⚛️ general relativity

Massive Spinning Fields During Inflation: Feynman rules and correlator comparison

This paper demonstrates that while two distinct theoretical descriptions of massive spinning fields during inflation yield identical signatures in the inflaton bispectrum due to the exclusive propagation of helicity-zero modes, they become distinguishable at the level of the trispectrum and higher-point correlation functions.

Original authors: Trevor Cheung, David Stefanyszyn

Published 2026-10-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Trevor Cheung, David Stefanyszyn

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the earliest moments of our universe, a fraction of a second after the Big Bang, space itself underwent a period of frantic, exponential expansion known as inflation. This era was not merely a quiet stretching of space; it was a high-energy laboratory where the laws of physics were pushed to their limits. During this time, the universe was filled with a field called the inflaton, which drove the expansion, and it is believed that many other heavy particles, too massive to be created in any terrestrial particle accelerator, were born and died in the chaos. As these heavy particles decayed, they left subtle ripples in the fabric of space, imprinting their mass and spin onto the distribution of matter we see today. Astronomers study these patterns, known as cosmological correlators, to reverse-engineer the physics of that ancient epoch. The challenge lies in the fact that we cannot observe these heavy particles directly; we can only infer their existence from the statistical shapes of the cosmic microwave background and the large-scale structure of galaxies.

A central puzzle in this field concerns how we mathematically describe these heavy, spinning particles during inflation. Because the universe was expanding rapidly, the usual symmetries of space and time were broken, leading physicists to propose two different ways to model these particles. One approach, often called the "cosmological collider" view, treats the particles as if they still respect the full symmetry of the expanding universe, even though the background is breaking it. The other, known as the "cosmological condensed matter" view, treats the particles more like vibrations in a solid, respecting only the rotational symmetry of the space at any given moment. For years, it was unclear whether these two mathematical descriptions were just different languages for the same physical reality or if they predicted genuinely different signatures that could be distinguished by future telescopes. If they were different, it would mean our current methods of searching for new physics might be missing crucial details. If they were the same, it would mean we could use the simpler mathematical framework to understand the more complex one, streamlining the search for new particles.

In a recent study, researchers set out to settle this question by rigorously comparing the predictions of these two models. They focused on the "bispectrum," a specific statistical measure that describes how three points in the universe's matter distribution are correlated, which is the first place where the influence of heavy particles can be seen. To do this, they had to first rewrite the rules of the game. They discovered that the standard way of calculating how these particles move and interact was incomplete. When using the more complex "cosmological collider" description, the particles sometimes behave in a way that requires an instantaneous connection, a mathematical feature that had been overlooked. This connection acts like a direct link that bypasses the usual flow of time, appearing as a sharp, instantaneous spike in the calculations. The researchers derived a new set of rules, or "Feynman rules," that included these instantaneous links, ensuring that the calculations for the complex model were accurate.

With these corrected rules in hand, the team compared the results of the two models. They found a surprising and reassuring result for the simplest case: the three-point correlation, or bispectrum. Despite the different starting assumptions and the complex instantaneous links in one model, the final predictions for the shape of the bispectrum were identical. The researchers showed that any signal produced by the complex model could be perfectly mimicked by the simpler model, provided that the simpler model included a few extra, direct interactions between the inflaton field itself. This means that for the specific signatures currently being hunted in the cosmic microwave background data, the two descriptions are indistinguishable. The universe, in this specific context, does not care which mathematical language we use to describe the heavy particles; the observable outcome is the same.

However, the story changes when the researchers looked at more complex patterns, specifically the "trispectrum," which involves four points, and other higher-order correlations. Here, the two models diverged. The complex model allowed for the exchange of multiple types of particle spins simultaneously in a way that the simpler model could not replicate without breaking the rules of locality, which would require interactions to happen across vast distances instantly. In these higher-order cases, the two descriptions produce distinct signatures that cannot be made to look the same. This distinction is crucial because it tells us that while the simpler model is sufficient for the current generation of searches focused on the bispectrum, future observations that can measure more complex patterns will need to be careful about which theoretical framework they use.

The study concludes that the "cosmological collider" signals currently being searched for in the data are universal. They do not depend on the intricate details of how we choose to describe the heavy spinning fields. This is a significant finding because the simpler mathematical framework is much easier to work with, allowing scientists to perform calculations more efficiently without losing accuracy for the signals we can currently detect. The researchers suggest that this universality holds true even when the interactions are constrained by the specific symmetries of the early universe. While the two descriptions are distinct in their deeper mathematical structure, they converge on the same observable reality for the most accessible signals. This convergence offers a powerful shortcut for cosmologists, allowing them to use the simpler tools to interpret the complex history of the infant universe, confident that they are not missing hidden physics in the three-point correlations that define the current frontier of observation.

Technical Summary: Massive Spinning Fields During Inflation: Feynman rules and correlator comparison

Problem Statement

The paper addresses a fundamental question in cosmological collider physics: do two distinct theoretical descriptions of massive spinning fields during inflation yield distinguishable signatures in cosmological observables?

During inflation, the background value of the inflaton breaks de Sitter boost symmetry. Consequently, there are two admissible frameworks for describing fluctuations of massive spinning fields:

Cosmological Collider (CC) Physics: The free theory of the spinning field is assumed to be de Sitter invariant (representations of the full $SO(1,4)$ group). Interactions with the inflaton break boosts. This approach is constrained by the Higuchi bound and requires unit sound speed.
Cosmological Condensed Matter (CCM) Physics: The free theory is not assumed to be de Sitter invariant; instead, fields are treated as irreducible representations of the unbroken $SO(3)$ symmetry of FRW cosmologies. This approach allows for reduced sound speeds and avoids the Higuchi bound.

The central question is whether the correlators (specifically the bispectrum and higher-point functions) generated in the CC framework can be mimicked by those in the CCM framework using local interactions and the same number of degrees of freedom. If they are indistinguishable, the simpler CCM framework could be used universally to predict cosmological collider signals.

Methodology

The authors systematically derive and compare the Feynman rules for both set-ups within the Poincaré patch of de Sitter space.

1. Derivation of Feynman Rules for CC

The authors develop a procedure to compute wavefunction coefficients and correlators for the CC set-up, where non-dynamical modes are required to form full de Sitter multiplets.

Decomposition: Fields are decomposed into helicity modes. For a spin- $S$ field $\Phi_{\mu_1...\mu_S}$ , the dynamical modes are contained in the traceless spatial components, while temporal and mixed components are non-dynamical.
Integrating Out Non-Dynamical Modes: Instead of constructing propagators from mode functions without solving for constraints, the authors solve the equations of motion for non-dynamical modes (e.g., $\Phi_0$ for spin-1) in terms of dynamical modes and the inflaton.
Effective Propagators: A key discovery is that the standard Schwinger-Keldysh propagators are insufficient. The correct effective propagators for non-dynamical modes (and mixed propagators connecting different components with the same helicity) must include delta function terms corresponding to instantaneous propagation.
- For spin-1, the effective propagator for the temporal mode $\Phi_0$ acquires a correction:
  $(G^{0}_{0,1 \to 0,1})_{\text{eff}} = G^{0}_{0,1} + \frac{i}{m^2 a^2} \delta(\eta - \eta')$
- Mixed propagators (e.g., connecting $\Phi_0$ and $\Phi_i$ ) are also derived, which are necessary because the creation/annihilation operators for helicity-zero modes in different components are identical in canonical quantization.
Spin-2 Extension: The procedure is extended to massive spin-2 fields, revealing a more complex structure of effective propagators involving derivatives of delta functions due to second-order differential operators relating the modes.

2. Conversion to Correlators

Using the Born rule, the authors convert the derived wavefunction coefficients into cosmological correlators (bispectrum, trispectrum). They demonstrate that the boundary terms arising from the wavefunction definition cancel out in the correlator calculation, leaving the delta function corrections in the propagators as the primary modification to the standard Feynman rules.

3. Correlator Comparison

The authors compare the resulting correlators from CC and CCM set-ups:

Single-Helicity Exchange: They analyze processes where only the helicity-zero ( $h=0$ ) mode is exchanged. This occurs in the tree-level bispectrum due to the linear-mixing vertex $\pi \Phi$ , which, by rotational invariance, decouples non-zero helicity modes.
Multi-Helicity Exchange: They analyze processes involving the exchange of multiple helicities (e.g., the trispectrum), where different helicity modes contribute simultaneously.

Key Contributions and Results

1. Novel Feynman Rules for CC Physics

The paper provides the first systematic derivation of Feynman rules for massive spinning fields in the CC set-up. The rules differ from standard expectations in two critical ways:

Mixed Propagators: Vertices involving different components of the spinning field (e.g., temporal vs. spatial) but the same helicity are connected via mixed propagators.
Delta Function Corrections: Propagators involving non-dynamical modes contain delta function terms ( $\delta(\eta - \eta')$ ). These terms generate additional contact diagrams that are of the same order in couplings as the exchange diagrams. These contact terms are essential for obtaining the correct perturbative results and are analogous to non-covariant terms found in flat-space canonical quantization of massive vector fields.

2. Indistinguishability of the Bispectrum

The authors prove that the bispectrum (three-point function) computed in the CC set-up can be exactly mimicked by a sum of local diagrams in the CCM set-up.

Mechanism: The delta function terms in the CC propagators map to local contact interactions in the CCM description. The momentum factors arising from the mixing of components in CC map to spatial derivatives in CCM.
Universality: This correspondence holds even when interactions are constrained by the symmetries of the Effective Field Theory of Inflation (EFToI).
Implication: Since the bispectrum is the primary observable currently searched for in CMB and LSS data to detect massive spinning fields, the cosmological collider signals in the bispectrum are universal. They do not depend on whether the underlying theory is de Sitter invariant (CC) or only rotationally invariant (CCM).

3. Distinguishability of Higher-Point Correlators

The correspondence breaks down for the trispectrum (four-point function) and other higher-point correlators.

Reason: Higher-point functions can involve the exchange of multiple helicities (e.g., $h=0$ and $h=\pm 1$ simultaneously). In the CC set-up, these modes are part of a single de Sitter multiplet. In the CCM set-up, they are independent fields with potentially different sound speeds.
Non-Locality: To mimic the CC trispectrum using CCM fields, one would need to construct non-local interactions (e.g., projecting onto transverse modes using $\nabla^{-2}$ operators) to decouple the helicity modes. Since the authors restrict the comparison to local interactions, the CC and CCM set-ups yield distinct signatures for higher-point functions.

4. Consistency with EFToI

The authors verify that the mapping between CC and CCM interactions preserves the symmetries of the Effective Field Theory of Inflation. The contact terms required in the CCM description to mimic CC exchanges can be constructed using standard EFToI building blocks (e.g., $\delta g_{00}$ , extrinsic curvature).

Significance and Claims

The paper claims that the search for cosmological collider signals in the bispectrum is robust against the specific theoretical description of the massive spinning field.

Practical Utility: Because the CCM framework is computationally simpler (avoiding the need for mixed propagators and delta function corrections), researchers can safely use CCM calculations to predict bispectrum signatures for any massive spinning field, regardless of whether the underlying theory respects full de Sitter invariance.
Limitations: The paper modestly claims that this universality does not extend to higher-point correlators. If future observations detect non-Gaussianities in the trispectrum or higher, the specific shape of these signals could potentially distinguish between the CC and CCM descriptions, provided the interactions are local.
Theoretical Insight: The work clarifies the role of non-dynamical modes in inflationary perturbation theory, showing that their integration leads to instantaneous propagation terms (delta functions) that are crucial for consistency, mirroring phenomena in flat-space canonical quantization but with distinct features due to the expanding background.

The authors conclude that while the CC and CCM set-ups are distinct at the level of the Lagrangian and higher-point functions, they are effectively indistinguishable at the level of the inflaton bispectrum, the primary target of current observational searches.