⚛️ general relativity

Explicit gauge-invariant variables in multifield inflation beyond linear order and Hamiltonian dynamics

This paper presents the first explicit construction of large-scale, gauge-invariant phase-space variables and the corresponding quadratic and cubic Hamiltonians for multifield inflation at second-order perturbation theory, demonstrating their consistency with standard gauge-fixed results and providing a framework for extending these calculations to higher orders.

Original authors: Julien Grain, Hugo Holland, Lucas Pinol

Published 2026-08-31

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Julien Grain, Hugo Holland, Lucas Pinol

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 story of our universe's beginning is written in the faint whispers of light that have traveled across the cosmos since its earliest moments. To understand these whispers, cosmologists rely on a theory called inflation, which suggests that the universe expanded at a mind-boggling speed in the fraction of a second after the Big Bang. This rapid stretching smoothed out the cosmos and planted the seeds for all the galaxies we see today. However, the physics behind this expansion remains a mystery. Scientists suspect that invisible fields, rather than the particles we know, drove this process. When there is more than one of these fields involved, the mathematics becomes incredibly tangled. A major hurdle in studying them is a problem of perspective: the equations used to describe the universe change depending on how you choose to measure time and space. This is not a physical difference, but a mathematical one, like describing a mountain's height from sea level versus from the base of the valley. If you do not account for this carefully, you might think you are seeing a new physical effect when you are actually just looking at the same thing from a different angle.

For decades, researchers have known how to handle this perspective problem when the ripples in the early universe were very small. They developed a set of tools to strip away the confusing angles and find the true, physical signals. But as the universe evolved, these ripples grew, and the simple tools were no longer enough. The interactions between the fields became complex, and the old methods of ignoring the perspective issue or fixing it in a single, rigid way began to break down. This left a gap in our understanding: we could not be sure if the complex patterns we were trying to predict were real physical features or just artifacts of the mathematical lens we were using. Without a way to see clearly through this lens, it is difficult to test which theories of the early universe are correct and which are not.

In this work, a team of physicists has built a new, robust framework to solve this problem for the first time when the ripples are large and the interactions are strong. They approached the universe not as a flowing river of space and time, but as a system of moving parts, similar to how a mechanic might look at the gears and springs of a clock. By using a specific mathematical language that tracks the energy and motion of these parts, they were able to construct a set of variables that remain unchanged no matter how you shift your perspective. These variables act as a universal standard, allowing scientists to describe the universe's behavior without getting lost in the confusion of different measurement choices. The researchers showed that they could build these stable variables even when the fields interact in complicated, non-linear ways, which happens when the universe is expanding rapidly.

The team demonstrated that their new method works by applying it to the most common scenario: a universe filled with multiple fields. They proved that their new variables are mathematically identical to the standard measurements used in previous studies, but with a crucial difference: they are guaranteed to be free of perspective errors. This means that when scientists use these new tools to calculate what the universe should look like today, they can be confident that the results are real. The researchers also showed that their approach can be extended to even more complex interactions, providing a clear path forward for studying the most extreme moments of the early universe. By establishing a way to see the universe's behavior without the distortion of mathematical perspective, this work provides a solid foundation for testing the deepest theories of how our cosmos began. It turns a confusing tangle of equations into a clear, reliable map, ensuring that when we look at the data from the early universe, we are seeing the truth of the cosmos, not just the reflection of our own mathematical choices.

Technical Summary: Explicit Gauge-Invariant Variables in Multifield Inflation Beyond Linear Order and Hamiltonian Dynamics

Problem Statement
General relativity coupled to multiple scalar fields constitutes a diffeomorphism-invariant constrained system. In the context of cosmological perturbation theory, a standard counting of perturbative degrees of freedom inevitably overestimates the number of physical modes because gauge redundancies and constraint equations eliminate non-dynamical variables. While this issue is well-resolved for linear fluctuations (via the Mukhanov-Sasaki variables), the construction of explicit, large-scale gauge-invariant phase-space variables at second order (and beyond) in multifield inflation has remained incomplete, particularly within a Hamiltonian framework. Most existing non-linear calculations rely on fixing a specific gauge (e.g., flat or comoving) without formally proving that the procedure removes all gauge redundancies or that the resulting predictions are robustly connected to gauge-invariant observables.

Methodology
The authors employ a covariant Hamiltonian formalism to address these issues. The methodology proceeds through several key steps:

Hamiltonian Formulation: The authors derive the Hamiltonian for general relativity coupled to $N$ scalar fields with a curved field space (non-linear sigma models) up to the third order in perturbation theory. They utilize the ADM formalism, introducing lapse ( $N$ ) and shift ( $N^i$ ) functions, and define covariant perturbations ( $Q^I, P_I$ ) to ensure the expressions are explicitly covariant in the field space, avoiding spurious Christoffel symbols that arise from standard expansions.
Gauge Transformation Analysis: The paper systematically derives the gauge transformations of configuration and momentum variables up to second order using Lie derivatives. This allows for the identification of how perturbative variables change under coordinate transformations.
Construction of Gauge-Invariant Variables:
- Linear Order: The authors recover the standard multifield Mukhanov-Sasaki variables, demonstrating they are gauge-invariant at linear order.
- Second Order (Large Scales): Restricting the analysis to super-Hubble scales (where spatial gradients are negligible), the authors explicitly construct a finite-dimensional basis of quadratic corrections. They solve for the specific combinations of field perturbations ( $Q^I, P_I$ ) and gravitational perturbations ( $\delta\gamma_1, \delta\pi_1$ ) that render the variables invariant under second-order gauge transformations.
- Canonical Structure: They verify that these corrected variables form canonical pairs (i.e., their Poisson brackets satisfy the canonical relations up to the relevant order).
Gauge Fixing and Consistency Checks: The authors demonstrate how to fix the gauge (specifically the flat gauge) within the derived Hamiltonian. They show that the large-scale limit of the gauge-fixed Hamiltonian for the corrected variables matches the results found in the literature (which typically use a Lagrangian approach), providing a robust consistency check.
Formal Proof: A formal algebraic proof is provided for the existence of gauge-invariant variables at quadratic order. This involves a phase-space parametrization that separates physical degrees of freedom from gauge and constraint degrees of freedom, generalizing the approach to multiple fields and outlining the procedure for higher orders.

Key Contributions and Results

Explicit Second-Order Variables: The paper presents the first explicit calculation of all large-scale gauge-invariant phase-space variables in multifield inflation at second order. These variables are constructed by correcting the standard Mukhanov-Sasaki variables with specific quadratic terms involving both field and gravitational perturbations.
Recovery of Standard Gauges: The authors prove that their constructed gauge-invariant variables reduce exactly to the standard field fluctuations in the flat gauge and the curvature perturbation in the comoving gauge. This establishes a rigorous link between gauge-fixed calculations and gauge-invariant observables.
Cubic Hamiltonian: The authors derive the cubic Hamiltonian for multifield inflation in a gauge-independent manner. They show that upon gauge fixing, this Hamiltonian reproduces the known cubic interactions in the flat gauge, validating the Hamiltonian approach against established Lagrangian results.
Canonical Consistency: A significant result is the demonstration that the corrected configuration variables and their conjugate momenta remain a canonical pair (up to second-order corrections), ensuring the consistency of the phase-space description.
Formal Existence Proof: Section 5 provides a formal construction proving that gauge-invariant variables exist at quadratic order for any number of fields. The authors outline a systematic procedure (using a specific basis in phase space) that can be extended to third order and beyond.

Significance
The paper claims to resolve the ambiguity surrounding gauge-fixing procedures in non-linear cosmological perturbation theory. By explicitly constructing gauge-invariant variables that retain the desirable properties of standard variables (such as reducing to flat-gauge fluctuations), the work provides a robust framework for connecting theoretical predictions above the horizon to cosmological observations.

The authors emphasize that while their explicit solutions are restricted to super-Hubble scales, this is sufficient for connecting to observations, as physical wavelengths eventually stretch above the horizon. The work offers a "robust consistency check" for the gauge-fixed procedures commonly used in the literature. Furthermore, the derivation of the gauge-independent cubic Hamiltonian is presented as a necessary ingredient for future calculations of loop corrections to the primordial power spectrum and bispectrum, as well as for applications in stochastic inflation where non-Gaussian noise terms are relevant. The paper concludes by suggesting that the algebraic structure identified could lead to a more systematic route via canonical transformations to map gauge-independent Hamiltonians directly to gauge-fixed ones in future work.

More like this