MPP-2024-7

Species Scale and Primordial Gravitational Waves

Marco Scalisi1, ∗
1
Max-Planck-Institut für Physik (Werner-Heisenberg-Institut), Föhringer Ring 6, 80805 München, Germany
The species scale is a field-dependent UV cut-off for any effective field theory weakly coupled to
gravity. In this letter we show that, in the context of inflationary cosmology, a detection of pri-
mordial gravitational waves will set an upper bound on the decay rate |Λ′s /Λs | of the species scale.
Specifically, we derive this in terms of the tensor-to-scalar ratio r of power spectra of primordial
perturbations. Given the targets of current and next generation experiments, we show that any
successful detection would signify that this upper limit is of the order of unity, which is consistent
with recent discussions in the literature.

I. INTRODUCTION The implications of such species scale drop-off for infla-

tion have first been studied in [22] (see also [23]), where
arXiv:2401.09533v1 [hep-th] 17 Jan 2024

Common lore suggests that effects of quantum grav- it was shown that a universal upper bound on the in-
ity are difficult to unravel. One of the reasons is that flaton range can be derived (in [24] the same bound was
the energy at which one expects quantum gravita- pointed out). It was shown that inflationary models, with
tional effects to become relevant is the Planck scale field ranges violating this bound, do not admit a gravi-
MP ≃ 2.48 × 1018 GeV. This is about 14 orders of magni- tational weakly coupled effective description, since their
tude above the highest energy reached on Earth, namely typical energy scale is above Λs .
at the Large Hadron Collider. The rate at which the species scale varies in field space,
In recent years, we have however learned that, for grav- namely λ ≡ |Λ′s /Λs |, is unknown and depends on the
itational effective field theories (EFTs) with a large num- string effective model at hand. It is determined by the
ber Ns of light species, a more natural UV cut-off is the decay rate of the tower of states and has been subject
species scale [1–5] (see also [6–8] for earlier works), a of some research activity, mainly aiming at identifying
renormalization of the Planck mass, which is given by1 lower bounds for it [15, 19, 25, 26]. The significance of
a lower bound for such decay rate is directly connected
MP to the validity of the SDC. Arbitrarily low values of λ
Λs = √ . (1)
Ns would in fact correspond to arbitrarily large distances in
Above the species scale, gravity becomes strongly cou- field space,2 which are forbidden by the SDC. Just more
pled and no effective description, with such a number of recently, it was suggested [15, 19, 20, 28] that the rate
light particles, is thus possible. In the limit of large num- λ be also upper bounded by an O(1) quantity. Namely,
ber of species, Λs will significantly deviate from MP so the species scale Λs cannot decrease (or increase) arbi-
that quantum gravity effects can become important for trarily fast in field space. One of the implications is that
effective field theories even at lower energies. certain finite scalar field variations must be allowed in
In the context of string theory, towers of light parti- string theory.
cles appear naturally in certain corners. A typical sit- The prevailing research approach so far has been to
uation is at the boundary of moduli space. Here one study properties of the species scale in the context of well
can rigorously test the Swampland Distance Conjecture controlled effective string models (see e.g. [12, 14, 18, 20])
(SDC) [9], which states that the limit of infinite scalar and then discuss the potential implications for phe-
field variations is always associated with an infinite tower nomenology. However, a risk of this approach is the po-
of states exponentially becoming light. Analogous con- tential oversight of critical details that a more realis-
clusions can be drawn in the more conservative case of tic model may incorporate. One common simplification
super-Planckian field excursions [10, 11], namely larger in these investigations is the assumption of isotropy in
that MP . In the limit of large distances, the mass decay Kaluza-Klein directions. Nevertheless, several more real-
of the tower determines in fact a drop-off of the species istic models opt for the utilization of anisotropic com-
scale, following eq. (1), since the number of light species pactification spaces, such as warped throats, as a means
increases. This simple argument already suggests that the to maintain control over supersymmetry breaking.
species scale must be a field-dependent UV cut-off (more In this work, we take the opposite approach, namely we
evidence about the moduli-dependence of Λs was recently show how (cosmological) observations can provide pre-
given in [12–21]). cious information about properties of the species scale.
Current and upcoming Cosmic Microwave Background

∗ mscalisi@mpp.mpg.de
1 In d dimensions, this formula has a simple generalization via 2 This argument can also be used to constrain the non-geodicity
√ 1
the substitution Ns → Nsd−2 and by considering MP as the of a trajectory in field space [27]. In an accelerating background,
d-dimensional reduced Planck mass. it has been shown that this must be negligible.
 2

(CMB) experiments are going to severely constraint this goes against the natural expectation for the
r ≡ At /As , the ratio of amplitutes of tensor and scalar entropy to be proportional at least to the number
primordial perturbations. The current upper bound is Ns of light species. One can verify that the mini-
r < 0.036 [29]. Next generation experiments, such as mum size for a black hole1 to solve this conundrum
CMB-S4 and the Simons Observatory [30, 31], will tar- is precisely Rmin = Nsd−2 MP−1 , which is nothing
get smaller values, namely r = O(10−3 ). Any detection but Λ−1
s .
would be an important indication that primordial gravi-
tational waves were generated during inflation. Therefore, both arguments lead to the formula eq. (1) in
As key result of this letter, we demonstrate that any 4 dimensions, as given in the Introduction.
future detection of the tensor-to-scalar ratio r will set a In string theory, one expects all parameters to be de-
universal upper bound on the species scale decay rate as pendent on vev of moduli or scalar fields. The number
of light species Ns is no exception. In fact, it has been
Λ′s c 108 recently pointed out that, in type II compactifications,
λ= ≲ √ log (2)
Λs r r one can identify Ns with the topological string free en-
ergy [12, 14] and, therefore, extract an explicit moduli-
with c = O(10−2 –10−3 ) depending on the specific class
dependence of the species scale.
of inflation models. As universal
√ and most conservative
For large distances in field space,3 namely in the regime
estimate, we find that c < 8/2N , with N being the to-
where one can successfully apply the SDC, the species
tal number of e-foldings of quasi-exponential expansion.
scale exhibits a typical exponential dependence in terms
We show how the constant c changes and is strengthened
of the field distance ∆ϕ. This can be understood by con-
in a variety of generic models of inflation. In particular,
sidering that the number of light species is expected to
we study the cases of chaotic monomial, Starobinsky-like,
increase inversely proportional to the typical mass scale
hilltop and inverse-hilltop (or brane inflation) potentials.
as
As key-result, we find that λ is upper bounded by an
order one quantity. This is in agreement with recent the- Ns ∼ m−α MPα , (3)
oretical arguments given in [15, 19, 20, 28].
The outline of this work is as follows: we first review with α some order one number. This implies that
some of the relevant aspects and formulas of the species
scale (in sec. II) and of inflation and the associated scalar α 1− α
Λs ∼ m 2 M P 2
, (4)
field variation (in sec. III). In sec. IV, we present our main
results. In sec. V, we draw our conclusions and offer an using eq. (1). Since the SDC predicts that the tower mass
outlook for such a research direction. exponentially decreases in field space as m ∼ e−γ∆ϕ , then
one obtains
II. FIELD-DEPENDENT SPECIES SCALE Λs = Λ0 e−λ∆ϕ , (5)

The species scale was introduced to account for modifi- where Λ0 ≤ MP is the value of the species scale corre-
cations of scale hierarchies in gravitational effective field sponding to zero field displacement and λ = αγ/2 cor-
theories with large number of species. There are at least responds to the decay rate of the species scale, as given
two different ways to understand its importance: in the Introduction of this article. As a consequence of
eq. (5), one obtains a universal upper bound on the scalar
• First, Λs can be defined as the scale at which grav- field range, such as
ity becomes strongly coupled and perturbation the-
ory breaks down. In a pertubative expansion of an 1 MP
EFT with a large number Ns of species weakly cou- ∆ϕ ≤ log . (6)
λ Λs
pled to gravity, Λs is in fact the scale at which the
one loop term is of the same order of the tree level The latter equation implies that certain finite scalar field
term. This happens when the graviton propagator variations in quantum gravity are allowed at the cost of
diverges, as it was shown in [2]. In this limit, higher decreasing the cut-off from the Planck scale. In fact, in
derivative terms cannot be neglected [15, 20, 32– the limit Λs → MP , no field displacement is permitted.
34]. Conversely, infinite scalar field variations are prohibited
in quantum gravity since the limit Λs → 0 corresponds
• A second, perhaps more intuitive, argument is to a complete breakdown of the EFT.
based on black holes physics [1, 2] (see also [35]).
In this case, Λs is defined as the scale correspond-
ing to the smallest black hole still consistent with
the EFT. Black holes of Planckian size, with RBH ≃ 3 We define large field variations as ∆ϕ ≳ MP . This can definitely
MP−1 , would in fact correspond to entropies of order happen towards the boundary of moduli space, where one can
one, using the standard Bekenstein-Hawking for- even probe infinite scalar field displacements. In the bulk, one
d−2
mula SBH ≃ RBH MPd−2 , in d dimensions. However, can also traverse modest super-Planckian field ranges.
 3

III. INFLATION AND FIELD-RANGE number of e-foldings N ∼ 60 required to explain homo-

geneity and isotropy of the CMB. In this class of models,
Cosmic inflation is a primary example of an effective we find some of the most studied inflationary scenarios,
field theory, which requires a certain finite scalar field such as chaotic monomial potentials (p = 1), inverse-
variation ∆ϕ to be successfully realized. The magnitude hilltop models (1 < p < 2), Starobinsky-like inflation
of this range is strictly model-dependent and can be ob- (p = 2) and hilltop potentials (p > 2).
tained by the integral For this class of models, it has been shown that the
Z √ inflaton field range ∆ϕ exhibits certain universal prop-
erties in the large-N limit and one can obtain a well-
∆ϕ = 2ϵ dN (7) approximated estimate of it. In particular, for any p < 2
the field range [38] is given by
calculated over the total period of inflation. In the above √
equation, ϵ is the acceleration parameter defined as 2 2β 1− p
∆ϕ ≃ N 2, (10)
ϵ ≡ d log H/dN ≤ 1, with H being the Hubble param- 2−p
eter and N the number of inflationary e-foldings, defined
where we have neglected the sub-leading contributions
by dN ≡ −Hdt. Knowing the exact functional form of
related to the point in field space where inflation ter-
ϵ(N ) allows to have a precise estimate of the inflaton
minates. As an example, quadratic inflation (V ∼ ϕ2 )
range ∆ϕ. From the theoretical viewpoint, this can be
is given by β = 1/2 and p√= 1 and, in fact, returns
obtained once we have the explicit form of an inflation-
ary scalar potential. It is instead more challenging to in- r = 8/N ≃ 0.13 and ∆ϕ ≃ 2 N ≃ 15 for N = 60.
fer the precise form of ϵ(N ) just from observations. One For p = 2, the leading part of the scalar field range is
of the reasons is that CMB measurements do not allow given by a
us to have access to the whole inflationary trajectory4 , p
∆ϕ ≃ 2β log N . (11)
rather just to a restricted window of around 8 e-foldings.
The result
p is that only a small part of the area under the One example of this class is given by the famous Starobin-
curve 2ϵ(N ) can be determined. sky model [42], which corresponds to β p = 3/4 so that it
A generic lower bound for ∆ϕ was pointed out by Lyth gives r = 12/N 2 ≃ 0.003 and ∆ϕ ≃ 3/2 log N ≃ 5
in his seminal paper [36]. Assuming that ϵ(N ) is a mono- for N = 60. The more general class of α-attractors mod-
tonic increasing (towards the end of inflation) function els [43–45] corresponds p instead to β = 3α/4 and, thus,
of N , one can extract the lower bound [37, 38] r = 12α/N 2 and ∆ϕ ≃ 3α/2 log N .
r For p > 2, the N -dependent piece of the scalar field
r range will be sub-leading, namely it will be proportional
∆ϕ ≳ , (8)
0.002 to a negative power of N (see eq. (10)). The leading part
will be instead corresponding to the point where inflation
where r is the tensor-to-scalar ratio and is related to ends. The benchmark class of models, in this case, is the
the acceleration parameter as r = 16 ϵ, calculated at one of hilltop scalar potentials and any small deviation
the moment of horizon exit. Moreover, in the previous from it, such as
formula, we have assumed a total of 60 e-foldings of in- "
flationary expansion. The significance of eq. (8) is that,  n +∞  q #
ϕ X ϕ
given the target of current or upcoming experiments be- V (ϕ) = V0 1 − + cq , (12)
µ µ
ing r = O(10−3 ), any future detection of r will point q=n+1
at inflationary EFTs with Planckian or super-Planckian
where n is a positive number related to the power p as
field displacements. This motivates the study of the SDC
n = 2(1 − p)/(2 − p), the coefficients cq parametrize de-
in this context.
viations from hilltop models. It was shown in [38] that in
An interesting and very generic class of inflationary
the small-µ limit (and in the large-N limit) one can still
models is characterized by the acceleration parameter
obtain universal predictions for the CMB cosmological
scaling as [38–41]
observables, in particular
β (p − 2)2p−2 µ2p−2
ϵ= , (9) r = 25−2p , (13)
Np (p − 1)p−2 N p
with β and p being some constants. In the latter equa- and for the inflaton range
tion, we display just the leading term and neglect higher
order terms in 1/N . This is justified given the large total  1− p2
p−2 2
∆ϕ ≃ √ µ2− p , (14)
2(p − 1)
where the coefficient cq enter only sub-leading terms.
4 This is mainly due to the suppression of the CMB power spec- Note that the leading term of ∆ϕ does not depend on
trum at small angular scales. N , as already announced above.
 4

IV. BOUNDS ON SPECIES SCALE Λs

Λs
12
Now, we are ready to discuss the universal upper bound
on the species scale decay rate. We will show that focus-
10
ing on specific classes of inflation models, as presented
above, allows to strengthen the bound. The starting point
8
is noting that eq. (6) can be recast into a bound on the
slope of the species scale such as 6

Λ′s 1 MP 4
≤ log , (15)
Λs ∆ϕ H
2
where we have used that H ≤ Λs for consistent gravita-
tional inflationary EFTs. Then, one can express the Hub- r
0.00 0.01 0.02 0.03 0.04
ble parameter H at CMB scales in terms of the ampli-
tudes of scalar perturbations
p As and the tensor to scalar FIG. 1. Upper bounds on the species scale decay rate |Λ′s /Λs |.
ratio r, such as H = π 2 As r/2 MP , and obtain The blue line is the most conservative estimate for an up-
per bound and it is reproduced by using the Lyth bound for
Λ′s 1 108 ∆ϕ. The other lines represents upper bounds on |Λ′s /Λs | for
≲ log , (16) inflation models belonging to the generic class eq. (9) with
Λs 2∆ϕ r
p = {1, 2, 3} respectively.
where we have used the measured valued As ≈ 2.1×10−9
[46]. From the expression above, one can already notice
that, for current and near-future targets of r (namely 1
that ∆ϕ ∼ r p . By keeping track of the correct coeffi-
r ≃ 10−2 −10−3 ), any scalar field range of order O(1−10) cients, this leads to
would correspond to an upper bound on the species scale
slope of order O(10 − 1), respectively. Λ′s 4 7 p − 2 1 108
It is possible to obtain a more precise estimate for the ≲ 2p−2 r− p log , (20)
Λs 15 r
upper bound on |Λ′s /Λs | by expressing the scalar field
range ∆ϕ as function of r. A first conservative result is again when setting N = 60. This function for p = 3
given by using the Lyth bound, that is eq. (8). This yields (n = 4) is displayed as an orange line in fig. 1. One can
notice that, within this class of models, a detection of
Λ′s 1 108 primordial gravitational waves with r close to the current
≲ √ log . (17)
Λs 30 2r r upper bound (r ≲ 0.036), would signify that to have a
strong upper bound on the decay rate of the species scale
This function is displayed in fig. 1 as a blue line. such as |Λ′s /Λs | ≲ 0.5.
One can strengthen the bound by considering specific Finally, we discuss the class of inverse hilltop models
n
√ monomial potentials (V ∼ ϕ )
class of models. Chaotic or brane inflation, characterized by a scalar potential of
have in fact ∆ϕ ≃ 30 2r (obtained by combining eq. (9) the form
and eq. (10) for p = 1 and N = 60), which strengthens   n 
µ
the bound of a factor of 2 and, therefore, yields V (ϕ) = V0 1 − + ... , (21)
ϕ
Λ′s 1 108
≲ √ log . (18) where the dots represents sub-leading deviations during
Λs 60 2r r inflation and n is a positive power, related to 1 < p < 2
as n = 2(p − 1)/(2 − p). This class has gained much at-
This function is displayed as a yellow line in fig. 1.
tention in the context of string and supergravity effec-
Starobinsky-like potentials with scalar spectral index tive models (see e.g. [47–49]). More recently, the authors
ns = 1−2/N (namely corresponding
√ to p = 2) are charac- of [50] have argued for an improved version of the fa-
terized by ∆ϕ ≃ 15 log(60) 2r, which strengthens again mous D3-D3 inflation model, which falls again into this
the previous bound such as class. The inflaton range can be expressed in terms of the
√
tensor-to-scalar ratio r as ∆ϕ ≃ 30 2r/(2 − p), which
Λ′s 1 108 leads to a strengthened bound
≲ √ log , (19)
Λs 30 log(60) 2r r
Λ′s 1 108
for a total of 60 e-foldings of inflation. This function is ≲ √ log , (22)
Λs (2 + n)30 2r r
displayed as a green line in fig. 1.
For the class of hilltop models and its deviations, as again when we set N = 60 and express p in terms of the
given by eq. (12) (corresponding to p > 2), one can show power n. It is interesting to notice that by appropriately
 5

choosing the power n in eq. (21) one can severely lower Some final comments are in order. First, we would like
the upper bound. For example, n = 4 (typical D3-D3 to note that we have assumed that the species scale varies
scenario [48, 50]) corresponds to an upper bound which exponentially in field space. As explained in sec. III, one
is lower than the one obtained for Starobinsky-like po- important motivation for this assumption is given by the
tentials (p = 2). A detection of r close to the current ex- Lyth bound eq. (8), which suggests that, for any target
perimental upper bound (r ≲ 0.036), would correspond of current and next-generation experiments, a detection
again to |Λ′s /Λs | ≲ 0.5. of r would generically indicate a super-Planckian scalar
field displacement. In this case, the Swampland Distance
Conjecture would in fact point at an exponential varia-
V. CONCLUSIONS tion of Λs in terms of the canonical inflaton ϕ. However,
one must also acknowledge that, in concrete string the-
In this letter, we have showed that a detection of pri- ory settings, the species scale Λs may exhibit a different
mordial gravitational waves will set an upper bound on functional form in the bulk of the moduli space for O(1)
the slope |Λ′s /Λs | of the species scale. We have found an field displacements [15, 20]. Here we would like to stress
explicit expression of this bound in terms of the tensor- that, in the case that Λs grows slower than an exponen-
to-scalar ratio r (see eq. (2)) and showed that this can tial, our results are nonetheless valid and therefore still
be strengthened depending on the class of inflation model provide an upper bound on the decay rate of the species
considered. For current and near-future target values of scale.
the tensor-to-scalar ratio, we have demonstrated that this The second comment regards the fact that, in order
bound must be an O(1) quantity, a result consistent with to find the upper bound on |Λ′s /Λs |, we have considered
recent theoretical investigations [15, 19, 20, 28]. As an that the exponential behaviour of Λs kicks in when Λ0 ,
example, the case of brane inflation [47–50] can yield as defined in eq. (5), is equal to its maximum possible
a stringent upper value such as |Λ′s /Λs | ≲ 0.5, for per- value, namely the Planck scale MP . However it might
centage values of r. These main results, together with a well be that the exponential behaviour kicks in when the
summarizing plot, are presented in sec. IV. species scale is already some orders below the Planck
The existence of an upper bound on the decay rate of mass and, therefore, Λ0 < MP . This situation would lead
the species scale Λs is intrinsically related to the existence to a tighter constraint on the decay rate.
on an infrared cut-off in the EFT [19]. In the context of In conclusion, we have presented additional evidence
this work, this situation becomes directly manifest: a de- showcasing the potential of precision cosmology to pro-
tection of primordial tensor perturbations fixes immedi- vide tangible insights into properties of quantum grav-
ately both an upper bound for |Λ′s /Λs | (main result of this ity. The Swampland program has offered a convenient
work) and also the scale of inflation H (a famous result framework and effective tools to establish such a con-
that the amplitude of tensor perturbations At ∼ H 2 ), nection and extract concrete results. We look forward to
which acts in fact as infrared cut-off for the effective the- addressing many other interesting open questions in the
ory. At the same time, since we have already measured near future.
the amplitude of scalar perturbations As , which is pro-
portional to the ratio H 2 /ϵ, a detection of r will give us
precious information also about the slope ϵ = (V ′ /V )2 /2 ACKNOWLEDGMENTS
of the scalar potential V (ϕ). Therefore, we have showed
that a detection of primordial gravitational waves would We would like to thank Alvaro Herráez, Dieter Lüst,
connect IR an UV properties of inflationary EFTs. Giada Maugeri, Carmine Montella for helpful comments.

[1] G. Dvali, “Black Holes and Large N Species Solution to 61 (2013) 768–778, arXiv:1206.2365 [hep-th].
the Hierarchy Problem,” Fortsch. Phys. 58 (2010) [6] G. Veneziano, “Large N bounds on, and compositeness
528–536, arXiv:0706.2050 [hep-th]. limit of, gauge and gravitational interactions,” JHEP
[2] G. Dvali and M. Redi, “Black Hole Bound on the 06 (2002) 051, arXiv:hep-th/0110129.
Number of Species and Quantum Gravity at LHC,” [7] N. Arkani-Hamed, S. Dimopoulos, and S. Kachru,
Phys. Rev. D 77 (2008) 045027, arXiv:0710.4344 “Predictive landscapes and new physics at a TeV,”
[hep-th]. arXiv:hep-th/0501082.
[3] G. Dvali and D. Lust, “Evaporation of Microscopic [8] J. Distler and U. Varadarajan, “Random polynomials
Black Holes in String Theory and the Bound on and the friendly landscape,” arXiv:hep-th/0507090.
Species,” Fortsch. Phys. 58 (2010) 505–527, [9] H. Ooguri and C. Vafa, “On the Geometry of the String
arXiv:0912.3167 [hep-th]. Landscape and the Swampland,” Nucl. Phys. B 766
[4] G. Dvali and C. Gomez, “Species and Strings,” (2007) 21–33, arXiv:hep-th/0605264.
arXiv:1004.3744 [hep-th]. [10] F. Baume and E. Palti, “Backreacted Axion Field
[5] G. Dvali, C. Gomez, and D. Lust, “Black Hole Quantum Ranges in String Theory,” JHEP 08 (2016) 043,
Mechanics in the Presence of Species,” Fortsch. Phys. arXiv:1602.06517 [hep-th].
 6

[11] D. Klaewer and E. Palti, “Super-Planckian Spatial JCAP 02 (2019) 056, arXiv:1808.07445
Field Variations and Quantum Gravity,” JHEP 01 [astro-ph.CO].
(2017) 088, arXiv:1610.00010 [hep-th]. [32] N. Cribiori, D. Lust, and C. Montella, “Species entropy
[12] D. van de Heisteeg, C. Vafa, M. Wiesner, and D. H. and thermodynamics,” JHEP 10 (2023) 059,
Wu, “Moduli-dependent Species Scale,” arXiv:2305.10489 [hep-th].
arXiv:2212.06841 [hep-th]. [33] J. Calderón-Infante, M. Delgado, and A. M. Uranga,
[13] A. Castellano, A. Herráez, and L. E. Ibáñez, “The “Emergence of species scale black hole horizons,” JHEP
emergence proposal in quantum gravity and the species 01 (2024) 003, arXiv:2310.04488 [hep-th].
scale,” JHEP 06 (2023) 047, arXiv:2212.03908 [34] I. Basile, D. Lust, and C. Montella, “Shedding black
[hep-th]. hole light on the emergent string conjecture,”
[14] N. Cribiori, D. Lust, and G. Staudt, “Black hole arXiv:2311.12113 [hep-th].
entropy and moduli-dependent species scale,” [35] I. Basile, N. Cribiori, D. Lust, and C. Montella,
arXiv:2212.10286 [hep-th]. “Minimal Black Holes and Species Thermodynamics,”
[15] D. van de Heisteeg, C. Vafa, and M. Wiesner, “Bounds arXiv:2401.06851 [hep-th].
on Species Scale and the Distance Conjecture,” [36] D. H. Lyth, “What would we learn by detecting a
arXiv:2303.13580 [hep-th]. gravitational wave signal in the cosmic microwave
[16] D. Andriot, “Bumping into the species scale with the background anisotropy?,” Phys. Rev. Lett. 78 (1997)
scalar potential,” arXiv:2305.07480 [hep-th]. 1861–1863, arXiv:hep-ph/9606387.
[17] R. Blumenhagen, A. Gligovic, and A. Paraskevopoulou, [37] L. Boubekeur and D. H. Lyth, “Hilltop inflation,”
“The Emergence Proposal and the Emergent String,” JCAP 07 (2005) 010, arXiv:hep-ph/0502047.
arXiv:2305.10490 [hep-th]. [38] J. Garcia-Bellido, D. Roest, M. Scalisi, and I. Zavala,
[18] N. Cribiori and D. Lust, “A note on modular invariant “Lyth bound of inflation with a tilt,” Phys. Rev. D 90
species scale and potentials,” arXiv:2306.08673 no. 12, (2014) 123539, arXiv:1408.6839 [hep-th].
[hep-th]. [39] V. Mukhanov, “Quantum Cosmological Perturbations:
[19] J. Calderón-Infante, A. Castellano, A. Herráez, and Predictions and Observations,” Eur. Phys. J. C 73
L. E. Ibáñez, “Entropy Bounds and the Species Scale (2013) 2486, arXiv:1303.3925 [astro-ph.CO].
Distance Conjecture,” arXiv:2306.16450 [hep-th]. [40] D. Roest, “Universality classes of inflation,” JCAP 01
[20] D. van de Heisteeg, C. Vafa, M. Wiesner, and D. H. (2014) 007, arXiv:1309.1285 [hep-th].
Wu, “Species Scale in Diverse Dimensions,” [41] J. Garcia-Bellido, D. Roest, M. Scalisi, and I. Zavala,
arXiv:2310.07213 [hep-th]. “Can CMB data constrain the inflationary field range?,”
[21] A. Castellano, A. Herráez, and L. E. Ibáñez, “On the JCAP 09 (2014) 006, arXiv:1405.7399 [hep-th].
Species Scale, Modular Invariance and the Gravitational [42] A. A. Starobinsky, “A New Type of Isotropic
EFT expansion,” arXiv:2310.07708 [hep-th]. Cosmological Models Without Singularity,” Phys. Lett.
[22] M. Scalisi and I. Valenzuela, “Swampland distance B 91 (1980) 99–102.
conjecture, inflation and α-attractors,” JHEP 08 (2019) [43] R. Kallosh, A. Linde, and D. Roest, “Superconformal
160, arXiv:1812.07558 [hep-th]. Inflationary α-Attractors,” JHEP 11 (2013) 198,
[23] M. Scalisi, “Inflation, Higher Spins and the arXiv:1311.0472 [hep-th].
Swampland,” Phys. Lett. B 808 (2020) 135683, [44] D. Roest and M. Scalisi, “Cosmological attractors from
arXiv:1912.04283 [hep-th]. α-scale supergravity,” Phys. Rev. D 92 (2015) 043525,
[24] D. van de Heisteeg, C. Vafa, M. Wiesner, and D. H. arXiv:1503.07909 [hep-th].
Wu, “Bounds on Field Range for Slowly Varying [45] M. Scalisi, “Cosmological α-attractors and de Sitter
Positive Potentials,” arXiv:2305.07701 [hep-th]. landscape,” JHEP 12 (2015) 134, arXiv:1506.01368
[25] A. Castellano, I. Ruiz, and I. Valenzuela, “A Universal [hep-th].
Pattern in Quantum Gravity at Infinite Distance,” [46] Planck Collaboration, N. Aghanim et al., “Planck 2018
arXiv:2311.01501 [hep-th]. results. VI. Cosmological parameters,” Astron.
[26] A. Castellano, I. Ruiz, and I. Valenzuela, “Stringy Astrophys. 641 (2020) A6, arXiv:1807.06209
Evidence for a Universal Pattern at Infinite Distance,” [astro-ph.CO]. [Erratum: Astron.Astrophys. 652, C4
arXiv:2311.01536 [hep-th]. (2021)].
[27] J. Freigang, D. Lust, G.-E. Nian, and M. Scalisi, [47] G. R. Dvali and S. H. H. Tye, “Brane inflation,” Phys.
“Cosmic Acceleration and Turns in the Swampland,” Lett. B 450 (1999) 72–82, arXiv:hep-ph/9812483.
arXiv:2306.17217 [hep-th]. [48] S. Kachru, R. Kallosh, A. D. Linde, J. M. Maldacena,
[28] D. Lust, J. Masias, B. Muntz, and M. Scalisi, L. P. McAllister, and S. P. Trivedi, “Towards inflation
“Starobinsky Inflation in the Swampland,” in string theory,” JCAP 10 (2003) 013,
arXiv:2312.13210 [hep-th]. arXiv:hep-th/0308055.
[29] BICEP, Keck Collaboration, P. A. R. Ade et al., [49] R. Kallosh and A. Linde, “Polynomial α-attractors,”
“Improved Constraints on Primordial Gravitational JCAP 04 no. 04, (2022) 017, arXiv:2202.06492
Waves using Planck, WMAP, and BICEP/Keck [astro-ph.CO].
Observations through the 2018 Observing Season,” [50] C. P. Burgess and F. Quevedo, “RG-induced modulus
Phys. Rev. Lett. 127 no. 15, (2021) 151301, stabilization: perturbative de Sitter vacua and improved
arXiv:2110.00483 [astro-ph.CO]. D3-D3 inflation,” JHEP 06 (2022) 167,
[30] CMB-S4 Collaboration, K. N. Abazajian et al., arXiv:2202.05344 [hep-th].
“CMB-S4 Science Book, First Edition,”
arXiv:1610.02743 [astro-ph.CO].
[31] Simons Observatory Collaboration, P. Ade et al.,
“The Simons Observatory: Science goals and forecasts,”