January 2015

Cyclic Entropy: An Alternative to Inflationary

arXiv:1501.03054v2 [gr-qc] 16 Mar 2015

Cosmology

Paul Howard Frampton

Courtyard Hotel, Whalley Avenue, New Haven, CT 06511, USA.

Abstract
We address how to construct an infinitely cyclic universe model. A major consid-
eration is to make the entropy cyclic which requires the entropy to be re-set to zero
in each cycle expansion → turnaround → contraction → bounce → etc. Here we
re-set entropy at the turnaround by selecting the introverse (visible universe) from
the extroverse which is generated by the accelerated expansion. In the model, the
observed homogeneity is explained by the low entropy at the bounce, The observed
flatness arises from the contraction together with the reduction in size between the
expanding and contracting universe. The present flatness is predicted to be very
precise.
1 Introduction
It is of broad interest among all physicists, indeed all scientists, to understand the early
universe especially whether a singular bang ever occurred as suggested by the most simple-
minded interpretation of the Friedmann equation. The idea of a big bang has proved
remarkably resilient despite the knowledge that extrapolating back the Friedmann equation
to the singularity must somehow be flawed.

The discovery [1] of the cosmic microwave background(CMB) resolved a dichotomy then
existing in theoretical cosmology between steady-state and big-bang theories. The inter-
pretation of the CMB as a relic of a big bang was compelling and the steady-state theory
died. Actually in 1965 it was really a trichotomy being reduced to a dichotomy because a
third theory, a bounce in a cyclic cosmology, had been under study since 1922 [2].

An infinitely cyclic cosmology was always the most popular alternative to an initial singu-
larity but its correct development was hindered by an impossible-seeming stumbling block
created by the monotonically increasing entropy of the universe. Entropy is sometimes
perceived as not a fundamental concept because it involves a very large number of particles
and is therefore not a property of a single particle. Nevertheless, in studying theoretical
cosmology we shall argue that the entropy of the universe is the single most important
concept. Interest in the pursuit of a successful cyclic model is reflected in the existence of
three recent popular books [3–5], as well as in [6].

Nevertheless, for purely theoretical reasons, the cyclic model had been discarded due to
the Tolman Entropy Conundrum(TEC) [7, 8]. The TEC, stated too simply, is that the
entropy of the universe necessarily increases, due to the second law of thermodynamics,
and therefore cycles become larger and longer in the future, smaller and shorter in the
past, implying that a big bang must have occurred at a finite time in the past.

Some progress towards a solution of the TEC was made in [9] using the Come Back
Empty (CBE) assumption in the so-called BF model. A huge entropy was there jettisoned
at turnaround and the significantly smaller universe, empty of matter, contracted adiabat-
ically to a bounce with zero entropy. The original BF model employed so-called phantom
dark energy with equation of state ω < −1. Since this violates every well-known energy
condition, it appears more sensible to employ a cosmological constant with ω = −1 as we
insist upon here, in a standard ΛCDM model of the present era . We shall find that CBE
at turnaround is still possible. An abbreviated presentation of some of the present results
appeared in [10].

The emphasis in this article is on entropy and the second law of thermodynamics. No
cause or mechanism will be provided for the turnaround from expansion to contraction of
the bounce from contraction to expansion. The only theory used will be general relativity,
in particular the Friedmann equation to describe the expansion and contraction era of

1
the space-time described by a homogeneous and isotropic FRLW metric. At almost all
times general relativity is sufficient. For the turnaround and bounce, an unknown extra
ingredient is necessary as a part of a C-theory where C means Complete. Our assumption
is that use of the Friedmann equation is sufficient to draw conclusions about the cyclic
cosmology.

To illustrate a phenomenological fudge factor to precipitate the turnaround and bounce

one could modify the Friedmann equation to
!2
˙
a(t) 8πG

a(t)2

a(tB )2

k(t)
= ρT OT (t) 1 − 1 − − (1)
a(t) 3 a(tT )2 a(t)2 a(t)2

where tT , tB are the turnaround, bounce times respectively and where the fudge factor in
square brackets has no derivation and violates diffeomorphism invariance so is merely to
illustrate a point, in the absence of any known C-theory. In [11] it is shown that infinite
cyclicity requires that tT ∼ 1.3 Ty, at which time the radius of the extroverse is REV (tT ) ∼
1067 m, a length scale far beyond where general relativity has been tested. At the bounce,
if it is at the Planck time, the radius of the extroverse is REV (tB = 10−44 s) ∼ 10−6 m,
which is a length scale where Newton’s law of gravity has not even been tested.

Eq.(1) is a good approximation to the Friedmann equation

!2
˙
a(t) 8πG k(t)
= ρT OT (t) − (2)
a(t) 3 a(t)2

for all times except very close to t ∼ tT or t ∼ tB because a(tT ) is extremely large and
a(tB ) extremely small. In this paper we shall use only Eq.(2) and make the reasonable
assumption that corrections from effects of C-theory near to the turnaround and bounce do
not significantly change conclusions about entropy and the second law of thermodynamics.

Most of research on cosmology and the early universe in the last few decades have assumed
an inflationary paradigm where there is a short burst of superluminal accelerated expansion
early on in the universe. The present cyclic model does not use any such assumption. An
inflationary stage is redundant because the geometry becomes flat without it. This is
because the value ΩT OT AL (t) = 1 is a stable fixed point of the contraction equation, just
as it is an unstable fixed point under expansion. This fact alone is sufficient to achieve the
observed flatness without inflation. The CBE model makes the flatness much more exact
to a degree that the departure from ΩT OT (t0 ) = 1 at the present time is too small to be
observable.

The plan of this paper is that in Section 2 we discuss the present expansion era; in Section
3 we study the turnaround from expansion to contraction; in Section 4 the topic is the
bounce from contraction to expansion; finally, in Section 5, there is a concluding discussion.

2
2 Present expansion era
Assuming the cosmological principle of homogeneity and isotropy leads to the FLRW
metric
dr 2
 
2 2 2 2 2 2 2
ds = dt − a(t) + r (dθ + sin θdφ ) , (3)
1 − k(t)r 2
where a(t) is the scale factor and k the curvature. Inserting this metric into the Einstein
equation leads to two Friedmann equations. The first is the expansion equation
 2
2 ȧ(t) 8πG k(t)
H(t) = = ρT OT (t) − , (4)
a(t) 3 a(t)2

where ρT OT (t) is the total density.

Using the continuity equation ρ̇(t) + 3H(t)(ρ(t) + p(t)) = 0, differentiation of Eq.(4) gives
rise to a second equation
ä(t) 4πG
=− (ρ(t) + 3p(t)). (5)
a(t) 3

The critical density is defined by ρc (t) = [3H(t)2 /8πG] and discussion of flatness involves
the proximity to unity of the quantity
ρT OT
ΩT OT (t) = (6)
ρc (t)

Let the present time be t = t0 ≃ 1.38 × 1010 and normalize a(t) by a(t0 ) = 1. Other useful
cosmic times, all measured relative to the would-have-been bang at t = 0, are the Planck
time at tP lanck = 10−44 s, the electroweak phase transition time at tEW = 10−10 s, the onset
of matter domination at tm = 4.7 × 104 y and the onset of dark energy domination at
tDE = 9.8Gy.

In Table 1 we summarize the well-established history of the present expansion era.

3
 Table 1: History of present expansion.

TIME Scale factor Comments

t a(t)
13.8Gy 1 present time
9.8Gy < t < 13.8Gy a(t) = exp[H0 (t − t0 )] dark energy dominated
tDE = 9.8Gy 0.75 onset of dark energy domination
47ky < t < 9.8Gy a(t) ∝ t2/3 matter dominated
47ky 2.1 × 10−4 onset of matter domination
t < 47ky a(t) ∝ t1/2 radiation dominated
t = 10−10 s 1.7 × 10−15 electroweak phase transition
10−44 s < t < 10−10 s a(t) ∝ t1/2 possible inflation or bounce
t = 10−44 s 1.7 × 10−32 Planck time

In the unadorned big bang theory, one has dependences of the scale factor a(t) ∼ t1/2 for
tP lanck < t < tm , a(t) ∼ t2/3 for 4tm < t < tDE , and a(t) = exp[H0 (t − t0 )] for tDE < t.
This leads to the values of a(t) displayed in Table 1.

The two most striking observations of the visible universe are isotropy to an accuracy
1 ± O(10−5 ) and [12] flatness ΩT OT (t0 ) = 1.00 ± 0.05. In big bang theory the latter implies
that in the early universe ΩT OT (t) is given at t = tEW , tP lanck by

|ΩT OT (tEW ) − 1| ∼ O(10−26), (7)

|ΩT OT (tP lanck ) = 1| ∼ O(10−60), (8)

as we shall re-derive later by considering the contraction of a cyclic universe. Once we
incorporate cyclic entropy, the resultant flatness will be even more extreme at the Planck
time than shown in Eq.(8), see e.g. Eq.(31) below for an example of this phenomenon.

We shall focus first on the geometrical condition of flatness, then return towards the end
to address homogeneity in terms of, what is an intimately related concept, the low entropy
of the universe at the bounce.

A convenient way to discuss flatness is to rearrange Eq.(4) after division by H(t)2 as

k(t)
(ΩT OT (t) − 1) = . (9)
ȧ(t)2

In a decelerating expansion the denominator of the RHS in Eq.(9) becomes smaller and
smaller and ΩT OT (t) deviates more and more from ΩT OT (t) = 1. This is why the proximity
of [ΩT OT (t0 ) − 1] to zero imposes the strong initial conditions in Eqs.(7) and (8).

4
The most popular explanation of flatness is “inflation” [13–17] which can be defined as
insertion of a brief period of highly accelerated superluminal expansion at a time t ≃
tinf lation during the radiation-dominated era tP lanck < tinf lation < tEW . While inflating,
ȧ(t) becomes extremely large enforcing flatness sufficiently so precisely in Eq.(9) that the
subsequent decelerating expansion during tinf lation < t < tDE does not remove it.

Inflation also explains homogeneity and has other successes including the prediction of
scale-invariant density perturbations. It has been claimed that inflation has been success-
fully incorporated into string theory [18, 19]. The successful discovery of the BEH boson
adds credibility to the existence of the scalar inflaton. One possible objection to inflation,
that it leads to eternal inflation [20] and hence to a multiverse in which predictivity is
hampered by the measure problem [21], is not a fatal flaw. Only a compelling alternative
theory could cast doubt on the correctness of inflation: cyclic entropy can provide such
an alternative as discussed in the remainder of this paper.

If one insisted on criticizing inflation (any theory can be criticized!), it would be that the
initial conditions before inflation are unspecified.

5
3 Turnaround from expansion to contraction
To solve the TEC, the entropy of the introverse immediately after turnaround must es-
sentially vanish as discussed in [9]. As an aid to discussion of cyclic entropy, we display
Tables 2 and 3 which illustrate the future evolution of ΛCDM models.

In Table 2 we start from the present time t = t0 = 13.8Gy and proceed into the future first
in steps of 10Gy, then 50Gy, then 250Gy and finally 500Gy. In the second column is the
corresponding value of the scale factor a(t) with the conventional normalization a(t0 ) = 1.

In the third column, we assume the number of galaxies inside the introverse at present,
t = t0 , is exactly one trillion, 1012 . Any departure would involve a corresponding rescaling
of all the entries in this column, but the entries ≤ 1 will be unchanged by any reasonable
such departure.

For the fourth column in Table 2, the radius of the introverse (IV) is calculated as usual
from Z tT
RIV (t) = 44Gly + c dt a(t)−1 , (10)
t0

with a(t) = exp[H0 (t − t0 )]. The value of RIV (t) gradually evolves to an asymptotic value
of RIV (∞) = 58Gly, limited by the speed of light.

Because of the superluminal expansion of space, RIV (t0 ) is increased to the much larger
radius REV (t) = a(t)RIV (t0 ) by the stretching of space, as shown in the fifth columnn of
Table 2. We refer to this larger space as the extroverse (EV).

It now requires very great care to identify the appropriate introverse at turnaround to
fulfill [9] the CBE assumption necessary for cyclic entropy. One might initially be tempted
to include the Milky Way galaxy which we inhabit. That would be too Ptolemaic and a
most grievous error! The correct choice of introverse at t = tT is instead a sphere which
contains no matter at all, luminous or dark, including no black holes. It contains instead
only dark energy with no entropy and small quantities of both curvature and radiation,
this last being inevitable and actually crucial to the ensuing derivation of flatness.

In the discussions of [9], a central role was played by the fraction (f ) defined by

RIV (tT ) RIV (tT )

f= = (11)
a(tT )REV (t0 ) REV (tT )

This all-important quantity is listed in the sixth and last column of Table 2.

Because the present time t = t0 is not special, we have taken the liberty to display Table
3 which is a slightly different but more physical version of Table 2. Table 3 begins at the

6
physically important time t = tDE = 9.8Gy which is the onset of dark energy domination
and hence of the superluminal accelerated expansion. We highlight the second row which is
the present time t = t0 = tDE +4Gy, because it illustrates that a extroverse presently exists
with REV = 52Gly and volume ∼ 60% larger than the introverse. This must contains some
600 billion additional galaxies which will remain forever unobservable, given a continued
accelerated expansion.

Table 2: Dependence on tT normalized to RIV (t0 ) = REV (t0 ).

Turnaround a(tT ) Galaxies RIV (tT ) REV (tT ) f

(tT − t0 ) in introverse
0 1 1012 44Gly 44Gly 1
10Gy 2.1 1.8 × 1011 51Gly 91Gly 0.56
20Gy 4.3 2.4 × 1010 55Gly 190Gly 0.29
30Gy 8.8 3.0 × 109 56Gly 390Gly 0.14
40Gy 18 3.6 × 108 57Gly 800Gly 0.071
50Gy 37 4.8 × 107 58Gly 1.6T ly 0.036
100Gy 1.5 × 103 710 58Gly 66T ly 8.8 × 10−4
150Gy 5.7 × 104 ≤1 58Gly 2.3P ly 2.5 × 10−5
200Gy 2.0 × 106 ≤1 58Gly 87P ly 6.7 × 10−7
250Gy 7.4 × 107 ≤1 58Gly 3.2Ely 1.8 × 10−8
500Gy 5.4 × 1015 ≤1 58Gly 240Y ly 2.4 × 10−16
1Ty 3.0 × 1031 ≤1 58Gly 1.3 × 1033 Gly 4.5 × 10−32

7
 Table 3: Dependence on tT normalized to RIV (tDE ) = REV (tDE ).

Turnaround a(tT ) Galaxies RIV (tT ) REV (tT ) f

(tT − tDE ) in introverse
0 0.75 1.6 × 1012 39Gly 39Gly 1
4Gy 1 1012 44Gly 52Gly 0.85
10Gy 1.6 5.5 × 1011 49Gly 70Gly 0.70
20Gy 3.2 8.8 × 1010 53Gly 141Gly 0.38
30Gy 6.6 1.1 × 1010 56Gly 290Gly 0.19
40Gy 13 1.6 × 109 57Gly 570Gly 0.10
50Gy 28 1.4 × 108 57Gly 1.2T ly 0.047
100Gy 1.1 × 103 2800 58Gly 48T ly 1.2 × 10−3
150Gy 4.3 × 104 ≤1 58Gly 1.9P ly 3.1 × 10−5
200Gy 1.5 × 106 ≤1 58Gly 66P ly 8.8 × 10−7
250Gy 5.5 × 107 ≤1 58Gly 2.4Ely 2.4 × 10−8
500Gy 4.0 × 1015 ≤1 58Gly 180Y ly 3.2 × 10−16
1Ty 2.2 × 1031 ≤1 58Gly 9.7 × 1032 Gly 6.1 × 10−32

4 Bounce from Contraction to Expansion

The CBE contracting universe contains no matter. It does, however, contain a known
amount of radiation and in approaching the bounce this radiation dominates the energy
and entropy. During the expansion era, at t = t0 its energy contribution is Ωγ (t0 ) =
1.3 × 10−4 . Its entropy will be discussed below.

The normal Friedmann equation without cyclic entropy is, from Eq.(5) and using the
radiation equation of state 3p = ρ

ä(t) 8πG
=− ργ (t), (12)
a(t) 3

and we need to calculate ȧ(tB ) which appears in this contraction version of Eq.(9)
 
 k 
|ΩT OT (tB ) − 1| = 
 . (13)
ȧ(tB )2 

Using
ργ (t0 )
ργ (t) = (14)
a(t)4

8
in Eq.(12) leads to the following
dȧ(t) 8πGργ (t0 )
=− a(t)−3 . (15)
dt 3

During the radiation-dominated era we have, from Table 1 that

a(t) = Cγ t1/2 , (16)

where
Cγ = 2.3 × 10−10 s−1/2 . (17)

Now we can use

8πGργ (t0 )
Ωγ (t0 ) = (18)
3H(t0 )2
and
|k| = H(t0 )2 (19)
to rewrite Eq.(15) in the more useful form

Cγ6
Z Z
d(ȧ) = dt t−3/2 (20)
4H(t0 )4 Ωγ (t0 )2
and inserting
|k| = H(t0 )2 (21)
in Eq.(13) and performing the integrals gives the flatness result

|ΩT OT AL (tB ) − 1| = CΩ tB , (22)

in which the coefficient, CΩ , is readily calculated to be

Cγ6
CΩ = = 3.9 × 10−16 s−1 . (23)
4H(t0 )2 Ωγ (t0 )2

Eqs.(22) and (23) now give, ignoring prefactors which are O(1), at the electroweak and
Planck times, tEW = 10−10 s and tP lanck = 10−44 s, respectively

|ΩT OT AL (tEW ) − 1| ∼ O(10−26 ) (24)

and
|ΩT OT AL (tP lanck ) − 1| ∼ O(10−60 ) (25)
which are nothing more than Eqs. (7) and (8) respectively, now rederived using time
reversal acting on the expansion era.

9
But the requirement of cyclic entropy importantly adds one more multiplicative factor
on the right-hand-side of Eq.(22) because the CBE contracting universe is described by a
modified scale factor â(t) = f a(t), where f is the usually small fraction given in the last
column of Tables 2 and 3 and in the fourth column of Table 4.

With cyclic entropy, for the contracting universe, Eq.(15) must therefore be replaced by
˙
dâ(t) 8πGργ (t0 )
=− â(t)−3 . (26)
dt 3

Inserting now the defining CBE relationship â(t) = f a(t) we easily find that Eq.(22)
inherits a multiplicative factor which is a fourth power of the CBE contraction fraction,
f 4 , giving the new flatness result

|ΩT OT AL (tB ) − 1| = f 4 CΩ tB . (27)

Let us illustrate the remarkable implications of Eq.(27) by two examples, using the results
displayed in Tables 2 and 3.

As a first example, consider the turnaround time tT satisfying (tT −tDE ) = 150Gly. This is
the shortest time after which the visible universe contains ≤ 1 galaxies and so is where the
CBE assumption first becomes straightforward to implement. For this case f = 3.1 × 10−5
and hence, ignoring prefactors which are O(1), f 4 ∼ 10−18 and therefore

|ΩT OT AL (tEW ) − 1| ∼ O(10−44 ) (28)

and
|ΩT OT AL (tP lanck ) − 1| ∼ O(10−78). (29)

As a second example, consider the turnaround time tT satisfying (tT −tDE ) = 1T y so that,
from Table 3, f 4 ∼ 10−126 . This leads naturally to the remarkable initial conditions

|ΩT OT AL (tEW ) − 1| ∼ O(10−152) (30)

and
|ΩT OT AL (tP lanck ) − 1| ∼ O(10−186 ). (31)

The total energy density at the present time t = t0 becomes extremely close to the critical
density in these two examples. It is respectively

|ΩT OT AL (t0 ) − 1| ∼ O(10−18), (32)

10
for the choice (tT − tDE ) = 150Gy,and

|ΩT OT AL (t0 ) − 1| ∼ O(10−126 ), (33)

for the case (tT − tDE ) = 1T y.

From a practical and observational point of view, however, the density perturbations of
the CMB render it impossible to check Eqs.(32) or (33) to more than five decimal places.

Homogeneity at the bounce is ensured by low entropy. Let us examine the value of the
dimensionless entropy ST OT AL (tB )/k which originates only from electromagnetic radiation,
ST OT AL (tB ) = Sγ (tB ).

First note that at the present time the CMB in the visible universe has dimensionless
entropy given in Weinberg’s book as [23]
 2
2π
Sγ (t0 )/k = g∗ VV U T 3 ∼ 1088 . (34)
45

The radiation entropy remains constant during adiabatic contraction so the bounce entropy
can be reliably estimated by using the extensive property that entropy is proportional to
volume:
Sγ (tB )/k ∼ 1088 f (tT )3 , (35)
where we have displayed explicitly the important dependence of f on the turnaround time
tT .

Using Table 4, the bounce entropy for (tT − tDE ) = 150Gy is thus Sγ (tB )/k ∼ 1074 which,
while a very large number, is nevertheless extremely small compared to the present total
entropy of the visible universe, ST OT AL (t0 ) ∼ 10124 , and may be sufficiently small to ensure
the observed homogeneity and isotropy.

On the other hand, if Sγ (tB )/k ∼ 1074 is still insufficiently low, then as can be confirmed
from the final column of Table 4, one may increase to a trillion years the turnaround time,
(tT − tDE ) = 1T y, and thus render the bounce entropy identically zero, Sγ (tB )/k = 0,
whereupon homogeneity and isotropy can become exact.

11
 Table 4: Dependence of bounce entropy Sγ (tB )/k on (tT − tDE )

Turnaround a(tT ) Galaxies f Bounce

(tT − tDE ) in introverse entropy (Sγ /k)
0 0.75 1.6 × 1012 1 -
4Gy 1 1012 0.85 -
10Gy 1.6 5.5 × 1011 0.70 -
20Gy 3.2 8.8 × 1010 0.38 -
30Gy 6.6 1.1 × 1010 0.19 -
40Gy 13 1.6 × 109 0.10 -
50Gy 28 1.4 × 108 0.047 -
100Gy 1.1 × 103 2800 1.2 × 10−3 -
150Gy 4.3 × 104 ≤1 3.1 × 10−5 ∼ 1074
200Gy 1.5 × 106 ≤1 8.8 × 10−7 ∼ 1070
250Gy 5.5 × 107 ≤1 2.4 × 10−8 ∼ 1065
500Gy 4.0 × 1015 ≤1 3.2 × 10−16 ∼ 1050
1Ty 2.2 × 1031 ≤1 6.1 × 10−32 ∼0

5 Discussion
Previous work on cyclic cosmology, especially back in the twentieth century, was stymied by
ignorance of the accelerated expansion which was first discovered only in 1998. The no-go
theorem or Tolman Entropy Conundrum (TEC) implicitly assumed decelerated expansion
and hence no extroverse surrounding the introverse into which to jettison the huge entropy
which accumulates from irreversible processes during expansion.

One prediction of the cyclic entropy model is that the presently observed flatness which
can be expressed as |Ω(t0 ) − 1| ≃ 0 is precise, possibly being accurate to 18 or even 128
decimal places depending on the turnaround time as exhibited in our Eqs.(32) and (33).
This precision arises from the Come Back Empty (CBE) assumption [9] where the visible
universe after turnaround generally has a radius only a very small fraction f of the radius
of the previously expanding visible universe. Such theoretical precision far exceeds that
of any possible observational measurement of Ω(t0 ).

In summary, cyclic entropy provides an alternative to inflationary cosmology, at least as

far as providing an alternative explanation of the observed homogeneity and flatness. To
be a more complete alternative would require an explanation of the density perturbations
which seed structure formation. In inflationary cosmology these are explained by quantum
fluctuations of an inflaton field. In the alternative of cyclic entropy, such fluctuations might
reasonably be expected to arise due to quantum effects during the contraction era.

12
Acknowledgement
Thanks are due to the University of Miami’s Department of Physics for kind hospitality
while writing up this paper.

References
[1] A.A. Penzias and R.W. Wilson, Astrophys. J. 142, 419 (1965).

[2] A. Friedmann, Z. Phys. 10, 377 (1922).

[3] P.J. Steinhardt and N. Turok, Endless Universe: Beyond the Big Bang. Doubleday
(2007).

[4] P.H. Frampton, Did Time Begin? Will Time End? World Scientific Publishing
Company (2009)

[5] R.Penrose, Cycles of Time: An Extraordinary New View of the Universe. Vintage
(2012).

[6] K.H. Shankar, A. Balaraman and K.C. Wali, Phys. Rev. D86, 024007 (2012).
arXiv:1203.5552[gr-qc].

[7] R.C. Tolman, Phys. Rev. 38, 1758 (1931).

[8] R.C. Tolman, Relativity, Thermodynamics and Cosmology. Oxford University Press
(1934).

[9] L. Baum and P.H. Frampton, Phys. Rev. Lett. 98, 071301 (2007). hep-th/0610213.

[10] P.H. Frampton. arXiv:1411.7887[gr-qc].

[11] P.H. Frampton. arXiv:1503.03519[gr-qc].

[12] R.H. Dicke, Gravitation and the Universe, Published by the American Philosophical
Society, Philadelphia (1970).

[13] A.A. Starobinsky, Phys. Lett. B91, 99 (1980)

[14] A.H. Guth, Phys. Rev. D23, 347 (1981).

[15] A.D. Linde, Phys. Lett. B108, 389 (1982); ibid B129, 177 (1983).

[16] A. Albrecht and P.J. Steinhardt, Phys. Rev. Lett. 48, 1220 (1982).

13
[17] A.R. Liddle and D.H. Lyth, Cosmological Inflation and Large-Scale Structure. Cam-
bridge University Press (2000).

[18] S. Kachru, R. Kallosh, A.D. Linde and S.P Trivedi, Phys. Rev. D68, 046005 (2003).
hep-th/0301240.

[19] S. Kachru, R. Kallosh, A.D. Linde, J.M. Maldacena, L.P. McAllister and S.P. Trivedi,
JCAP 0310, 013 (2003). hep-th/0308055.

[20] A.H. Guth, J. Phys. A40, 6811 (2007). hep-th/0702178

[21] A. Vilenkin, AIP Conf. Proc. 1514, 7 (2012). arXiv:1301.0121[hep-th]

[22] M. Tegmark, in “Universe or Multiverse” Editor, B. Carr. Cambridge University

Press (2007). pp 99-125. arXiv:0905.1283[physics.pop-ph]

[23] S.Weinberg, Relativity and Cosmology: Principles and Applications of the General
Theory of Relativity. John Wiley & Sons (1972).

14