 265 [26] F. K. Hansen, A. J. Banday and K. M. Gorski, (2003); C. l. Kuo et al. [ACBAR collaboration], Astro- Mon. Not. Roy. Astron. Soc. 354, 1714 (2004) phys. J. 600, 32 (2004). [arXiv:astro-ph/0404206]. [11] T. J. Pearson et al., Astrophys. J. 591, 556 (2003); [27] S. Prunet, J. P. Uzan, F. Bernardeau and A. C. S. Readhead et al., Astrophys. J. 609, 498 (2004). T. Brunier, Phys. Rev. D 71, 083508 (2005) [12] P. F. Scott et al., Mon. Not. Roy. Astron. Soc. 341, 1076 [arXiv:astro-ph/0406364]. (2003); K. Grainge et al., Mon. Not. Roy. Astron. Soc. [28] H. K. Eriksen, A. J. Banday, K. M. Gorski and P. B. Lilje, 341, L23 (2003); C. Dickinson et al., extended Mon. Not. Astrophys. J. 622, 58 (2005) [arXiv:astro-ph/0407271]. Roy. Astron. Soc. 353, 732 (2004). [29] K. Land and J. Magueijo, Phys. Rev. Lett. 95, 071301 [13] A. Benoit et al. [Archeops Collaboration], Astron. As- (2005) [arXiv:astro-ph/0502237]. trophys. 399, L19 (2003); M. Tristram et al., Astron. [30] T. R. Jaﬀe, A. J. Banday, H. K. Eriksen, K. M. Gorski Astrophys. 436, 785 (2005). and F. K. Hansen, Astrophys. J. 629, L1 (2005) [14] N. W. Halverson et al., Background Astrophys. J. 568, [arXiv:astro-ph/0503213]. 38 (2002). [31] C. J. Copi, D. Huterer, D. J. Schwarz and G. D. Stark- 7

man, Mon. Not. Roy. Astron. Soc. 367, 79 (2006) [48] J. G. Cresswell, A. R. Liddle, P. Mukherjee and [arXiv:astro-ph/0508047]. A. Riazuelo, Phys. Rev. D 73, 041302 (2006) [32] K. Land and J. Magueijo, Mon. Not. Roy. Astron. Soc. [arXiv:astro-ph/0512017]. 367, 1714 (2006) [arXiv:astro-ph/0509752]. [49] S. H. S. Alexander, arXiv:hep-th/0601034. [33] A. Bernui, T. Villela, C. A. Wuensche, R. Leonardi and [50] R. A. Battye and A. Moss, Phys. Rev. D 74, 041301 I. Ferreira, arXiv:astro-ph/0601593. (2006) [arXiv:astro-ph/0602377]. [34] Y. Wiaux, P. Vielva, E. Martinez-Gonzalez and [51] L. Campanelli, P. Cea and L. Tedesco, Phys. Rev. Lett. P. Vandergheynst, Phys. Rev. Lett. 96, 151303 (2006) 97, 131302 (2006) [Erratum-ibid. 97, 209903 (2006)] [arXiv:astro-ph/0603367]. [arXiv:astro-ph/0606266]. [35] L. R. Abramo, A. Bernui, I. S. Ferreira, T. Villela [52] A. E. Gümrük¸cüo˘glu, C. R. Contaldi and M. Peloso, and C. A. Wuensche, Phys. Rev. D 74, 063506 (2006) arXiv:astro-ph/0608405. [arXiv:astro-ph/0604346]. [53] E. A. Lim, Phys. Rev. D 71, 063504 (2005) [36] J. Magueijo and R. D. Sorkin, arXiv:astro-ph/0604410. [arXiv:astro-ph/0407437]. [37] A. Hajian and T. Souradeep, Phys. Rev. D 74, 123521 [54] G. Dvali, A. Gruzinov and M. Zaldarriaga, Phys. Rev. (2006) [arXiv:astro-ph/0607153]. D 69, 023505 (2004) [arXiv:astro-ph/030359]; G. Dvali, [38] C. G. Park, C. Park and J. R. I. Gott, A. Gruzinov and M. Zaldarriaga, Phys. Rev. D 69, arXiv:astro-ph/0608129. 083505 (2004) [arXiv:astro-ph/0305548]; L. Kofman, [39] D. Huterer, New Astron. Rev. 50, 868 (2006) plenary talk COSMO-02, Chicago, September 2002, [arXiv:astro-ph/0608318]. astro-ph/0303614. [40] P. Vielva, Y. Wiaux, E. Martinez-Gonzalez and [55] V. A. Kostelecky and S. Samuel, Phys. Rev. D 40, 1886 P. Vandergheynst, New Astron. Rev. 50, 880 (2006) (1989). [arXiv:astro-ph/0609147]. [56] T. Jacobson and D. Mattingly, Phys. Rev. D 64, 024028 [41] K. Land and J. Magueijo, arXiv:astro-ph/0611518. (2001) [arXiv:gr-qc/0007031]. [42] H. K. Eriksen, A. J. Banday, K. M. Gorski, F. K. Hansen [57] T. Jacobson and D. Mattingly, Phys. Rev. D 70, 024003 and P. B. Lilje, arXiv:astro-ph/0701089. (2004) [arXiv:gr-qc/0402005]. [43] A. Berera, R. V. Buniy and T. W. Kephart, JCAP 0410, [58] S. M. Carroll and E. A. Lim, Phys. Rev. D 70, 123525 016 (2004) [arXiv:hep-ph/0311233]. (2004) [arXiv:hep-th/0407149]. [44] E. P. Donoghue and J. F. Donoghue, Phys. Rev. D 71, [59] C. Eling and T. Jacobson, Phys. Rev. D 69, 064005 043002 (2005) [arXiv:astro-ph/0411237]. (2004) [arXiv:gr-qc/0310044]. [45] C. Gordon, W. Hu, D. Huterer and T. Crawford, Phys. [60] V. A. Kostelecky, Phys. Rev. D 69, 105009 (2004) Rev. D 72, 103002 (2005) [arXiv:astro-ph/0509301]. [arXiv:hep-th/0312310]. [46] C. Armendariz-Picon, JCAP 0603, 002 (2006) [61] R. W. Wald, Phys. Rev. D 28 (1983) 2118. [arXiv:astro-ph/0509893]. [62] S. Weinberg, Phys. Rev. D 72, 043514 (2005) [47] R. V. Buniy, A. Berera and T. W. Kephart, Phys. Rev. [arXiv:hep-th/0506236]. D 73, 063529 (2006) [arXiv:hep-th/0511115].