The Massive Flat Space Limit of Cosmological Correlators

This paper introduces a novel massive flat-space limit for cosmological correlators that preserves finite internal masses through a double-scaling procedure, deriving a general reduction formula to express these correlators in terms of flat-space amplitudes and applying it to reveal new non-Gaussian bispectrum shapes from heavy particle exchanges that cannot be captured by standard local effective field theory operators.

The Gist

The Big Picture: Listening to the Universe's Echo

Imagine the universe during its earliest moments (inflation) as a giant, expanding drum. When you hit a drum, it vibrates, and those vibrations travel across the surface. In cosmology, these "vibrations" are quantum fluctuations—tiny ripples in space and time. Scientists call the mathematical description of these ripples "cosmological correlators."

Usually, figuring out what these ripples tell us about the particles inside the drum is incredibly hard. The math is messy because the universe is expanding, and the rules of physics there (curved space) are different from the flat, empty space we are used to in our daily lives.

The Old Way: The "Massless" Shortcut

For a while, physicists had a clever trick to simplify this. They realized that if they looked at the "total energy" of a specific pattern of ripples and pushed it toward zero, the messy, expanding-universe math would magically turn into the clean, simple math of flat space (like a calm, non-expanding pool).

However, there was a catch. In this "old trick," the heavy particles inside the drum (the internal lines in the math diagrams) acted as if they had no mass at all. It was like trying to understand a heavy bowling ball by pretending it was a feather. You could get the shape of the wave, but you lost all the information about how heavy the ball actually was.

The New Discovery: The "Massive" Shortcut

This paper introduces a brand new trick, called the Massive Flat-Space (MFS) Limit.

Think of it like this: Imagine you are trying to hear a specific sound in a noisy, expanding room.

• The Old Trick: You turned down the volume of the room so much that the heavy objects in the room seemed to disappear (become massless). You could hear the sound clearly, but you couldn't tell if the source was a heavy rock or a light feather.
• The New Trick (MFS): The authors found a way to turn down the volume and simultaneously make the heavy objects feel heavier, in a perfectly balanced way.

By doing this "double-scaling" act (making external energies tiny while making internal masses huge, but keeping their ratio constant), the messy math of the expanding universe simplifies into the clean math of flat space without losing the mass of the particles.

The "Reduction Formula": The Magic Translator

The authors created a "Reduction Formula." Think of this as a universal translator or a recipe converter.

1. The Input: You have a complex, multi-layered cake (a complicated diagram of particles interacting in the expanding universe).
2. The Process: You apply the MFS limit.
3. The Output: The cake magically transforms into a simple, flat cookie (a standard Feynman diagram in flat space), but the cookie still has the exact "weight" (mass) of the original ingredients.

This is huge because it allows physicists to use the simple, well-understood rules of flat-space physics to predict what happens with heavy particles in the early universe, without having to solve the impossible math of the expanding universe directly.

The "Cosmological Phonon Collider"

The authors used this new tool to look at a specific scenario: what happens when heavy particles interact with "sound waves" (phonons) in the early universe, specifically when those sound waves travel very slowly (a "small sound speed").

They found that these heavy particles leave a unique fingerprint on the universe's patterns (the "bispectrum").

• The Old Expectation: If you just added heavy particles to the standard theory, you would expect the patterns to look like simple, local bumps (like adding a stone to a pond).
• The New Discovery: The heavy particles create strange, non-local shapes. Specifically, they found new shapes around the "equilateral" configuration (where the three points of the pattern are equal).

These shapes are so unique that you cannot fake them by just adding simple, local rules to the theory. You have to include "non-local" rules—rules that say, "What happens here depends on what happened over there."

Why This Matters (According to the Paper)

1. It connects two worlds: It bridges the gap between the complex, curved universe of the Big Bang and the simple, flat universe we use for standard particle physics.
2. It keeps the mass: Unlike previous methods, this one remembers that the particles were heavy. This is crucial because the mass of a particle often holds the key to its identity.
3. It finds new signals: It predicts specific patterns in the cosmic microwave background (the afterglow of the Big Bang) that look like "low-speed collider signals." These are distinct from the "cosmological collider" signals scientists have been looking for before.

Summary in One Sentence

The authors invented a new mathematical "lens" that lets us view the complex, expanding early universe as a simple, flat space, but with a special filter that keeps the heavy particles heavy, revealing unique new patterns that could help us identify heavy, invisible particles from the birth of the cosmos.

Technical Summary

Technical Summary: The Massive Flat Space Limit of Cosmological Correlators

Problem Statement
Cosmological correlators, which describe the statistical properties of quantum fields at the end of inflation, are central to understanding the microscopic details of the early universe. While the "amplitude limit" of these correlators—where the total energy $k_T \to 0$—has been successfully utilized to extract flat-space scattering amplitudes, this limit is inherently "massless." In this regime, the masses of internal propagators become negligible compared to the kinetic energies, causing massive exchange processes to degenerate into massless ones. Consequently, the rich analytic structure of massive Feynman integrals is lost, and the resulting constraints on cosmological bootstrap programs are limited. Furthermore, existing methods struggle to compute exact results for massive loop diagrams beyond specific soft limits (e.g., squeezed or collapsed limits) due to mathematical complexity. The authors identify a gap in the ability to extract information about finite internal masses from cosmological correlators in a flat-space limit.

Methodology
The authors propose a novel Massive Flat-Space (MFS) limit applicable to "heavy graphs," defined as diagrams with light external legs (e.g., massless or conformally coupled fields) and heavy internal lines (mass $m \gg H$). The methodology relies on a double-scaling limit:

1. The internal mass $m$ is sent to infinity ($m/H \to \infty$).
2. The external energies $\omega_i$ (introduced as off-shell variables independent of momenta $k_i$) are sent to zero.
3. Crucially, these limits are taken such that the product $\omega_i \times m$ remains finite.

To derive the behavior of correlators in this limit, the authors employ three complementary approaches:

• Direct Integration: They analyze the in-in (Schwinger-Keldysh) time integrals for tree-level exchange and one-loop bubble diagrams. Using the WKB approximation for massive mode functions in the large mass limit, they demonstrate that the integrals are dominated by contributions near the diagonal of the time variables ($\eta_1 \approx \eta_2$). This effectively collapses the internal structure of the diagram into a single vertex.
• Effective Action Formalism: By integrating out the heavy field in the curved spacetime path integral, they show that the non-time-ordered propagators are exponentially suppressed. The resulting effective action becomes local in time but spatially non-local, allowing the correlator to be expressed in terms of flat-space amputated diagrams with red-shifted momenta.
• Spectral Decomposition: For loop diagrams, they utilize the Källén-Lehmann spectral representation, showing that in the MFS limit, the de Sitter spectral density converges to its Minkowski counterpart, weighted by the massive propagator structure.

Key Contributions

1. The MFS Reduction Formula: The central result is a general reduction formula (Eq. 2.20) that expresses an $n$-point cosmological correlator in the MFS limit as a contact diagram in de Sitter space. The vertex of this contact diagram is determined by an amputated Feynman graph in flat space ($G_n$), computed with massive propagators and time-dependent, red-shifted external momenta $p_i^\mu(\eta) = (0, \mathbf{k}_i/a(\eta))$.
   $$F_n \xrightarrow{MFS} 2 \text{Re} \int_{-\infty}^0 d\eta \, a^4(\eta) \, G_n[p_i(\eta); m] \prod_{i=1}^n \mathcal{O}_i K_+(\omega_i, \eta)$$
   Unlike the standard amplitude limit, the internal propagators in $G_n$ retain their finite mass $m$.

2. Off-Shell Extension: The authors introduce a systematic off-shell extension of cosmological correlators by replacing bulk-to-boundary propagators $K(k_i, \eta)$ with $K(\omega_i, \eta)$, where $\omega_i$ are independent energy variables. This allows the definition of the MFS limit as a specific trajectory in the space of these variables.

3. Application to Inflationary Phenomenology: The framework is applied to the Cosmological Phonon Collider, specifically computing tree-level and one-loop contributions to the bispectrum of curvature perturbations in the regime of small sound speed ($c_s \ll 1$). The authors consider heavy scalars with masses in the window $H \ll m \lesssim \mathcal{O}(1) H/c_s$.

Results

• Analytical Expressions: The authors derive explicit analytical expressions for the bispectrum contributions from heavy scalar exchanges (SE1, SE2) and one-loop bubbles (SB1, SB2). These results are expressed in terms of Mellin transforms and series expansions involving polygamma functions.
• Novel Bispectrum Shapes:
  • Tree-Level: The results reproduce the known "low-speed collider signal," characterized by resonances in the mildly squeezed limit when $\alpha = c_s m/H \ll 1$.
  • One-Loop: The one-loop contributions reveal novel shapes that cannot be reproduced by adding local operators to the Effective Field Theory (EFT) of single-field inflation. Specifically, the SB2 diagram exhibits a local minimum or inflection point near the equilateral configuration, transitioning to a maximum as parameters vary.
• Non-Locality: The paper demonstrates that these one-loop signals correspond to spatially non-local operators in the EFT of inflation. While the effective action remains local in time, the spatial dependence is non-analytic in momentum space, distinguishing these signals from standard local EFT predictions.

Significance and Claims
The paper claims that the MFS limit provides a powerful new tool for the cosmological bootstrap program by accessing the analytic structure of massive Feynman integrals, which are otherwise obscured in the standard massless amplitude limit. By preserving the internal mass in the flat-space limit, the authors provide a more stringent constraint on the form of cosmological correlators.

Phenomenologically, the work highlights that massive fields in the intermediate mass range ($H \ll m \lesssim H/c_s$) leave distinct non-Gaussian imprints on inflationary correlators. These imprints, particularly at the one-loop level, are not captured by standard local EFT operators and require the inclusion of prescribed non-local operators. The authors position this as a method to probe heavy particle physics during inflation that goes beyond the traditional "cosmological collider" oscillations found in the squeezed limit, offering new observational templates for future CMB and large-scale structure surveys. The work remains modest regarding on-shell limits, noting that particle production effects (exponentially suppressed in the MFS limit) may complicate a direct on-shell formulation, which is left for future investigation.