The Supersymmetry of Cuts in Pure Gauge Theory and Gravity

Jacob L. Bourjaily · 2025

Read The Supersymmetry of Cuts in Pure Gauge Theory and Gravity by Jacob L. Bourjaily (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-24984.txt.

Introduction

Jacob Bourjaily shows that scattering amplitudes in pure gauge theory and gravity, which lack supersymmetry, can still be grouped by general superfunctions far beyond tree level. At tree level, gluon and graviton amplitudes are efficiently pulled from fully supersymmetric theories, but loop fixes were expected to break that ease badly. Instead, all leading singularities of non-supersymmetric theories fit the general superfunction form.

Those functions differ from the fully supersymmetric ones only by an extent fixed by loop order and related data. Particle theory readers get lean codings of all helicity parts. The result keeps strong supersymmetric tracking useful for real-world theories like quantum chromodynamics and gravity.

Catalog record

CREATORJacob L. Bourjaily
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE48 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-24984

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The Supersymmetry of Cuts in Pure Gauge Theory and Gravity Jacob L. Bourjaily Abstract At tree-level, scattering amplitudes involving only gluons or gravitons are unaffected by supersymmetry, allowing them to be efficiently encoded by and extracted from those of maximally supersymmetric ( 𝒩 = 4 , 8 \mathcal{N}\!=\!4,8 ) theories. This fails beyond tree-level, of course, but much less than would be expected. We show that all the leading singularities of (sub-maximally or) non-supersymmetric theories can be organized into generalized superfunctions, in terms of which all helicity components can be effectively encoded. These functions differ from those of maximally supersymmetric theories by an extent determined by loop-order—broken into a sum over 2 L 2^{L} supersymmetric pieces. I Introduction The use of supersymmetry has long proven extremely useful for the computation of amplitudes in (sub-maximally or) non -supersymmetric theories. At tree-level, maximal supersymmetry can be viewed entirely as a book-keeping device, as gluons or gravitons cannot interact with other states except through loops. This has been put to powerful use in a variety of public code and computational tools for trees (see e.g Dixon:2010ik ; Bourjaily:2010wh ; Bourjaily:2023uln ). Beyond tree-level, however, amplitudes in theories with different degrees of supersymmetry obviously differ. Perturbative scattering amplitudes in gauge theory can be computed in terms of their leading singularities according to generalized unitarity Bern:1993tz ; Bern:1993mq ; Bern:1994zx ; Bern:1994cg ; Bern:1995db ; Bern:1995ix ; Bern:1996ja (see also Roiban:2004ix ; Berger:2006ci ; Berger:2006vq ; Ellis:2006ss ; Ellis:2007br ; Berger:2008sj ; Ellis:2008ir ; Giele:2008ve ; Degrande:2018neu ; Bourjaily:2020qca ), which has led to impressive improvements in both our understanding of scattering amplitudes and our ability to compute them (see e.g. Su:2006vs ; Xiao:2006vt ; Dixon:2009uk ; Abreu:2018aqd ; Abreu:2019rpt ). This is no more apparent than in the case of amplitudes in theories with maximal supersymmetry (see e.g. Bourjaily:2013mma ; Bourjaily:2015jna ; Bourjaily:2017wjl ; Abreu:2018aqd ; Bourjaily:2018omh ; Bourjaily:2019gqu ; Abreu:2019rpt ; Bourjaily:2021iyq ), integrands of which led to the discovery of on-shell recursion relations at tree-level BCF ; BCFW ; ArkaniHamed:2008yf ; Elvang:2011fx and beyond ArkaniHamed:2010kv ; Bourjaily:2023apy . Despite this tremendous progress, there is considerably less known about amplitudes in pure (or even less supersymmetric) gauge theory and gravity. Partly, this can be attributed to the difficulties of considering each of the \binom{n}{k} ‘helicity’ (or super-) component amplitudes separately (see e.g. Bourjaily:2021ujs ). In this Letter, we show that the full power of supersymmetric book-keeping can be preserved beyond tree-level. Specifically, we show that any L -loop leading singularity 1 1 1 Specifically, leading singularities constructed in terms of four-dimensional tree amplitudes—not those whose definition requires ( e.g. dimensional) regularization. of pure (or less-supersymmetric) gauge or gravity theory can be expressed in terms of generalized superfunctions which simultaneously encode all the particular configurations of external component states related by supersymmetry. In notation clarified below, an L -loop leading singularity of N {}^{k\text{\makebox[3.01389pt][c]{$\hskip-150.0pt\displaystyle\shortminus\hskip-150.0pt$}}2} MHV-degree in a theory \mathcal{N} supersymmetries can be expressed in terms of functions of the form \mathfrak{f}_{\sigma}^{{\mathcal{N}}}(\vec{p},\widetilde{\eta},\theta)\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\hskip 0.0pt\text{\makebox[0.0pt][r]{$\raisebox{0.47pt}{\hskip 1.25pt:\hskip-4.0pt}$}}=\text{\makebox[0.0pt][l]{$$}}\hskip-150.0pt$}}f_{\sigma}(\vec{p})\!\times\!\delta^{k\times\mathcal{N}}\!\big(C\!\cdot\!\widetilde{\eta}\big)\hskip-2.0pt\sum_{\vec{s}\in[2]^{L}}\hskip-4.0pt\delta^{k\times\overline{\mathcal{N}}}\!\big(D_{\vec{s}}\!\cdot\!\theta\big)\vskip-5.0pt\vskip-0.5pt (1) where \sigma{=}1,2 for sYM and sGR, respectively; f_{\sigma}(\vec{p}) is an ordinary function of on-shell momenta; C and each of the D_{\vec{s}} are k\!\times\!n matrices of ordinary functions; for each a\!\in\![n] and I\!\in\![\mathcal{N}] , \widetilde{\eta}_{a}^{I} is an auxiliary Grasmman parameter used to connect states related to the highest-helicity h\!=\!{+}\sigma state by supersymmetry, and \theta_{a}\!\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\hskip 0.0pt\text{\makebox[0.0pt][r]{$\raisebox{0.47pt}{\hskip 1.25pt:\hskip-4.0pt}$}}=\text{\makebox[0.0pt][l]{$$}}\hskip-150.0pt$}}\!\prod_{\overline{I}\in[\overline{\mathcal{N}}]}\widetilde{\eta}_{a}^{\overline{I}},\vskip-10.0pt\vskip-0.5pt (2) with [\overline{\mathcal{N}}]\!\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\hskip 0.0pt\text{\makebox[0.0pt][r]{$\raisebox{0.47pt}{\hskip 1.25pt:\hskip-4.0pt}$}}=\text{\makebox[0.0pt][l]{$$}}\hskip-150.0pt$}}\![4\sigma]\backslash[\mathcal{N}] introduced to label states connected to these by parity. Notice that in ( 1 ), the function f_{\sigma}(\vec{p}) does not depend on the degree of supersymmetry, but will be different for sYM and sGR; the matrices C and D_{\vec{s}} are the same for both theories. Generalized superfunctions are defined in terms of an L -loop graph with vertices denoting tree-amplitudes with L marked internal lines—the cutting of which would result in a tree of tree-amplitudes; for example, \begin{array}[]{@{}c@{}}\includegraphics[scale]{figures/eg_three_loop_ls}\end{array}\raisebox{-0.95pt}{\scalebox{1.25}{$\Rightarrow$}}\begin{array}[]{@{}c@{}}\includegraphics[scale]{figures/eg_three_loop_ls_cut}\end{array}\text{\makebox[0.0pt][l]{$\!\!.$}}\vskip-10.0pt\vskip-0.5pt (3) The cut diagram will have (n{+}2L) legs and N {}^{k\text{\makebox[3.01389pt][c]{$\hskip-150.0pt\displaystyle\shortminus\hskip-150.0pt$}}2\text{\makebox[3.01389pt][c]{$\hskip-150.0pt\displaystyle\text{\makebox[0.0pt][c]{$\hskip-150.0pt\displaystyle\shortminus\hskip-150.0pt$}}\text{\makebox[0.0pt][c]{$\hskip-150.0pt\displaystyle\rotatebox[origin]{90.0}{$\shortminus$}\hskip-150.0pt$}}\hskip-150.0pt$}}L} MHV-degree; because it is built entirely from tree amplitudes for which supersymmetric bookkeeping can be used it can be defined independent of the degree (or existence) of supersymmetry in terms of ordinary (maximally-supersymmetric) superfunctions. From these, all component amplitudes can be extracted, including those involving only the subset of states relevant to \mathcal{N}\!\geq\!0 sYM or sGR to define the new functions. The distinction between theories with different degrees of supersymmetry, therefore, arises entirely through the restricted set of states to be summed for each pair of legs sewn together: \sum_{\begin{subarray}{c}\text{conjugate}\\ \text{states }{\color[rgb]{0.575,0,0.225}X},{\color[rgb]{0.575,0,0.225}Y}\end{subarray}}\hskip-5.0pt\raisebox{-0.95pt}{\scalebox{1.25}{$\Rightarrow$}}\int\!\!\big[d\theta_{\!{\color[rgb]{0.575,0,0.225}X}}{+}d\theta_{\hskip-0.5pt{\color[rgb]{0.575,0,0.225}Y}}\!\big]\!\!\!\!\prod_{I\in[\mathcal{N}]}\!\!\!\big(d\widetilde{\eta}_{{\color[rgb]{0.575,0,0.225}X}}^{I}{+}d\widetilde{\eta}_{{\color[rgb]{0.575,0,0.225}Y}}^{I}\big)\,.\vskip-5.0pt\vskip-0.5pt (4) In this state sum, the Grassmann integral involving d\theta_{{\color[rgb]{0.575,0,0.225}X}} includes all those states for which |{\color[rgb]{0.575,0,0.225}X}\rangle is connected to |{\color[rgb]{0.575,0,0.225}X}\rangle^{\text{\makebox[3.01389pt][c]{$\hskip-150.0pt\displaystyle\shortminus\hskip-150.0pt$}}\sigma} by supersymmetry, and similarly for the integral over d\theta_{{\color[rgb]{0.575,0,0.225}Y}} . Choosing one parity for each pair of legs to be merged results in one of the 2^{L} terms in ( 1 ): f_{\sigma}(\vec{p})\!\times\!\delta^{k\times\mathcal{N}}\!\big(C\!\cdot\!\widetilde{\eta}\big)\hskip 0.0pt\hskip 0.0pt\delta^{k\times\overline{\mathcal{N}}}\!\big(D_{\vec{s}}\!\cdot\!\theta\big).\vskip-4.0pt\vskip-0.5pt (5) This particular contribution depends sensitively on the choice of edges cut in the original diagram; only when all 2^{L} terms are summed will the resulting generalized superfunction be independent of this choice. It is worth noting that because the cut diagrams are manifestly tree-graphs of tree-amplitudes, they necessarily admit a plane embedding. The ordering of legs on the boundary will depend on the orderings of partial amplitudes at each vertex—which can be defined unambiguously (if conventionally) for both sYM DelDuca:1999rs ; Arkani-Hamed:2014bca and sGR Paranjape:2023qsq ; Koefler:2024pzv ; Lemmon:2025dhq ; Herrmann:2016qea ; Heslop:2016plj ; Bourjaily:2023ycy . Thus, they are definable in terms of positroid varieties ArkaniHamed:book ; Bourjaily:2012gy ; Bourjaily:2016mnp . This is not terribly surprising: any non-planar on-shell diagram can be understood as constructed from a plane-embedded tree upon merging legs which need not be adjacent. But it is striking that the vocabulary of planar diagrams suffices not only to describe the various non-planar leading singularities of sYM, but also of pure YM and pure GR. II Formalism: Superfunctions and Coherent States Let us briefly review the notation used to describe superfunctions of \mathcal{N} -supersymmetric Yang-Mills and gravity theories (‘sYM’ and ‘sGR’). Let \widetilde{Q}_{I}^{\dot{\alpha}} for I\!\in\![\mathcal{N}] be generators of supersymmetry (if any exist). For each external particle indexed by a\!\in\![n] , with on-shell momentum p_{a}^{\alpha\dot{\alpha}}\!\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\hskip 0.0pt\text{\makebox[0.0pt][r]{$\raisebox{0.47pt}{\hskip 1.25pt:\hskip-4.0pt}$}}=\text{\makebox[0.0pt][l]{$$}}\hskip-150.0pt$}}p_{a}^{\mu}\sigma_{\mu}^{\alpha\dot{\alpha}}\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\text{\makebox[0.0pt][r]{$$}}=\text{\makebox[0.0pt][l]{$\raisebox{0.47pt}{\hskip-4.0pt:\hskip 1.25pt}$}}\hskip-150.0pt$}}\lambda_{a}^{\alpha}\widetilde{\lambda}_{a}^{\dot{\alpha}} , there will be a variety of particular states related by supersymmetry/parity to those with the highest helicity h{=}{+}\sigma with \sigma{=}1,2 for sYM and sGR, respectively. We gather these states into a coherent state Arkani-Hamed:2008owk as follows. For each particle we introduce 4\sigma auxiliary Grassmann parameters \widetilde{\eta}_{a}^{I} with I\!\in\![4\sigma]{=}[\mathcal{N}]\raisebox{-0.95pt}{\scalebox{1.25}{$\cup$}}\big([4\sigma]\backslash[\mathcal{N}]\big) —regardless of the degree of supersymmetry—to be used for book-keeping. We define a spinor \widetilde{\lambda}^{\dot{\alpha}}_{\overline{a}} such that [\overline{a}\,a]\!\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\hskip 0.0pt\text{\makebox[0.0pt][r]{$\raisebox{0.47pt}{\hskip 1.25pt:\hskip-4.0pt}$}}=\text{\makebox[0.0pt][l]{$$}}\hskip-150.0pt$}}\!\epsilon_{\dot{\alpha}\dot{\beta}}\widetilde{\lambda}_{\overline{a}}^{\dot{\alpha}}\widetilde{\lambda}_{a}^{\dot{\beta}}\!\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\hskip 0.0pt\text{\makebox[0.0pt][r]{$\raisebox{0.47pt}{\hskip 1.25pt:\hskip-4.0pt}$}}=\text{\makebox[0.0pt][l]{$$}}\hskip-150.0pt$}}\!1 , and for each supersymmetry generator indexed by I\!\in\![\mathcal{N}] we define \widetilde{Q}^{a}_{I}\!\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\hskip 0.0pt\text{\makebox[0.0pt][r]{$\raisebox{0.47pt}{\hskip 1.25pt:\hskip-4.0pt}$}}=\text{\makebox[0.0pt][l]{$$}}\hskip-150.0pt$}}\!\epsilon_{\dot{\alpha}\dot{\beta}}\widetilde{Q}_{I}^{\dot{\alpha}}\widetilde{\lambda}^{\dot{\beta}}_{\overline{a}} . The coherent state ‘ |a\rangle ’ is then defined to be |a\rangle\text{\makebox[15.5pt][c]{$\hskip-150.0pt\displaystyle\hskip 0.0pt\text{\makebox[0.0pt][r]{$\raisebox{0.47pt}{\hskip 1.25pt:\hskip-4.0pt}$}}=\text{\makebox[0.0pt][l]{$$}}\hskip-150.0pt$}}\mathrm{exp}\big(\widetilde{Q}_{I}^{a}\widetilde{\eta}_{a}^{I}\big)\Big(|a\rangle^{\!\text{\makebox[3.01389pt][c]{$\hskip-150.0pt\displaystyle\text{\makebox[0.0pt][c]{$\hskip-150.0pt\displaystyle\shortminus\hskip-150.0pt$}}\text{\makebox[0.0pt][c]{$\hskip-150.0pt\displaystyle\rotatebox[origin]{90.0}{$\shortminus$}\hskip-150.0pt$}}\hskip-150.0pt$}}\sigma}{+}\theta_{a}\,|a\rangle^{\frac{\mathcal{N}}{2}\text{\makebox[3.01389pt][c]{$\hskip-150.0pt\displaystyle\shortminus\hskip-150.0pt$}}\sigma}\Big)\,\vskip-4.0pt\vskip-0.5pt (6) with \theta_{a} defined in ( 2 ) in terms of the subset of \widetilde{\eta}_{a}^{I} ’s with indices I\!>\!\mathcal{N} . From the coherent state, particular component states can be extracted via Grassmann integration. For example, the state |a\rangle^{\text{\makebox[3.01389pt][c]{$\hskip-150.0pt\displaystyle\shortminus\hskip-150.0pt$}}\sigma} would be obtained from ( 6 ) via |a\rangle^{\text{\makebox[3.01389pt][c]{$\hskip-150.0pt\displaystyle\shortminus\hskip-150.0pt$}}\sigma}=\int\!\!d\widetilde{\eta}_{a}^{1}\cdots d\widetilde{\eta}_{a}^{4\sigma}\,\,|a\rangle=\int\!\!\Big(\!d\widetilde{\eta}_{a}^{1}\cdots d\widetilde{\eta}_{a}^{\mathcal{N}}\!\Big)d\theta_{a}\,\,|a\rangle\,.\vskip-4.0pt\vskip-0.5pt (7) Superfunctions are defined directly in terms of coherent states, and therefore depend not just on the momenta of the particles but also \widetilde{\eta} ’s and \theta ’s used to package the various states

Read the complete text