Cosmic Inflation/8. Open Questions & Alternatives

Open Questions & Alternatives to Inflation

Despite its remarkable successes, inflation raises deep conceptual puzzles — from trans-Planckian physics and eternal self-reproduction to the string swampland conjectures — while alternative scenarios challenge whether inflation is truly the unique explanation for our universe's initial conditions.

8.1 Introduction: Triumphs and Puzzles

Inflation has achieved extraordinary empirical success. It explains why the universe is spatially flat to one part in $10^{4}$, why the CMB temperature is uniform to one part in $10^{5}$ across causally disconnected regions, and why we observe no magnetic monopoles. Most impressively, it predicts a nearly scale-invariant, adiabatic, Gaussian spectrum of primordial perturbations — precisely what Planck observes with $n_s = 0.9649 \pm 0.0042$and $r < 0.036$ (95% CL).

Yet inflation also raises profound theoretical questions that remain unresolved:

●Trans-Planckian problem: Observable modes had sub-Planckian wavelengths at early times — do we trust our predictions?
●Eternal inflation: Generic potentials lead to self-reproducing, never-ending inflation — how do we compute probabilities?
●UV sensitivity: Slow-roll is notoriously difficult to achieve in supergravity and string theory (the $\eta$ problem).
●Swampland conjectures: String theory considerations may forbid the flat potentials inflation requires.
●Alternatives exist: Ekpyrotic/cyclic and bouncing cosmologies can reproduce key observational signatures.

Chapter Goals

In this chapter we derive the trans-Planckian problem quantitatively, show when and why inflation becomes eternal, expose the $\eta$ problem in supergravity, derive the ekpyrotic mechanism, and confront inflation with the swampland conjectures. We conclude with a stochastic inflation simulation demonstrating eternal self-reproduction.

8.2 Derivation 1: The Trans-Planckian Problem

The modes we observe in the CMB today correspond to comoving wavenumber $k$ that exited the Hubble horizon during inflation. Let us trace these modes back to the onset of inflation and compute their physical wavelength.

Setup

A comoving mode $k$ has physical wavelength $\lambda_\text{phys} = a(t)/k$and physical momentum $p_\text{phys} = k / a(t)$. During inflation with $N$ e-folds, the scale factor grows as:

$$a(t_\text{end}) = a(t_\text{start}) \, e^N$$

Tracing modes backward

A mode that exits the horizon at $N_k$ e-folds before the end of inflation has$k = a_k H_k$ at horizon crossing. At the onset of inflation (which occurred $N$ e-folds before the end), the scale factor was smaller by $e^{N - N_k}$. The physical momentum of this mode at the onset is:

$$p_\text{phys}(t_\text{start}) = \frac{k}{a(t_\text{start})} = \frac{a_k H_k}{a(t_\text{start})} = H_k \, e^{N - N_k}$$

For a mode that exits the horizon $N_k \sim 60$ e-folds before the end, with total e-folds$N > 70$, and $H \sim 10^{13}$ GeV (GUT-scale inflation):

$$p_\text{phys}(t_\text{start}) \sim 10^{13} \, e^{N - 60} \; \text{GeV}$$

For $N = 70$, this gives $p_\text{phys} \sim 10^{13} \times e^{10} \sim 2 \times 10^{17}$ GeV, already approaching $M_P \approx 2.4 \times 10^{18}$ GeV. For $N = 80$:

$$p_\text{phys} \sim 10^{13} \times e^{20} \sim 5 \times 10^{21} \; \text{GeV} \gg M_P$$

The physical wavelength criterion

Equivalently, the physical wavelength at the start of inflation is:

$$\lambda_\text{phys}(t_\text{start}) = \frac{1}{p_\text{phys}} = \frac{e^{-(N-N_k)}}{H_k} \ll \ell_P = \frac{1}{M_P}$$

when $N - N_k > \ln(M_P / H) \approx \ln(10^5) \approx 12$. Since $N - N_k$ can easily exceed 12 for generic models, the modes we observe in the CMB today had trans-Planckian physical wavelengths at the start of inflation.

Implications

The trans-Planckian problem does not necessarily invalidate inflation, but it raises the question: can unknown Planck-scale physics modify the initial vacuum state of perturbations? If so, the standard prediction $n_s - 1 = -6\epsilon + 2\eta$ could receive corrections of order $H/M_P$. Models with modified dispersion relations (e.g., Corley-Jacobson) show that corrections are typically small ($\sim (H/M_P)^2$) but model-dependent. The trans-Planckian censorship conjecture (TCC) of Bedroya and Vafa (2019) proposes that no trans-Planckian mode should ever cross the Hubble horizon, which severely constrains inflationary model building.

8.3 Derivation 2: Eternal Inflation and the Measure Problem

During slow-roll inflation, the inflaton field is not perfectly classical. It receives stochastic quantum kicks from short-wavelength modes that are continuously being stretched beyond the Hubble radius. When these quantum fluctuations dominate over the classical slow-roll drift, inflation becomes eternal: it never ends globally.

Classical drift vs. quantum diffusion

In one Hubble time $\Delta t = H^{-1}$, the inflaton experiences:

$$\text{Classical drift:} \quad \Delta\phi_\text{cl} = \dot{\phi} \cdot H^{-1} = -\frac{V'}{3H} \cdot H^{-1} = -\frac{V'}{3H^2}$$

$$\text{Quantum kick:} \quad \delta\phi_\text{qu} \sim \frac{H}{2\pi}$$

The quantum kick arises because each Hubble time, modes with $k \sim aH$ exit the horizon, contributing a variance $\langle \delta\phi^2 \rangle = (H/2\pi)^2$ per e-fold.

Condition for eternal inflation

Eternal inflation occurs when the quantum diffusion dominates the classical drift:

$$\delta\phi_\text{qu} > |\Delta\phi_\text{cl}| \quad \Longrightarrow \quad \frac{H}{2\pi} > \frac{|V'|}{3H^2}$$

Rearranging and using $3H^2 M_P^2 = V$ (Friedmann equation during slow-roll):

$$\frac{3H^3}{2\pi} > |V'| \quad \Longrightarrow \quad \frac{H^2}{|V'|} > \frac{2\pi}{3H}$$

In terms of slow-roll parameters

Recall the slow-roll parameter $\epsilon_V = \frac{M_P^2}{2}\left(\frac{V'}{V}\right)^2$. The condition $H/(2\pi) > |V'|/(3H^2)$ can be rewritten using $V = 3H^2 M_P^2$:

$$\frac{H}{2\pi} > \frac{|V'|}{3H^2} \quad \Longrightarrow \quad \frac{V}{24\pi^2 M_P^4} > \frac{M_P^2}{2}\left(\frac{V'}{V}\right)^2 = \epsilon_V$$

But the left side is precisely the amplitude of the scalar power spectrum$\mathcal{P}_\mathcal{R} = V / (24\pi^2 M_P^4 \epsilon_V)$. Therefore the eternal inflation condition becomes:

$$\boxed{\mathcal{P}_\mathcal{R} \cdot \epsilon_V > \epsilon_V \quad \Longrightarrow \quad \mathcal{P}_\mathcal{R} > 1}$$

More precisely, one can show that eternal inflation occurs when:

$$\frac{V^{3/2}}{|V'| M_P^3} > \frac{6\pi}{\sqrt{2}} \approx 13.3$$

This is generically satisfied high up on the potential where $V$ is large and $|V'|$ is relatively small. For chaotic inflation $V = m^2\phi^2/2$, eternal inflation occurs for $\phi > \phi_* \sim M_P^2 / m$, which for typical values is far above the Planck scale.

The Measure Problem

In an eternally inflating spacetime, every possible outcome is realized infinitely many times. To compute the probability of observing any particular vacuum or set of physical constants, we must regulate these infinities — this is the measure problem. Different prescriptions (proper time cutoff, scale factor cutoff, causal diamond, etc.) yield vastly different probability distributions, leading to different "predictions" for low-energy physics.

The measure problem is arguably the most severe conceptual challenge for inflationary cosmology: without a solution, inflation in combination with the string landscape cannot make definite predictions. The "youngness paradox" (proper time measure predicts we should be Boltzmann brains) and the "Q-catastrophe" illustrate the sensitivity of predictions to the choice of measure.

8.4 Derivation 3: The $\eta$ Problem in Supergravity

Any UV completion of inflation must be compatible with supergravity (SUGRA), the local version of supersymmetry, since SUSY is the leading framework for physics beyond the Standard Model. However, the generic structure of the $\mathcal{N}=1$ SUGRA scalar potential makes slow-roll inflation extremely difficult to achieve — this is the celebrated $\eta$ problem.

The F-term scalar potential

In $\mathcal{N}=1$ supergravity, the F-term scalar potential is determined by the Kähler potential $K(\phi, \bar{\phi})$ and the superpotential $W(\phi)$:

$$V_F = e^{K/M_P^2} \left[ K^{i\bar{j}} D_i W \overline{D_j W} - \frac{3}{M_P^2} |W|^2 \right]$$

where $D_i W = \partial_i W + (K_i / M_P^2) W$ is the Kähler-covariant derivative and $K^{i\bar{j}}$ is the inverse Kähler metric.

The dangerous exponential prefactor

The crucial feature is the overall factor $e^{K/M_P^2}$. For a canonical Kähler potential $K = |\phi|^2$ (the minimal choice), this becomes:

$$e^{K/M_P^2} = e^{|\phi|^2 / M_P^2}$$

Expanding this around the inflationary trajectory, any scalar field $\phi$ acquires a mass contribution from this exponential. Computing $V''$ with respect to the inflaton:

$$\frac{\partial^2 V}{\partial \phi^2} \supset \frac{V}{M_P^2} + \cdots$$

The $\eta$ problem explicitly

The slow-roll parameter $\eta_V$ receives a contribution:

$$\eta_V = M_P^2 \frac{V''}{V} \supset M_P^2 \cdot \frac{1}{M_P^2} = 1$$

This means that generically, $|\eta_V| \sim \mathcal{O}(1)$, which violates the slow-roll condition $|\eta_V| \ll 1$. The inflaton mass is of order the Hubble scale: $m_\phi^2 \sim H^2$, causing the field to roll too fast for sustained inflation.

$$\boxed{\eta_V = M_P^2 \frac{V''}{V} \sim \mathcal{O}(1) \quad \text{(generic SUGRA)}}$$

Solutions to the $\eta$ problem

Shift Symmetry

If $K$ depends only on $\phi + \bar{\phi}$ (shift-symmetric Kähler potential), the inflaton direction $\text{Im}(\phi)$ is protected. The dangerous exponential does not generate a mass for the shift-symmetric direction. This is the basis of natural inflation embeddings in SUGRA.

Kähler Moduli Stabilization

In string compactifications, Kähler moduli can be stabilized via non-perturbative effects (KKLT, Large Volume Scenario). Careful tuning of the stabilization mechanism can cancel the$\mathcal{O}(1)$ contribution to $\eta$, though this typically requires fine-tuning of order $10^{-2}$.

D-brane Inflation

In D-brane inflation (KKLMMT), the inflaton is the separation between branes in the compact dimensions. The potential arises from brane-antibrane attraction, and the $\eta$problem manifests as a Coulomb-like correction. Warped throats can alleviate but not fully eliminate the issue.

Starobinsky / $R^2$ in SUGRA

The Starobinsky model can be embedded in old-minimal SUGRA with a specific higher-derivative structure that naturally produces a plateau potential. The no-scale structure of the Kähler potential cancels the dangerous mass term, giving $\eta_V \sim 2/N \ll 1$.

8.5 Derivation 4: The Ekpyrotic/Cyclic Alternative

The ekpyrotic scenario, proposed by Khoury, Ovrut, Steinhardt, and Turok (2001), generates a scale-invariant spectrum not through exponential expansion but through a slowly contracting phase with equation of state $w \gg 1$. This is inspired by M-theory brane collisions.

The ekpyrotic potential

The key ingredient is a steep, negative exponential potential:

$$V(\phi) = -V_0 \, e^{-c\,\phi / M_P}$$

with $c \gg 1$ (typically $c > \sqrt{6}$) and $V_0 > 0$. The scalar field rolls rapidly down this steep negative potential during a contracting phase.

Equation of state

For a scalar field rolling in a potential, the equation of state is:

$$w = \frac{p}{\rho} = \frac{\frac{1}{2}\dot{\phi}^2 + V}{\frac{1}{2}\dot{\phi}^2 - V}$$

For the ekpyrotic solution with $V < 0$ and $\dot{\phi}^2 \gg |V|$, the kinetic energy dominates both numerator and denominator. More precisely, the attractor solution gives:

$$w = \frac{c^2}{3} - 1$$

For $c \gg 1$, we get $w \gg 1$, which is the defining feature of the ekpyrotic phase. The scale factor contracts as a power law:

$$a(t) \propto (-t)^{2/c^2}$$

Scale-invariant spectrum from contraction

In an expanding universe, the comoving Hubble radius $(aH)^{-1}$ shrinks during inflation, causing modes to exit the horizon. In an ekpyrotic contracting universe, the Hubble radius $|aH|^{-1}$ also shrinks (because $|H|$ grows faster than $a$ shrinks). The spectrum of the Bardeen variable $\Phi$ is:

$$\mathcal{P}_\Phi(k) \propto k^{n_s - 1} \quad \text{with} \quad n_s - 1 = \frac{2}{1 - w/(w+1)} - 2 \approx \frac{4}{c^2}$$

For large $c$, $n_s \to 1$: the spectrum is nearly scale-invariant, just as in inflation. The spectral tilt is slightly blue ($n_s > 1$) in single-field ekpyrosis, which is disfavored by Planck data. However, two-field models (the "new ekpyrotic" scenario) can produce a red tilt $n_s < 1$ matching observations.

Comparison with inflation

Observable	Inflation	Ekpyrotic
$n_s$	$\approx 0.965$ (red tilt)	$\approx 1$ (1-field), red tilt possible (2-field)
$r$	$\sim 10^{-3}$ to $0.01$	Negligible ($r \ll 10^{-6}$)
$f_\text{NL}$	$\sim \mathcal{O}(\epsilon, \eta)$ (small)	$\sim \mathcal{O}(c^2) \gg 1$ (large, local shape)
Gravitational waves	Primordial B-modes predicted	No primordial B-modes

Key Challenges for Ekpyrosis

The ekpyrotic scenario must navigate the bounce problem: transitioning from contraction to expansion through a cosmological bounce. This requires violation of the null energy condition or new physics (e.g., ghost condensates, Galileon fields). The matching of perturbations through the bounce remains debated, with different prescriptions (Deruelle-Mukhanov, Durrer-Vernizzi) yielding different answers for whether the scale-invariant spectrum survives.

8.6 Derivation 5: The Swampland Conjectures

The string swampland program aims to identify which effective field theories (EFTs) can be consistently coupled to quantum gravity and which cannot. Two conjectures from this program pose a direct challenge to slow-roll inflation.

Conjecture 1: The de Sitter Conjecture

Obied, Ooguri, Spodyneiko, and Vafa (2018) conjectured that any scalar potential in a consistent theory of quantum gravity must satisfy:

$$\frac{|\nabla V|}{V} \geq \frac{c}{M_P}$$

where $c \sim \mathcal{O}(1)$ is a positive constant. For a single-field potential, this reads:

$$\frac{|V'|}{V} \geq \frac{c}{M_P}$$

Tension with slow-roll inflation

The first slow-roll parameter is:

$$\epsilon_V = \frac{M_P^2}{2} \left(\frac{V'}{V}\right)^2$$

The de Sitter conjecture implies:

$$\epsilon_V = \frac{M_P^2}{2} \left(\frac{V'}{V}\right)^2 \geq \frac{c^2}{2} \sim \mathcal{O}(1)$$

But slow-roll inflation requires $\epsilon_V \ll 1$! If $c \sim 1$, then $\epsilon_V \geq 1/2$, which is incompatible with inflation (no accelerated expansion when$\epsilon_V \geq 1$, and even $\epsilon_V \sim 0.5$ gives too few e-folds and too large a tensor-to-scalar ratio).

$$\boxed{\text{de Sitter conjecture:} \quad \epsilon_V \geq \frac{c^2}{2} \sim \mathcal{O}(1) \quad \Longleftrightarrow \quad \text{No slow-roll inflation}}$$

Conjecture 2: The Swampland Distance Conjecture

The distance conjecture (Ooguri and Vafa, 2007) states that as a scalar field traverses a distance $\Delta\phi$ in field space, a tower of states becomes exponentially light:

$$m_\text{tower} \sim m_0 \, e^{-\lambda \, \Delta\phi / M_P}$$

with $\lambda \sim \mathcal{O}(1)$. This implies the EFT breaks down for super-Planckian field excursions $\Delta\phi > d \cdot M_P$ with $d \sim \mathcal{O}(1)$.

Implications for large-field inflation

Many inflationary models (chaotic, natural, axion monodromy) require super-Planckian field excursions $\Delta\phi > M_P$. The Lyth bound connects the field range to the tensor-to-scalar ratio:

$$\frac{\Delta\phi}{M_P} \geq \frac{1}{\sqrt{8}} \left(\frac{r}{0.01}\right)^{1/2}$$

If the distance conjecture restricts $\Delta\phi \lesssim M_P$, then$r \lesssim 0.01$, which rules out large classes of inflationary models. Combined with the de Sitter conjecture, the swampland program potentially excludes all standard slow-roll inflation.

The Landscape vs. Swampland Debate

The swampland conjectures remain controversial. Many string theorists argue that the de Sitter conjecture is too strong and that metastable de Sitter vacua (as in KKLT) exist in the landscape. The refined de Sitter conjecture allows $|\nabla V|/V \geq c/M_P$ OR $\min(\nabla_i \nabla_j V)/V \leq -c'/M_P^2$, which permits hilltop inflation. The debate is one of the most active areas at the intersection of string theory and cosmology.

8.7 Applications and Observational Tests

Multiverse & Anthropic Principle

If eternal inflation populates a landscape of $\sim 10^{500}$ string vacua, the cosmological constant and other parameters may be determined anthropically. Weinberg's (1987) prediction of $\Lambda > 0$ using anthropic reasoning was confirmed in 1998. However, without solving the measure problem, the multiverse framework lacks predictive power.

String Cosmology Landscape

The string landscape provides concrete mechanisms for inflation: brane inflation (KKLMMT), axion monodromy (Silverstein-Westphal), fiber inflation, and Kähler moduli inflation. Each model makes specific predictions for $n_s$, $r$, and non-Gaussianity. Current data constrain but do not uniquely select a string inflation model.

Bouncing Cosmologies

Beyond ekpyrosis, other bouncing cosmologies include the matter bounce (Wands, 1999; Finelli-Brandenberger, 2002), which has a matter-dominated contraction producing scale-invariant perturbations; and string gas cosmology (Brandenberger-Vafa, 1989), which uses string winding modes to explain three large spatial dimensions and generates a scale-invariant spectrum via thermal fluctuations.

Observational Discriminators

Key observables to distinguish inflation from alternatives: (1) Primordial gravitational waves ($r > 0.001$ favors inflation); (2) Non-Gaussianity ($f_\text{NL}^\text{local} \sim 1$ for ekpyrotic, $\sim 0$ for inflation); (3) Spectral running ($dn_s/d\ln k$); (4) Isocurvature modes (constrained by Planck); (5) CMB B-mode polarization from next-generation experiments (CMB-S4, LiteBIRD).

8.8 Historical Context

1983

Vilenkin shows that inflation can be eternal to the future: quantum fluctuations allow the inflaton to diffuse up the potential in some regions, leading to perpetual self-reproduction of inflating domains. This introduces the concept of eternal inflation.

1986

Linde develops the theory of self-reproducing inflationary universe (chaotic eternal inflation). He shows that in chaotic inflation with $V = \lambda \phi^4$, regions where $\phi$ is sufficiently large undergo eternal self-reproduction, creating an infinite fractal multiverse.

2001

Khoury, Ovrut, Steinhardt, and Turok propose the ekpyrotic scenario, inspired by the Hořava-Witten picture of M-theory. Two boundary branes collide in the extra dimension, producing a big bang-like event. Steinhardt and Turok extend this to the cyclic model, where bang/crunch cycles repeat eternally.

2003

KKLT (Kachru, Kallosh, Linde, Trivedi) construct the first explicit de Sitter vacua in string theory, establishing the string landscape framework and enabling model-building for string inflation.

2006

Bousso and Polchinski and others solidify the landscape picture with $\sim 10^{500}$ vacua, igniting the landscape/anthropic debate that continues to this day.

2018

Obied, Ooguri, Spodyneiko, and Vafa propose the de Sitter swampland conjecture, reigniting debate about whether slow-roll inflation is compatible with quantum gravity. The conjecture sparks hundreds of papers exploring its implications and potential loopholes.

2019

Bedroya and Vafa propose the Trans-Planckian Censorship Conjecture (TCC), which forbids any quantum fluctuation from crossing the Hubble horizon if it was ever trans-Planckian. This severely constrains inflationary model building (requiring $V^{1/4} \lesssim 10^9$ GeV) and revives interest in alternatives.

8.9 Python Simulation: Stochastic & Ekpyrotic Inflation

We simulate the stochastic inflation dynamics: the inflaton evolves under classical slow-roll drift plus random quantum kicks of amplitude $H/(2\pi)$ per Hubble time. Multiple realizations illustrate how some regions inflate eternally while others roll to the minimum. We also plot the ekpyrotic potential and its effective slow-roll parameters for comparison.

Stochastic Inflation and Ekpyrotic Potential Comparison

Python

script.py143 lines

import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

# ===================================================================
# PART 1: Stochastic Inflation (classical drift + quantum kicks)
# ===================================================================
# We use a simple m^2 phi^2 / 2 potential in Planck units (M_P = 1)
# Slow-roll: dphi/dN = -V'/V * M_P^2 = -M_P^2 * m^2 phi / (m^2 phi^2/2) = -2/phi
# Quantum kick per e-fold: delta_phi = H/(2*pi), H^2 = V/(3 M_P^2) = m^2 phi^2 / 6

np.random.seed(42)

m = 1e-6       # inflaton mass in Planck units (realistic)
M_P = 1.0

N_efolds = 300      # maximum e-folds to simulate
dN = 0.1            # step size in e-folds
N_steps = int(N_efolds / dN)
N_realizations = 12

phi_end = np.sqrt(2) * M_P   # inflation ends when epsilon_V = 1 -> phi = sqrt(2)

fig, axes = plt.subplots(2, 2, figsize=(14, 11))

# --- Panel 1: Multiple stochastic realizations ---
ax1 = axes[0, 0]
colors_stoch = plt.cm.plasma(np.linspace(0.15, 0.85, N_realizations))

phi_start_values = [15.0, 18.0, 20.0, 22.0, 25.0, 28.0,
                    15.0, 18.0, 20.0, 22.0, 25.0, 28.0]

for i in range(N_realizations):
    phi = np.zeros(N_steps + 1)
    phi[0] = phi_start_values[i]
    ended = False
    for n in range(N_steps):
        if phi[n] < phi_end or phi[n] < 0.1:
            phi[n+1:] = phi[n]
            ended = True
            break
        # Classical drift per e-fold: dphi/dN = -2 M_P^2 / phi
        drift = -2.0 * M_P**2 / phi[n] * dN
        # Hubble parameter
        H = m * phi[n] / np.sqrt(6.0) / M_P
        # Quantum kick per e-fold
        kick = H / (2.0 * np.pi) * np.sqrt(dN) * np.random.randn()
        phi[n+1] = phi[n] + drift + kick
        if phi[n+1] < 0.01:
            phi[n+1:] = phi[n+1]
            break

N_arr = np.arange(N_steps + 1) * dN
    label_str = f'phi_0 = {phi_start_values[i]:.0f}'
    ax1.plot(N_arr, phi, color=colors_stoch[i], alpha=0.7, linewidth=1.2, label=label_str)

ax1.axhline(y=phi_end, color='red', linestyle='--', linewidth=1.5, label='phi_end (epsilon=1)')
ax1.set_xlabel('e-folds N', fontsize=12)
ax1.set_ylabel('phi / M_P', fontsize=12)
ax1.set_title('Stochastic Inflation: Multiple Realizations', fontsize=13, fontweight='bold')
ax1.set_xlim(0, 200)
ax1.set_ylim(0, 40)
ax1.legend(fontsize=7, ncol=2, loc='upper right')
ax1.grid(True, alpha=0.3)

# --- Panel 2: Classical vs quantum kick magnitudes ---
ax2 = axes[0, 1]
phi_range = np.linspace(1.0, 50.0, 500)
classical_drift = np.abs(-2.0 * M_P**2 / phi_range)  # |dphi/dN|_classical
H_vals = m * phi_range / np.sqrt(6.0) / M_P
quantum_kick = H_vals / (2.0 * np.pi)                 # sigma per e-fold

ax2.semilogy(phi_range, classical_drift, 'cyan', linewidth=2, label='|Classical drift| per e-fold')
ax2.semilogy(phi_range, quantum_kick, 'magenta', linewidth=2, label='Quantum kick (rms) per e-fold')

# Find crossing point
ratio = quantum_kick / classical_drift
cross_idx = np.argmin(np.abs(ratio - 1.0))
phi_eternal = phi_range[cross_idx]
ax2.axvline(x=phi_eternal, color='yellow', linestyle='--', linewidth=1.5,
            label=f'Eternal inflation: phi* = {phi_eternal:.1f} M_P')
ax2.fill_betweenx([1e-12, 1e2], phi_eternal, 50, alpha=0.1, color='magenta')
ax2.text(phi_eternal + 1, 1e-8, 'Eternal\nregime', color='magenta', fontsize=10)

ax2.set_xlabel('phi / M_P', fontsize=12)
ax2.set_ylabel('Field displacement per e-fold', fontsize=12)
ax2.set_title('Classical Drift vs Quantum Kicks', fontsize=13, fontweight='bold')
ax2.legend(fontsize=9)
ax2.grid(True, alpha=0.3)
ax2.set_ylim(1e-12, 1e-2)

# --- Panel 3: Ekpyrotic potential ---
ax3 = axes[1, 0]
V0 = 1.0
c_values = [3, 6, 10, 15]
phi_ek = np.linspace(0, 5, 500)
colors_ek = ['#FF6B6B', '#FFD93D', '#6BCB77', '#4D96FF']

for idx_c, c_val in enumerate(c_values):
    V_ek = -V0 * np.exp(-c_val * phi_ek / M_P)
    ax3.plot(phi_ek, V_ek, color=colors_ek[idx_c], linewidth=2, label=f'c = {c_val}')

ax3.axhline(y=0, color='white', linestyle=':', linewidth=0.5)
ax3.set_xlabel('phi / M_P', fontsize=12)
ax3.set_ylabel('V(phi) / V_0', fontsize=12)
ax3.set_title('Ekpyrotic Potential V = -V_0 exp(-c phi/M_P)', fontsize=13, fontweight='bold')
ax3.legend(fontsize=10)
ax3.grid(True, alpha=0.3)

# --- Panel 4: Ekpyrotic effective parameters ---
ax4 = axes[1, 1]
phi_ek2 = np.linspace(0.01, 5, 500)

for idx_c, c_val in enumerate(c_values):
    V_ek = -V0 * np.exp(-c_val * phi_ek2 / M_P)
    Vp_ek = V0 * c_val / M_P * np.exp(-c_val * phi_ek2 / M_P)
    Vpp_ek = -V0 * (c_val / M_P)**2 * np.exp(-c_val * phi_ek2 / M_P)

# epsilon_V = (M_P^2 / 2) (V'/V)^2 -- note V < 0
    eps_ek = 0.5 * M_P**2 * (Vp_ek / V_ek)**2
    # w = c^2/3 - 1 (constant for the attractor)
    w_ek = c_val**2 / 3.0 - 1.0

ax4.plot(phi_ek2, eps_ek, color=colors_ek[idx_c], linewidth=2,
             label=f'c={c_val}, eps={c_val**2/2:.0f}, w={w_ek:.0f}')

ax4.axhline(y=1, color='white', linestyle='--', linewidth=1, label='epsilon = 1')
ax4.set_xlabel('phi / M_P', fontsize=12)
ax4.set_ylabel('epsilon_V', fontsize=12)
ax4.set_title('Ekpyrotic Slow-Roll Parameter', fontsize=13, fontweight='bold')
ax4.legend(fontsize=8)
ax4.set_ylim(0, 120)
ax4.grid(True, alpha=0.3)

plt.tight_layout()
plt.savefig('output.png', dpi=120, bbox_inches='tight', facecolor='#1a1a2e')
print("Plot saved successfully.")
print(f"Eternal inflation threshold: phi* = {phi_eternal:.2f} M_P")
print(f"For m = {m}, H at phi=20: {m*20/np.sqrt(6):.2e} M_P")
print(f"Quantum kick at phi=20: {m*20/np.sqrt(6)/(2*np.pi):.2e} M_P per e-fold")
print(f"Classical drift at phi=20: {2.0/20:.2e} M_P per e-fold")

Click Run to execute the Python code

Code will be executed with Python 3 on the server

←Previous: Observational Signatures Back to Course Overview→

References & Further Reading

●Martin, J. & Brandenberger, R. (2001). "The Trans-Planckian Problem of Inflationary Cosmology." Phys. Rev. D 63, 123501.
●Vilenkin, A. (1983). "Birth of Inflationary Universes." Phys. Rev. D 27, 2848.
●Linde, A. (1986). "Eternally Existing Self-Reproducing Chaotic Inflationary Universe." Phys. Lett. B 175, 395.
●Khoury, J., Ovrut, B.A., Steinhardt, P.J. & Turok, N. (2001). "The Ekpyrotic Universe." Phys. Rev. D 64, 123522.
●Obied, G., Ooguri, H., Spodyneiko, L. & Vafa, C. (2018). "De Sitter Space and the Swampland." arXiv:1806.08362.
●Baumann, D. & McAllister, L. (2015). Inflation and String Theory. Cambridge University Press.
●Bedroya, A. & Vafa, C. (2019). "Trans-Planckian Censorship and the Swampland." arXiv:1909.11063.
●Brandenberger, R. (2017). "Initial Conditions for Inflation — A Short Review." Int. J. Mod. Phys. D 26, 1740002.