⚛️ general relativity

Classical soft theorems meet the post-Newtonian expansion

This paper establishes a direct correspondence between gravitational memory effects, their higher-order tail corrections, and the soft expansion by analyzing the infrared logarithmic structure of gravitational waveforms from compact binaries within the multipolar post-Newtonian framework, demonstrating consistency with Sahoo-Sen logarithmic soft theorems.

Original authors: Samim Akhtar, Riccardo Sturani

Published 2026-09-25

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Samim Akhtar, Riccardo Sturani

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

Gravity is not merely a force that pulls objects together; it is a ripple in the fabric of space and time itself. When massive objects like black holes or neutron stars orbit each other, they stir this fabric, sending out waves that travel across the universe at the speed of light. For decades, scientists have been learning to listen to these waves, using them to understand the violent collisions that create them. However, to hear the full story, one must understand not just the main crash, but also the lingering echoes that follow. These echoes are known as "memory" effects. Unlike the rhythmic chirp of an orbiting pair, memory is a permanent change: after a gravitational wave passes, the space between two floating objects does not return to exactly where it was. It shifts, just a tiny bit, forever. This shift is the universe's way of remembering that energy was carried away.

For a long time, physicists understood these memory effects well in the context of objects flying past one another and scattering apart. But a more complex and common scenario involves objects that are locked in a gravitational embrace, orbiting each other until they finally merge. In these bound systems, the rules of the game are different. The objects never reach a state of free flight, which changes how the memory of their interaction is written into the waves. A new study by researchers at the ICTP South American Institute for Fundamental Research and the Universidade Estadual Paulista has now bridged this gap. They have shown how the permanent shifts in space caused by orbiting binaries are mathematically connected to a deeper, universal principle governing how gravity behaves at very low frequencies. Their work reveals that the complex, lingering echoes of these cosmic dances are not random noise, but a precise, predictable pattern that follows a specific mathematical ladder.

The researchers focused on the "tail" of these gravitational waves. Imagine a sound wave traveling through a room; if the room has corners, the sound bounces off the walls and returns to the listener slightly later, creating a tail of sound that lingers after the original note. In the universe, gravitational waves do something similar. As they travel away from a binary system, they interact with the curved space created by the total mass of the system itself. This interaction causes the waves to scatter off the background gravity, creating a "tail" that modifies the signal. The team discovered that this tailing effect happens in layers. The first layer is the basic memory shift. The second layer is a "tail of the memory," where the memory effect itself gets scattered. The third layer is a "tail of the tail of the memory," and so on.

Using a sophisticated framework that treats gravity like a collection of interacting particles, the team calculated the strength of these effects for binary systems. They found that each additional layer of tailing adds a specific, predictable change to the signal. The first layer, which occurs at a specific stage of the orbital evolution known as 2.5 post-Newtonian order, produces a permanent shift that depends only on the total energy radiated away. The next layer, appearing at 4 post-Newtonian order, introduces a logarithmic growth, meaning the effect increases slowly but steadily as the frequency of the wave drops. The researchers showed that this pattern continues indefinitely. Each new layer of tailing adds another power of a logarithm to the signal, creating a tower of effects that grows in a very specific way.

What makes this discovery significant is that it connects two different ways of thinking about gravity. One way looks at the slow, steady orbit of binary stars, while the other looks at the fundamental rules of how particles scatter at the lowest possible energies. The study proves that these two perspectives are actually describing the same phenomenon. The complex, hereditary effects seen in orbiting binaries are the physical realization of a universal theorem about soft gravitons, which are the lowest-energy packets of gravitational waves. The researchers demonstrated that the infinite series of these tail effects can be summed up into a single, elegant mathematical expression, much like how a series of small steps can be described as a single smooth slope.

The team also looked at what happens at very high energies, where the details of the source matter more than the universal laws. They found that while the low-frequency effects follow a perfect, universal pattern, the high-frequency parts of the signal contain their own logarithmic features that depend on the specific details of the binary system. These high-energy features require a different kind of mathematical cleanup to make sense of them, but they do not disrupt the beautiful, universal pattern found in the low-frequency tails.

This work provides a complete map of how gravitational waves remember the history of the systems that created them. By showing that the memory effects in bound systems follow the same rules as the most fundamental scattering processes, the researchers have unified our understanding of gravitational radiation. They have confirmed that the universe keeps a precise record of every interaction, written in the permanent shifts of space and the logarithmic tails of the waves. This understanding is crucial for the next generation of gravitational wave detectors, which will be sensitive enough to hear these subtle, lingering echoes. With this new map, scientists can now predict exactly what these detectors should see, turning the faint whispers of the universe into a clear and readable story of cosmic history.

Technical Summary: Classical Soft Theorems meet the post-Newtonian expansion

Problem Statement
The paper addresses the theoretical gap between two established frameworks for describing gravitational radiation: the Post-Newtonian (PN) expansion, specifically within the Multipolar Post-Minkowskian (MPM) and Non-Relativistic General Relativity (NRGR) effective field theory approaches, and the modern understanding of classical soft theorems. While soft theorems govern the non-analytic (infrared) structure of scattering amplitudes at low frequencies—linking the $1/\omega$ pole to displacement memory and logarithmic terms to "tail" effects—the explicit realization of this hierarchy in gravitationally bound systems (compact binaries) has remained less explored than in hyperbolic scattering. In bound systems, particles do not reach free asymptotic states, eliminating trajectory-based logarithmic contributions found in scattering, leaving only backscattering (tail) effects as the source of infrared (IR) logarithms. The authors aim to bridge this gap by explicitly deriving the NRGR amplitudes for memory and tail-of-memory effects and demonstrating their direct correspondence to the Sahoo–Sen logarithmic soft theorems.

Methodology
The authors work within the NRGR framework in $d = 3 - \epsilon$ spatial dimensions, utilizing retarded propagators and a specific renormalization scale $\tilde{\mu}^2 = \pi\mu^2 e^{-\gamma}$ . The methodology involves:

Diagrammatic Computation: Calculating Feynman diagrams for hereditary processes. The memory effect is treated as a one-loop process, while tail-of-memory and higher-order tails are treated as multi-loop processes involving static mass insertions ( $M$ ) coupled via the $\sigma\sigma\phi$ vertex (where $\phi$ is the Newtonian potential).
Isolation of IR/UV Structure: The authors isolate the infrared divergences (poles in $1/\epsilon_{IR}$ ) and the associated physical logarithms ( $\log \omega$ ) arising from the integration over static propagators. They distinguish these from ultraviolet (UV) divergences arising from short-distance sensitivity.
Frequency Space Analysis: By transforming time-domain hereditary integrals (convolutions involving logarithmic kernels) into frequency space, the authors analyze the non-analytic behavior of the amplitudes as $\omega \to 0$ .
Exponentiation: The authors sum the series of amplitudes with $n$ tail insertions to demonstrate an exponentiation structure, comparing the result to the all-order predictions of soft theorems.

Key Contributions and Results

Memory Effect (2.5PN): The authors compute the nonlinear (Christodoulou) memory amplitude as a one-loop diagram with no static propagators. They confirm that the Transverse-Traceless (TT) radiative component is finite (no IR divergence) and yields only the Weinberg soft pole $\sim 1/\omega$ . This provides the PN realization of the leading soft coefficient $\tilde{h}^{(0)}_{\mu\nu}$ , corresponding to the displacement memory effect proportional to the total radiated energy.
Tail-of-Memory (4PN): The tail-of-memory is computed as a two-loop process involving one static propagator. This diagram generates a single IR pole ( $1/\epsilon_{IR}$ ) and a physical logarithmic term $\log \omega$ . The result reproduces the PN realization of the Sahoo–Sen coefficient $\tilde{h}^{(\log)}_{\mu\nu}$ , establishing that the "tail" of the memory is the backscattering of the memory radiation off the static background curvature.
Generalization to $n$ -Tail Insertions: The authors generalize the computation to $n$ static mass insertions (the $n$ -tail-of-memory), corresponding to $(2.5 + 1.5n)$ PN order. They derive the universal soft behavior for the amplitude $A_n$ in the limit $\omega \to 0$ :
$A_n \big|_{\omega \to 0} \sim (GM)^n \cdot E_{rad} \cdot \omega^{n-1} (\ln \omega)^n$
This result provides the first explicit PN/NRGR derivation of the all-order leading-logarithm tower conjectured in previous literature.
IR Exponentiation: By summing the series of amplitudes over all $n$ , the authors demonstrate that the IR logarithms exponentiate into a universal soft-dressed factor:
$\mathcal{A} \sim \frac{e^{i GM \omega L(\omega)}}{\omega} \cdot \tilde{Q}_{ij}$
where $L(\omega)$ contains the IR pole and logarithmic terms. This confirms the IR exponentiation structure of the radiation EFT and establishes a direct mapping between the PN hereditary hierarchy and the Sahoo–Sen soft expansion.
UV Logarithms: The paper notes that UV logarithms, arising from the short-distance sensitivity of the multipole expansion, first appear at $n \ge 2$ . These are subleading relative to the dominant IR tower and are associated with renormalization-group running.

Significance
The paper claims to establish a direct, explicit correspondence between the PN hereditary amplitude hierarchy (as developed in NRGR) and the Sahoo–Sen logarithmic soft theorems for gravitationally bound binaries. By isolating the specific role of static mass insertions in generating the IR logarithmic tower, the work clarifies that in bound systems, the entire soft logarithmic structure arises solely from the backscattering (tail) sector, distinct from the trajectory-based contributions present in hyperbolic scattering. The derivation of the all-order leading-logarithm tower $(GM)^n \omega^{n-1} (\ln \omega)^n$ from first principles in the PN framework validates the universality of the soft theorem predictions in the context of compact binary coalescence. Furthermore, the work clarifies the distinction between the "tail" in the PN-MPM literature (radiative drag) and the "tail to the memory" in soft theorems, showing they describe the same long-distance radiative mechanism in the bound-system limit.

More like this