⚛️ general relativity

Reconstructing cosmological correlators via dispersion: from cutting to dressing rules

This paper demonstrates that tree-level cosmological correlators in de Sitter space can be systematically reconstructed from their discontinuities using momentum-space dispersion relations and cutting rules, thereby deriving dressing rules that connect these correlators to flat-space Feynman diagrams while emphasizing the fundamental role of unitarity.

Original authors: Shibam Das, Debanjan Karan, Babli Khatun, Nilay Kundu

Published 2026-02-06

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Shibam Das, Debanjan Karan, Babli Khatun, Nilay Kundu

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

The Big Picture: Rebuilding a Broken Vase

Imagine the universe during its earliest moments (inflation) as a giant, expanding balloon. Physicists want to know how different parts of this balloon "talked" to each other back then. They call these conversations cosmological correlators.

Usually, calculating these conversations is like trying to rebuild a shattered vase by watching a slow-motion video of it breaking, frame by frame, while the pieces are flying through a hurricane. It's messy, complicated, and requires doing thousands of difficult math integrals (summing up tiny pieces of time).

This paper proposes a smarter way. Instead of watching the whole video, the authors say: "Let's just look at the cracks."

The Core Idea: The "Cutting" and "Dispersion" Trick

The authors use a mathematical tool called dispersion, which is like a reconstruction kit. The logic goes like this:

The Cut (The Crack): Imagine you have a complex Lego structure (a cosmological diagram). If you snap one of the connecting bricks (an internal line), the structure falls apart into two smaller, simpler pieces. In physics, this "snapping" is called a cut.
The Discontinuity (The Gap): When you snap that brick, there is a sudden jump or "discontinuity" in the math. The authors realized that if you know exactly what happens at the moment of the snap (the gap), you can mathematically glue the pieces back together.
The Reconstruction (Dispersion): They use a formula (the dispersion formula) that takes these "gaps" and integrates them (adds them up) to rebuild the entire original structure.

The Analogy: Think of a complex song. Instead of trying to write the whole song from scratch, you listen to the silence between the notes (the gaps). The authors found that if you know the pattern of the silence, you can mathematically reconstruct the entire melody without ever needing to hear the full song first.

The Step-by-Step Process

The paper breaks this down into a logical flow:

Step 1: Start Small. They first looked at simple structures (2-site and 3-site diagrams). They showed that if you cut the internal lines one by one, you eventually break the whole thing down into the simplest possible building blocks (1-site objects).
Step 2: The Chain Reaction. For a huge, complex diagram with many interaction points (vertices), they don't cut everything at once. They cut one line, then another, then another, successively. This turns a giant, scary problem into a chain of tiny, easy problems.
Step 3: The Magic Formula. Once they have all these tiny "gaps" (discontinuities), they apply the dispersion formula. This formula acts like a 3D printer, taking the data from the gaps and printing out the full, complex cosmological correlator.

Key Result: They proved that you can rebuild the entire history of the universe's interactions using only the data from the simplest building blocks and the rules of how they break apart. You don't need to do the hard time-integrals anymore.

The "Dressing Rules": A New Uniform

The paper also rediscovered a set of instructions called dressing rules.

The Old Way: Previously, to get the answer for our expanding universe (De Sitter space), physicists had to take a standard flat-space diagram (like a drawing of particles in a vacuum) and manually apply a complex, theory-specific "costume" or "dressing" to it. It was like having to sew a custom suit for every single type of particle interaction.
The New Insight: The authors showed that these "dressing rules" aren't magic; they naturally emerge from the "cutting and gluing" method they developed.
The Generalization: They created a universal rulebook. Instead of sewing a new suit for every theory, they gave a single set of instructions (color-coding vertices blue or red) that works for any scalar field theory.
- Blue Vertices: Represent standard interactions.
- Red Vertices: Represent "auxiliary" or imaginary parts of the interaction.
- The Rule: You draw the diagram, color the vertices in different combinations (ensuring an even number of red ones), and apply specific mathematical "dressings" (integrals) to the lines.

Why This Matters (According to the Paper)

Simplicity: It bypasses the tedious, time-consuming integrals that usually plague cosmological calculations.
Universality: Their "dressing rules" work for a wide variety of theories (massless particles, conformally coupled particles, different interaction strengths) without needing to re-derive the math for each one.
Connection: It bridges the gap between the "cutting rules" (which rely on the principle of unitarity, or conservation of probability) and the "dressing rules" (which look like a recipe for modifying flat-space diagrams). It shows that the recipe isn't arbitrary; it's a direct consequence of how the universe breaks and rebuilds itself mathematically.

Summary

In short, the authors found a shortcut. Instead of calculating the complex history of the universe by summing up every moment in time, they showed you can calculate it by looking at the "cracks" (discontinuities) in the math, breaking the problem down into tiny pieces, and using a universal set of "dressing" instructions to stitch it all back together. They turned a difficult, time-consuming puzzle into a systematic, rule-based assembly line.

Technical Summary: Reconstructing Cosmological Correlators via Dispersion: from Cutting to Dressing Rules

Problem Statement
The computation of cosmological correlators—equal-time correlation functions of quantum fields on the late-time slice of de Sitter space—is traditionally performed using the Schwinger-Keldysh (in-in) formalism or wave-function methods. While these approaches are principled, they become technically intractable for diagrams with many external legs, multiple interaction vertices, or loop structures due to the complexity of time-ordered integrals in expanding spacetime.

Recent developments in the "cosmological bootstrap" program have sought to reconstruct correlators using fundamental principles like unitarity and analyticity, often borrowing intuition from flat-space scattering amplitudes. Specifically, it was established that wave-function coefficients could be reconstructed from their discontinuities using momentum-space dispersion formulas. However, a direct reconstruction of the correlators themselves (rather than the wave-function coefficients) via dispersion relations had not been fully formulated until recently. Furthermore, while "dressing rules" exist to map flat-space Feynman diagrams to de Sitter correlators, their derivation via the shadow formalism is theory-dependent and lacks a direct connection to the fundamental unitarity principles (cutting rules) governing the correlators.

Methodology
The authors propose a framework to reconstruct tree-level cosmological correlators for conformally coupled and massless polynomial scalar interactions (both IR-divergent and IR-convergent) by applying momentum-space dispersion formulas directly to the discontinuities of the correlators.

Discontinuity Relations: The work builds upon recent single-cut discontinuity relations for cosmological correlators (specifically those derived in [1]). These relations express the discontinuity of an $r$ -site correlator (obtained by cutting an internal line) as a sum of products of lower-point data. Crucially, this lower-point data includes not only standard correlators ( $B^{(1)}$ ) but also "auxiliary counterparts" ( $\tilde{B}^{(1)}$ ), which involve the imaginary parts of wave-function coefficients.
Successive Cutting: For diagrams with multiple vertices ( $r$ -site), the authors apply the single-cut rule successively to all internal lines. This decomposes the full correlator's discontinuity into a product of discontinuities of elementary 1-site objects (contact terms).
Dispersion Integration: Once the successive discontinuity is expressed in terms of 1-site data, the authors apply the dispersion formula iteratively. By integrating over the internal momenta, they reconstruct the full correlator from these discontinuities, up to contact-term ambiguities (which the dispersion method inherently cannot capture).
Inductive Generalization: The authors formulate a general algorithm for an arbitrary $r$ -site correlator using mathematical induction. They demonstrate that the successive discontinuity of an $r$ -site correlator can be systematically constructed from the discontinuities of 1-site objects, respecting specific symmetry and sign rules.
Derivation of Dressing Rules: By analyzing the structure of the reconstructed integrals, the authors derive a set of "diagrammatic dressing rules." These rules provide a direct prescription for obtaining late-time de Sitter correlators from flat-space Feynman diagrams without invoking the shadow formalism.

Key Contributions and Results

Direct Correlator Reconstruction: The paper demonstrates that cosmological correlators can be reconstructed directly from their discontinuities using dispersion relations, bypassing the intermediate step of reconstructing wave-function coefficients. This is achieved by utilizing the single-cut relations for correlators derived in [1].
Generalized Inductive Algorithm: The authors provide a rigorous inductive proof for the successive single-cut discontinuity of an arbitrary $r$ -site tree-level correlator. They show that the full discontinuity decomposes into a specific linear combination of products of 1-site correlators ( $B^{(1)}$ ) and auxiliary objects ( $\tilde{B}^{(1)}$ ), governed by the parity of the number of auxiliary insertions.
Derivation of Dressing Rules: A primary result is the derivation of a generalized set of diagrammatic dressing rules. These rules allow one to write the late-time de Sitter correlator as an integral relation acting on the flat-space Feynman propagator.
- Rule Structure: The rules involve "color-coding" vertices of a flat-space Feynman diagram (blue or red) with the constraint that the number of red vertices must be even.
- Vertex Contributions: Blue vertices contribute terms involving the real part of the wave-function data ( $Disc B^{(1)}$ ), while red vertices contribute terms involving the imaginary part ( $\tilde{Disc} \tilde{B}^{(1)}$ ).
- Internal Lines: Internal lines are "dressed" by integrating over their energy with a specific kernel involving the de Sitter bulk-to-bulk propagator pole structure.
Explicit Checks: The authors explicitly verify their generalized dressing rules against known results for conformally coupled $\lambda\phi^4$ and $\lambda\phi^3$ theories, as well as massless $\lambda\chi^3$ theories. They also reconstruct the 3-site correlator in $\lambda\phi^5$ theory, confirming that the method reproduces standard in-in calculation results.

Significance and Claims
The paper claims to provide an alternative, principle-based perspective on the "miraculous" workings of the dressing rules previously derived in [59] via the shadow formalism.

Unification of Principles: The authors emphasize that their derivation stems directly from basic principles—specifically unitarity (via cutting rules) and analyticity (via dispersion)—rather than the more abstract shadow formalism. This establishes a clearer conceptual link between cosmological cutting rules and the resulting diagrammatic prescriptions.
Generalization: Unlike previous dressing rules which required theory-specific construction of auxiliary propagators, the authors propose a generalized version applicable to all scalar theories with $\phi^n$ self-interactions. The details of the interaction are encoded directly into the rules themselves.
Limitations: The authors modestly note that because the method relies on dispersion integration over discontinuities, it cannot reconstruct the "contact pieces" (analytic terms) of the correlator. These must be added separately or determined by other means.

In summary, the work bridges the gap between the analytic structure of cosmological correlators (discontinuities) and practical computational tools (dressing rules), offering a unified framework to reconstruct de Sitter observables from flat-space data using fundamental quantum mechanical principles.