What does inflation really predict?

Max TegmarkJournal of Cosmology and Astroparticle Physics 2005 · DOI · arXiv

Abstract

If the inflaton potential has multiple minima, as expected in the string theory "landscape", inflation predicts a probability distribution for cosmological parameters. This study computes the multivariate probability distribution for various classes of single-field slow-roll models, exploring dependence on inflationary energy scales, potential shape, and the choice of measure. Unless the characteristic scale $\Delta\phi$ is near the Planck scale, the only aspect of the potential that matters observationally is the statistical distribution of its peaks and troughs. For all energy scales and plausible measures considered, we obtain robust predictions: $\Omega_{tot} \approx 1 \pm 10^{-5}$, $w = -1$, and $\rho_\Lambda$ in the observed ballpark. High-energy limits predict $n_s \approx 0.96$ and $r \approx 0.15$, consistent with observational data. We discuss the coolness and smoothness problems which limit viable models. Our findings suggest a detectable gravitational wave signature ($r \sim 0.03$) for well-motivated single-field models.

Introduction: The Landscape Challenge

Cosmic inflation is the leading theory for the origin of the universe's structure, yet it is often criticized for being "too flexible." Critics argue that by fine-tuning the shape of the inflaton potential $V(\phi)$, one can "predict" almost any set of cosmological parameters. This paper tackles the "Landscape" challenge: if string theory is correct, there isn't just one potential, but a vast landscape with perhaps $10^{300}$ different minima. Max Tegmark asks: If we don't know the exact potential, but only its statistical properties, what does inflation really predict?

A simulated random inflaton potential illustrating different inflationary outcomes across a "messy" landscape.

Methods and Simulation

The author employs a numerical and analytical approach to derive probability distributions for cosmological parameters. The study utilizes Monte Carlo simulations to generate $10^9$ random inflaton potentials using Gaussian random fields. Two key scales are varied: the horizontal scale $\Delta\phi$ (the "wiggliness" of the potential) and the vertical scale $M$ (the energy density of inflation). For each random potential, standard observables—spatial curvature ($\Omega_{tot}$), dark energy properties ($\rho_\Lambda, w$), and primordial fluctuations ($Q, n_s, r, \alpha$)—are calculated using the slow-roll approximation. These results are then filtered by anthropic selection effects, such as the requirement that fluctuations are large enough to form galaxies but small enough to avoid immediate recollapse.

Key Results: Robust Predictions in a Messy Universe

The analysis reveals that despite the underlying complexity of the landscape, inflation makes remarkably robust and testable predictions.

Probability distributions for key cosmological parameters derived from the landscape simulations, showing strong peaks for observed values.
  • Flatness and Dark Energy: Across almost all models, inflation predicts a flat universe ($\Omega_{tot} \approx 1$) to within a part in $10^5$. The predicted equation of state for dark energy is $w = -1$, consistent with a cosmological constant.
  • High-Energy Universality: In the high-energy limit (where the field rolls over large distances), almost any messy potential behaves like simple $V \propto \phi^2$ inflation. This leads to a robust prediction of $n_s \approx 0.96$ and a significant gravitational wave signature ($r \approx 0.15$).
  • Gravitational Waves: The paper argues that well-motivated single-field models favor a tensor-to-scalar ratio of $r \sim 0.03$. This provides a clear target for future CMB polarization experiments like BICEP/Keck and LiteBIRD.
  • The Smoothness Problem: Tegmark identifies a conflict where anthropic selection for dark energy tends to favor much larger fluctuations ($Q$) than observed. This suggests that "natural" low-energy models are unlikely unless specific selection effects or measures are at play.

Discussion and the Measure Problem

A significant portion of the work addresses the "measure problem"—the mathematical difficulty of defining probabilities in an eternally inflating multiverse. Tegmark uses the "coolness problem" (or youngness paradox) to rule out global time-based measures. These measures predict we should live in a much hotter, younger universe; the fact that our CMB is only 2.7K rules them out with massive statistical significance.

While the study focuses on single-field inflation, it provides a rigorous framework for evaluating the predictive power of the landscape. The findings suggest that the "messiness" of fundamental physics does not preclude precise cosmological predictions, but rather shifts the focus to the statistical properties of the underlying theory.

Summary updated

Abstract

If the inflaton potential has multiple minima, as may be expected in, e.g., the string theory "landscape", inflation predicts a probability distribution for the cosmological parameters describing spatial curvature (Omega_tot), dark energy (rho_Lambda, w, etc.), the primordial density fluctuations (Omega_tot, dark energy (rho_Lambda, w, etc.). We compute this multivariate probability distribution for various classes of single-field slow-roll models, exploring its dependence on the characteristic inflationary energy scales, the shape of the potential V and and the choice of measure underlying the calculation. We find that unless the characteristic scale Delta-phi on which V varies happens to be near the Planck scale, the only aspect of V that matters observationally is the statistical distribution of its peaks and troughs. For all energy scales and plausible measures considered, we obtain the predictions Omega_tot ~ 1+-0.00001, w=-1 and rho_Lambda in the observed ballpark but uncomfortably high. The high energy limit predicts n_s ~ 0.96, dn_s/dlnk ~ -0.0006, r ~ 0.15 and n_t ~ -0.02, consistent with observational data and indistinguishable from eternal phi^2-inflation. The low-energy limit predicts 5 parameters but prefers larger Q and redder n_s than observed. We discuss the coolness problem, the smoothness problem and the pothole paradox, which severely limit the viable class of models and measures. Our findings bode well for detecting an inflationary gravitational wave signature with future CMB polarization experiments, with the arguably best-motivated single-field models favoring the detectable level r ~ 0.03. (Abridged)
BibTeX
@article{tegmark2005what,
  author = {Max Tegmark},
  title = {{What does inflation really predict?}},
  journal = {Journal of Cosmology and Astroparticle Physics},
  year = {2005},
  doi = {10.1088/1475-7516/2005/04/001},
}
Figure