The Kerr-Newman Two-Twistor Particle Preprint: CALT-TH 2026-011 Joon-Hwi Kim Abstract An all-orders worldline effective action for Kerr-Newman black hole is achieved in twistor particle theory. Exact hidden symmetries are identified in self-dual backgrounds. Introduction An important topic in the current gravitational physics literature has been the construction of worldline effective actions Goldberger:2004jt ; Porto:2005ac ; Levi:2015msa ; Porto:2016pyg ; Levi:2018nxp ; Kalin:2020mvi of spinning black holes. Recently, Ref. njmagic has derived an exact first-order action of the Kerr black hole and provided its explicit all-orders formula, by using massive twistor theory Penrose:1974di ; Perjes:1974ra ; tod1977some and the tangent bundle formalism for geodesic deviation gde . Ref. njmagic ’s derivation of the Kerr action provides the test-particle counterpart of the Newman-Janis trick njmagic ; probe-nj . The Newman-Janis trick is a solution-generating technique that rederived the Kerr metric by performing a complex-geometrical transformation on the Schwarzschild metric Newman:1965tw-janis . This trick then facilitated the very historical derivation of the Kerr-Newman solution Newman:1965my-kerrmetric . According to a modern view nja , the Newman-Janis trick encodes that the Kerr solution is diffeomorphic to a pair of self-dual and anti-self-dual Taub-NUT instantons when taken as a holomorphic saddle. The present paper generalizes Ref. njmagic ’s construction to derive the first-order action of the Kerr-Newman black hole coupled to background gravitational and electromagnetic fields. Again, the systematic formalism of Ref. gde proves instrumental in obtaining explicit formulae for the action. We identify a Hamiltonian system that can represent the exact, all-orders dynamics of the Kerr-Newman black hole in its point-particle effective theory. It is found that the conclusions of Ref. njmagic are preserved. Nonlinear Newman-Janis shift manifests in self-dual backgrounds. A chiral (“googly”) formulation of the Kerr-Newman action is viable for generic, non-self-dual external fields. In perturbation theory, the spin exponentiation of same-helicity Compton amplitudes Guevara:2018wpp ; Guevara:2019fsj ; aho2020 ; Johansson:2019dnu ; Aoude:2020onz ; Lazopoulos:2021mna is manifested to all graviphoton multiplicities. We also derive the classical equations of motion in self-dual backgrounds. We confirm the complexified, spinning analog of equivalence principle proposed by Ref. sst-asym . This implies dynamical hidden symmetries in the presence of Killing-Yano tensors, generalizing Ref. probe-nj ’s previous result about the Kerr particle. Geodesic Deviation in Einstein-Maxwell Geometry First, we present the new application of Ref. gde ’s tangent bundle formalism for geodesic deviation. Let ({\mathcal{M}},g) be a real-analytic pseudo-Riemannian four-manifold, and let F{\>\in\>}\Omega^{2}({\mathcal{M}}) be a two-form. Let \nabla be the Levi-Civita connection. Let D be the covariant exterior derivative due to \nabla . Let N\in T(T{\mathcal{M}}) be the horizontal lift of the tautological vector field of the tangent bundle T{\mathcal{M}} with respect to \nabla . This is the geodesic spray, i.e., the generator of geodesic deviation in T{\mathcal{M}} . We explicitly evaluate ({\iota}_{N}d)^{\ell-1}\hskip 1.00006pt{\iota}_{N}F=({\iota}_{N}D)^{\ell-1}\hskip 1.00006pt{\iota}_{N}F for all integers \ell\geq 1 . The result is \displaystyle\begin{split}({\iota}_{N}D)^{\ell-1}\hskip 1.00006pt{\iota}_{N}F\hskip 1.00006pt\,=\,\hskip 1.00006pt\begin{aligned} &(P_{\ell})_{\mu}\hskip 1.00006ptdx^{\mu}+(\ell{\hskip 1.00006pt-\hskip 1.00006pt}1)\hskip 1.00006pt(P_{\ell-1})_{\mu}\hskip 1.00006pt\,Dy^{\mu}\\ &+\sum_{p=2}^{\lfloor{(\ell+1)/2}\rfloor}\kern-3.99994pt\sum_{\alpha\in\Omega_{p}\hskip-0.40974pt(\ell+1)}\kern-6.00006pt\mathscr{K}_{\alpha}\left(\hskip 1.00006pt{\hskip 0.50003pt\begin{aligned} &\big(\hskip 1.00006pt{\hskip-0.50003ptP_{\alpha_{1}-1}Q_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p}}\hskip-0.50003pt}\hskip 1.00006pt\big)_{\hskip-0.81949pt\mu}\,dx^{\mu}\\ &+(\alpha_{p}{\hskip 1.00006pt-\hskip 1.00006pt}2)\big(\hskip 1.00006pt{\hskip-0.50003ptP_{\alpha_{1}-1}Q_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p-1}}Q_{\alpha_{p}-1}\hskip-0.50003pt}\hskip 1.00006pt\big)_{\hskip-0.81949pt\mu}\,Dy^{\mu}\hskip-1.00006pt\end{aligned}}\hskip 1.00006pt\right)\,.\end{aligned}\end{split} (1) . This is a new calculation, not presented in our previous works gde ; njmagic . Yet, our notations are the same. Each \alpha=(\alpha_{1},\alpha_{2},\cdots,\alpha_{p})\in\Omega_{p}(\ell) is an ordered partition such that \alpha_{1}+\alpha_{2}+\cdots+\alpha_{p}=\ell and \alpha_{i}\geq 2 . The coefficient \mathscr{K}_{\alpha} for \alpha\in\Omega_{p}(\ell) is a product of binomial coefficients: \displaystyle\mathscr{K}_{\alpha}\,=\,\bigg[\,{\prod_{i=1}^{p}\binom{\big(\hskip 1.00006pt{\sum_{j=i}^{p}\alpha_{j}}\hskip 1.00006pt\big){\hskip 1.00006pt-\hskip 1.00006pt}2\hskip 0.50003pt}{\alpha_{i}{\hskip 1.00006pt-\hskip 1.00006pt}2}\hskip-1.00006pt\hskip-1.00006pt}\,\bigg]\,. (2) For any \alpha=(\alpha_{1},{\cdots},\alpha_{p}) such that \min\{\alpha_{1},{\cdots},\alpha_{p}\}<2 , we set \mathscr{K}_{\alpha}=0 . The P - and Q -tensors gde are \displaystyle(P_{\ell})_{\mu} \displaystyle\,=\,F_{\rho_{1}\mu;\rho_{2};\cdots;\rho_{\ell}}(x)\,y^{\rho_{1}}y^{\rho_{2}}{\hskip 0.50003pt\cdots\hskip 1.00006pt}y^{\rho_{\ell}}\,, (3a) \displaystyle(Q_{\ell})^{\mu}{}_{\sigma} \displaystyle\,=\,R^{\mu}{}_{\rho_{1}\rho_{2}\sigma;\rho_{3};\cdots;\rho_{\ell}}\hskip-0.50003pt(x)\,y^{\rho_{1}}y^{\rho_{2}}y^{\rho_{3}}{\hskip 0.50003pt\cdots\hskip 1.00006pt}y^{\rho_{\ell}}\,, (3b) while P_{\ell} and Q_{\ell} are set to zero for \ell=0 and \ell=0,1 , respectively. Eq. ( 1 ) is readily proven by mathematical induction, based on the recursive structure illustrated in Fig. 2 . This applies the “organic chemistry” intuition of Ref. gde . Amusingly, Eq. ( 1 ) can be quickly seen by a “covariant color-kinematics” cck mapping, which views gravity as a gauge theory of the Lorentz group. Namely, Eq. ( 1 ) can be deduced from the formula for ({\iota}_{N}D)^{\ell-1}\hskip 1.00006pt{\iota}_{N}R^{m}{}_{n} given in Ref. gde ; njmagic . By viewing the antisymmetric pair of local Lorentz indices as the adjoint index of the Lorentz group as a gauge group, ({\iota}_{N}D)^{\ell-1}\hskip 1.00006pt{\iota}_{N}R^{m}{}_{n} corresponds to ({\iota}_{N}D)^{\ell-1}\hskip 1.00006pt{\iota}_{N}F^{a} where F^{a} is a nonabelian field strength. Taking the abelian limit then reproduces Eq. ( 1 ). The Kerr-Newman Action Second, we implement the probe counterpart of Newman-Janis algorithm, in the version of either Ref. njmagic (twistorial, top-down) or Ref. probe-nj (relatively bottom-up), to derive the exact effective action of the Kerr-Newman black hole in the presence of external gravitational and electromagnetic fields. The gravitational part has been already provided by Ref. njmagic : the Kerr two-twistor particle. In this paper, we simply need to add the electromagnetic interaction action. Let A\in\Omega^{1}({\mathcal{M}}) be the electromagnetic gauge potential, and let F=dA\in Z^{2}({\mathcal{M}})\subset\Omega^{2}({\mathcal{M}}) now denote the electromagnetic field strength. A 0 {\iota}_{N}F 0 {\iota}_{N}D\hskip 1.00006pt{\iota}_{N}F 0 \vdots {\iota}_{N}{\ast F} {\iota}_{N}D\hskip 1.00006pt{\iota}_{N}{\ast F} \vdots 0 0 {d}\hskip 0.81949pt{\iota}_{N} d {\iota}_{N}{d} dual Figure 1: The “ \pounds_{N} sequence” for Einstein-Maxwell geometry. The “ \pounds_{N} sequence” for A is shown in Fig. 1 . The recipe of Refs. njmagic ; probe-nj applies Hodge duality \ell times on the 2^{\ell} -pole moment term in this sequence. This determines the electromagnetic interaction symplectic potential. Consequently, the full symplectic potential of the Kerr-Newman particle is \displaystyle\theta_{\text{KN}}\,=\, \displaystyle\Big(\,{\cos(\pounds_{N})+\sinc(\pounds_{N})\,J\hskip 1.00006pt{\iota}_{N}\hskip 0.50003ptd}\,\Big)\big(\hskip 1.00006pt{p_{\mu}\hskip 0.24994ptdx^{\mu}}\hskip 1.00006pt\big) (4) \displaystyle+\Big(\,{\cos(\pounds_{N})+\sinc(\pounds_{N})\,{\iota}_{N}\hskip 1.00006pt{*}\hskip 1.00006ptd}\,\Big)\big(\hskip 1.00006pt{A_{\mu}(x)\hskip 1.00006ptdx^{\mu}}\hskip 1.00006pt\big)\,. For simplicity, we have set the electric charge q to a unit value; to reinstate it, rescale A\mapsto qA . We are now working in the the massive correspondence space \mathcal{K} defined in Ref. njmagic . Due to limited space, we leave the details to Ref. njmagic . Recall that \mathcal{K} is equipped with the geometric data N\in\Gamma(T\mathcal{K}) and J:T^{*}\mathcal{K}\to T^{*}\mathcal{K} , where J^{2}=-\mathrm{id} . Eq. ( 1 ) facilitates the explicit evaluation of Eq. ( 4 ): \displaystyle\begin{split}\theta_{\text{KN}}\hskip 1.00006pt=\,{}&{}\theta_{(0)}+A_{\mu}(x)\hskip 1.00006ptdx^{\mu}\\ {}&{}+\hskip 1.00006pt\hskip 0.24994pt\sum_{\ell=2}^{\infty}\frac{1}{\ell!}\sum_{p=1}^{\lfloor{\ell/2}\rfloor}\kern-1.99997pt\sum_{\alpha\in\Omega_{p}(\ell)}\kern-1.99997pt\mathscr{K}_{\alpha}\hskip 1.00006pt\,p_{\mu}\left(\hskip 1.00006pt{\begin{aligned} &\big(\hskip 1.00006pt{\hskip-0.50003pt({*}^{\ell}Q_{\alpha_{1}}\hskip-0.50003pt)\hskip 1.00006ptQ_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p}}\hskip-0.50003pt}\hskip 1.00006pt\big){}^{\mu}{}_{\sigma}\,dx^{\sigma}\\ &+(\alpha_{p}{\hskip 1.00006pt-\hskip 1.00006pt}2)\hskip 1.00006pt\big(\hskip 1.00006pt{\hskip-0.50003pt({*}^{\ell}Q_{\alpha_{1}}\hskip-0.50003pt)\hskip 1.00006ptQ_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p-1}}Q_{\alpha_{p}-1}\hskip-0.50003pt}\hskip 1.00006pt\big){}^{\mu}{}_{\sigma}\,Dy^{\sigma}\hskip-0.50003pt\end{aligned}}\hskip 1.00006pt\right)\\ {}&{}+\hskip 1.00006pt\hskip 0.24994pt\sum_{\ell=1}^{\infty}\frac{1}{\ell!}\,\bigg(\,{\hskip-0.50003pt\big(\hskip 1.00006pt{\hskip-0.50003pt{*}^{\ell}P_{\ell}\hskip-0.50003pt}\hskip 1.00006pt\big){}_{\mu}\,dx^{\mu}+(\ell{\hskip 1.00006pt-\hskip 1.00006pt}1)\hskip 1.00006pt\big(\hskip 1.00006pt{\hskip-0.50003pt{*}^{\ell}P_{\ell-1}\hskip-0.50003pt}\hskip 1.00006pt\big){}_{\mu}\,Dy^{\mu}}\,\bigg)\\ {}&{}+\hskip 1.00006pt\hskip 0.24994pt\sum_{\ell=2}^{\infty}\frac{1}{\ell!}\sum_{p=2}^{\lfloor{(\ell+1)/2}\rfloor}\kern-3.99994pt\sum_{\alpha\in\Omega_{p}\hskip-0.40974pt(\ell+1)}\kern-6.00006pt\mathscr{K}_{\alpha}\left(\hskip 1.00006pt{\begin{aligned} &\big(\hskip 1.00006pt{\hskip-0.50003pt({*}^{\ell}P_{\alpha_{1}-1})\hskip 1.00006ptQ_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p}}\hskip-0.50003pt}\hskip 1.00006pt\big)_{\hskip-0.81949pt\mu}\,dx^{\mu}\\ &+(\alpha_{p}{\hskip 1.00006pt-\hskip 1.00006pt}2)\big(\hskip 1.00006pt{\hskip-0.50003pt({*}^{\ell}P_{\alpha_{1}-1})\hskip 1.00006ptQ_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p-1}}Q_{\alpha_{p}-1}\hskip-0.50003pt}\hskip 1.00006pt\big)_{\hskip-0.81949pt\mu}\,Dy^{\mu}\end{aligned}\hskip-0.50003pt}\hskip 1.00006pt\right)\,,\end{split} (5) . Here, {*}^{\ell}P_{j} and {*}^{\ell}Q_{j} means to act on the Hodge star \ell times on the electromagnetic field strength F and the Riemann tensor R inside P_{j} and Q_{j} . Again, \theta_{(0)} is defined in Ref. njmagic : the covariantization of the free-theory symplectic potential of the two-twistor particle. For the reader’s sake, we reproduce it below: \displaystyle\theta_{(0)}\,=\,p_{\mu}\hskip 0.24994ptdx^{\mu}+i\hskip 1.00006pty^{{\dot{\alpha}}\alpha}\hskip 1.00006pt\Big(\,{\lambda_{\alpha}{}^{I}\hskip 0.50003ptD\bar{\lambda}_{I{\dot{\alpha}}}{\,-\,}D\lambda_{\alpha}{}^{I}\hskip 1.00006pt\bar{\lambda}_{I{\dot{\alpha}}}}\,\Big)\,. (6) By incorporating the first-class constraints \phi_{0}=\frac{1}{2}\hskip 1.00006pt(p^{2}{\,+\,}m^{2}) and \phi_{1}=-p{\hskip 1.00006pt\cdot\hskip 1.00006pt}y via Lagrange multipliers, Eq. ( 5 ) concretely defines an all-orders first-order worldline action in which each term is fully and explicitly evaluated as a manifestly covariant worldline operator. It should be clear that Eq. ( 5 ) nicely reduces to the Kerr and \kern-0.74997pt\sqrt{\text{Kerr\hskip 0.50003pt}} symplectic potentials in Ref. njmagic . Also, it is straightforward that the nonabelian Kerr-Newman particle is obtained from Eq. ( 5 ) by simple replacements A_{\mu}(x)\hskip 1.00006ptdx^{\mu}\mapsto i\hskip 1.00006pt\bar{\psi}_{i}\hskip 1.00006ptD\psi^{i} and (P_{\ell})_{\sigma}\mapsto q_{a}\hskip 0.50003pt(P_{\ell})^{a}{}_{\sigma} (see the appendix of Ref. njmagic for the relevant setup). Nonlinear Newman-Janis Shift Third, we impose self-duality condition on both electromagnetic and Riemannian curvatures to reveal the dynamical, nonlinear Newman-Janis shift: {*}F_{\mu\nu}=i\hskip 0.50003ptF_{\mu\nu} , {*}R^{\mu}{}_{\nu\rho\sigma}=i\hskip 0.50003ptR^{\mu}{}_{\nu\rho\sigma} . To this end, we simply replace {*}^{\ell} with i^{\ell} in Eq. ( 5 ). Recalling the explanations given in Ref. njmagic , it is seen that this replacement results in \displaystyle\theta_{\text{KN}}\,\sim\,{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{i\pounds_{N}}\Big(\,{p_{\mu}\hskip 0.24994ptdx^{\mu}\hskip 0.50003pt+\hskip 1.00006pt\hskip 0.24994ptA_{\mu}(x)\hskip 1.00006ptdx^{\mu}\hskip 0.50003pt+\hskip 1.00006pt\hskip 0.24994pt\theta_{\text{G2A}}}\,\Big)\,, (7) where \sim signifies equivalence modulo total derivative. Specifically, we have discarded i\hskip 1.00006pt\sinc(\pounds_{N})\hskip 1.00006ptd(A_{\mu}(x)\hskip 1.00006pty^{\mu}) . Eq. ( 7 ) is nothing but a minimally coupled charged scalar particle—i.e., the Reissner-Nordström particle—geodesically deviated into complex spacetime, up to the “Gilbert-to-Ampère” replacement njmagic implemented by \theta_{\text{G2A}} which reflects a kinematical detail: \displaystyle\theta_{\text{G2A}}\,=\,i\hskip 1.00006pty^{{\dot{\alpha}}\alpha}\hskip 1.00006pt\Big(\,{\lambda_{\alpha}{}^{I}\hskip 0.50003ptD\smash{\tilde{\lambda}}_{I{\dot{\alpha}}}{\,-\,}D\lambda_{\alpha}{}^{I}\hskip 1.00006pt\smash{\tilde{\lambda}}_{I{\dot{\alpha}}}}\,\Big)-i\hskip 1.00006ptp_{\mu}Dy^{\mu}\,. (8) Clearly, this incarnates the spirit of the Newman-Janis algorithm Newman:1965tw-janis and its very historical application Newman:1965my-kerrmetric to the Reissner-Nordström solution for the derivation of the Kerr-Newman solution. The exponentiated Lie derivative {\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{i\pounds_{N}} in Eq. ( 7 ) describes the map \gamma defined in Ref. njmagic that brings us to the “primed chart” (z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}},y^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}},\lambda_{\alpha^{\prime}}{}^{I}) on the (complexified) massive correspondence space \mathcal{K} . With this understanding, Eq. ( 7 ) boils down to \displaystyle\theta_{\text{KN}}^{+}\,=\,p_{\mu^{\prime}}\hskip 0.24994ptdz^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}-2i\hskip 1.00006pt\smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}}\hskip 1.00006pty^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\hskip 1.00006ptD\lambda_{\alpha^{\prime}}{}^{I}+A_{\mu^{\prime}}(z)\hskip 1.00006ptdz^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\,, (9) where we have discarded a total derivative just as in Ref. njmagic . Here, z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}=\delta^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\mu}\hskip 1.00006pt(x^{\mu}+iy^{\mu}+\Gamma^{\mu}{}_{\rho\sigma}(x)\hskip 1.00006pty^{\rho}y^{\sigma}+\mathcal{O}(y^{3})) are the holomorphic coordinates of curved spinspacetime. Reviving the analysis in Ref. probe-nj , one sees that the worldline perturbation theory due to Eq. ( 9 ) manifests the same-helicity (positive-helicity, in particular) spin exponentiation of the graviphotonic Compton amplitudes at all multiplicities and any massless species. Heavenly Equations of Motion The classical equations of motion due to the symplectic potential in Eq. ( 9 ) and the first-class constraints \phi_{0} and \phi_{1} are easy to derive by the technique of covariant symplectic perturbations spt-math . By fixing their Lagrange multipliers respectively as \kappa^{0}(\tau)=1 and \kappa^{1}(\tau)=0 , we find \displaystyle\begin{split}\dot{z}^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\hskip 1.00006pt=\hskip 1.00006ptp^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}&\,,\,\,\,\frac{Dy^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}}{d\tau}\hskip 1.00006pt=\hskip 1.00006ptF^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\nu^{\prime}\hskip-0.40974pt}(z)\,y^{\nu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\,,\\ \frac{D\lambda_{\alpha^{\prime}}{}^{I}}{d\tau}\hskip 1.00006pt=\hskip 1.00006pt0&\,,\,\,\,\frac{D\smash{\tilde{\lambda}}_{I\alpha{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}}{d\tau}\hskip 1.00006pt=\hskip 1.00006pt\smash{\tilde{\lambda}}_{I\beta{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\,\tilde{F}^{{\dot{\beta}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{{\dot{\alpha}}^{\prime}\hskip-0.40974pt}(z)\,.\end{split} (10) This shows that the motion of our Kerr-Newman particle describes a complex geodesic worldline z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} in spinspacetime on which the left-handed spin frame \lambda_{\alpha^{\prime}}{}^{I} is covariantly constant (parallel-transported), while the right-handed spin frame \smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}} and the spin length pseudovector y^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} are transported according to the Lorentz force law due to the dynamically Newman-Janis shifted field strength. Based on the fact that the left-handed spinor bundle is flat for the moment, D\lambda_{\alpha^{\prime}}{}^{I}\hskip-1.00006pt/d\tau=0 explicitly verifies the spinning, complexified analog of equivalence principle, which was proposed by Ref. sst-asym as the property characterizing spinning black holes. Namely, there is no anti-self-dual component in the spin precession, in self-dual backgrounds. When re-covariantized at x^{\mu} in spacetime (i.e., transcribed to the defining chart of \mathcal{K} ), Eq. ( 10 ) yields an all-orders extension of the Bargmann-Michel-Telegdi Bargmann:1959gz and Mathisson-Papapetrou-Dixon Mathisson:1937zz ; Papapetrou:1951pa ; Dixon:1970zza equations with all the multipole coefficients, both electromagnetic and gravitational, set to unity. This is indeed the desired behavior for the Kerr-Newman particle. Crucial point here is that the unity of the gravitational multipole coefficients is achieved because we have utilized geodesic deviation in generating the spin-curvature couplings in the action. Similarly, it is also easily seen that the holomorphic coordinates z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} of the Kerr black hole’s curved spinspacetime are Poisson-commutative in this self-dual limit: \displaystyle{\{\hskip 0.50003pt{\hskip 0.50003ptz^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\hskip 0.50003pt},{z^{\nu^{\prime}}}\hskip 0.50003pt\}}\,=\,0\,. (11) Hidden Symmetry Fourth, let us suppose the presence of a Killing vector X\in\Gamma(T{\mathcal{M}}) as well as a Killing-Yano tensor Y\in\Omega^{2}({\mathcal{M}}) . These shall satisfy \displaystyle\pounds_{X}g=0\,,\quad\pounds_{X}F=0\,,\quad Y^{\mu}{}_{\rho}\hskip 0.50003ptF^{\rho}{}_{\nu}\,=\,F^{\mu}{}_{\rho}\hskip 0.50003ptY^{\rho}{}_{\nu}\,, (12) and Y_{\mu\nu;\rho}{\>=\>}Y_{[\mu\nu;\rho]} . The last condition describes a commuting property, which holds for the Kerr-Newman background by the type-D property Hughston:1972qf and also for its self-dual sector: the self-dual charged Taub-Newman-Unti-Tamburino (NUT) solution in which F^{\mu}{}_{\nu}\propto Y^{\mu}{}_{\nu}/|\vec{x}|^{3} . See Ref. nja and also Ref. Adamo:2023fbj . As is well-known Hughston:1972qf , the Reissner-Nordström particle, governed by the equations of motion \dot{x}^{\mu}{\>=\>}p^{\mu} , Dp^{\mu}\hskip-1.00006pt/d\tau=F^{\mu}{}_{\nu}(x)\hskip 1.00006ptp^{\nu} , enjoys the conserved quantities \displaystyle\mathds{Q}\, \displaystyle=\,p_{\mu}\hskip 0.50003ptX^{\mu}(x)+\alpha(x)\,, (13a) \displaystyle\mathds{C}\, \displaystyle=\,-p_{\mu}\hskip 1.00006ptY^{\mu}{}_{\rho}(x)\hskip 1.00006ptY^{\rho}{}_{\nu}(x)\,p^{\nu}\,. (13b) In Eq. ( 13a ), \alpha\in C^{\infty}\hskip-0.50003pt({\mathcal{M}}) is a scalar field such that d\alpha=-{\iota}_{X}F Hughston:1972qf , whose local existence is guaranteed because -d{\iota}_{X}F={\iota}_{X}dF=0 . Concretely, one can take \alpha={\iota}_{X}A if one is willing to make an explicit gauge choice such that \pounds_{X}A=0 . To derive Eq. ( 13b ), one notes that the covariant precession of the vector Y^{\mu}{}_{\nu}(x)\hskip 1.00006ptp^{\nu} is the same as p^{\mu} by virtue of the commuting property in Eq. ( 12 ). Ref. probe-nj has established the bold proposal that exact (all-orders in spin and coupling) conserved charges of the spinning black hole probe directly arise by dynamically Newman-Janis shifting the conserved quantities of the non-spinning black hole probe, in the self-dual sector. By appreciating this idea, we identify the following exact conserved charges for our Kerr-Newman probe: \displaystyle\mathds{Q}\, \displaystyle=\,p_{\mu^{\prime}}X^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}(z)+\alpha(z)\,, (14a) \displaystyle\mathds{R}\, \displaystyle=\,p_{\mu^{\prime}}Y^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\nu^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006pty^{\nu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\,, (14b) \displaystyle\mathds{C}\, \displaystyle=\,-p_{\mu^{\prime}}Y^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\rho^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006ptY^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\nu^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006pt\hskip 1.00006ptp^{\nu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\,. (14c) In the explicit gauge choice for the electromagnetic gauge potential, one finds \alpha(z){\>=\>}A_{\mu^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006ptX^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}(z) . It is easy to verify that \mathds{Q} , \mathds{R} , \mathds{C} in Eq. ( 14 ) are all conserved by the equations of motion in Eq. ( 10 ). Note that Dp^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\hskip-1.00006pt/d\tau=F^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\nu^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006ptp^{\nu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} . The covariant precession behaviors exhibited by p^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} , y^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} , and Y^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\nu^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006ptp^{\nu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} are all identical. The conserved charges in Eq. ( 14 ) are sufficient to establish integrability of the Kerr-Newman probe dynamics in self-dual Einstein-Maxwell backgrounds admitting at least two Killing vectors and a Killing-Yano tensor. In Ref. probe-nj , such integrability has been identified as the property that characterizes black holes among all possible massive spinning objects. While Ref. probe-nj has substantiated this proposal by the Kerr particle, this paper reports the same conclusion for the Kerr-Newman particle. In fact, one can establish superintegrability by incorporating the anti-self-dual Plebaǹski two-forms Plebanski:1977zz , just as in Ref. probe-nj : \displaystyle\mathds{K}_{a}\, \displaystyle=\,p_{\mu^{\prime}}(\Sigma_{a})^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\rho^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006ptY^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\nu^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006pt\hskip 1.00006ptp^{\nu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\,. (15) It should be also highlighted that the symmetry algebra formed by \mathds{Q} , \mathds{C} , and \mathds{K}_{a} is identical to the non-spinning case up to complexification, by virtue of the very important property that {\{\hskip 0.50003pt{z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}},{z^{\nu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}}\hskip 0.50003pt\}}=0 (Eq. ( 11 )) which is characteristic of black hole spinspacetime. In short, we identify superintegrability in the dynamics of the Kerr-Newman black hole binary system at the zeroth self-force order, yet in the self-dual sector. Curved Massive Twistor Space Fifth, we identify the “mu variables” to return to the massive twistor space. To this end, we trivialize the anti-self-dual spin connection coefficients, employ tetrad formalism, and expand the holomorphic coframe around a flat background as e^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha}=dz^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha}+h^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha} which may utilize the second Plebański coordinates plebanski1975some . Then Eq. ( 9 ) becomes \displaystyle\theta_{\text{KN}}^{+}\,=\,-\smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}}\hskip 0.50003ptdz^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha}\hskip 0.50003pt\lambda_{\alpha}{}^{I}-2i\hskip 1.00006pt\smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}}\hskip 1.00006pty^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha}\hskip 1.00006ptd\lambda_{\alpha}{}^{I}+\theta^{\prime}_{\text{KN}}\,. (16) We have discarded a total derivative, while \displaystyle\theta^{\prime}_{\text{KN}}\,=\,\Big(\,{-\smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}}\hskip 0.50003pth^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha}{}_{\mu^{\prime}\hskip-0.40974pt}(z)\hskip 0.50003pt\lambda_{\alpha}{}^{I}+A_{\mu^{\prime}\hskip-0.40974pt}(z)}\,\Big)\,dz^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\,. (17) As a result, the same identification for the mu variables as in Ref. njmagic are made: \displaystyle\begin{split}\mu^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\hskip 0.81949ptI}\,=\,z^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha}\hskip 0.50003pt\hskip 0.24994pt\lambda_{\alpha}{}^{I}\,,\quad\tilde{\mu}_{I}{}^{\alpha}\,&=\,\smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}}\big(\hskip 1.00006pt{z^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha}{\hskip 0.50003pt-\hskip 1.00006pt}2iy^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha}}\hskip 1.00006pt\big)\,.\end{split} (18) Eventually, we find \displaystyle\begin{split}\theta^{+}_{\text{KN}}\,&=\,-\smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}}\hskip 0.50003ptd\mu^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\hskip 0.81949ptI}+\tilde{\mu}_{I}{}^{\alpha}\hskip 0.50003ptd\lambda_{\alpha}{}^{I}+\theta^{\prime}_{\text{KN}}\,,\\ \omega_{\text{KN}}\,&=\,d\mu^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\hskip 0.81949ptI}{\hskip 1.00006pt{\wedge}\,}d\smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}}+d\tilde{\mu}_{I}{}^{\alpha}{\hskip 1.00006pt{\wedge}\,}d\lambda_{\alpha}{}^{I}+d\theta^{\prime}_{\text{KN}}\,.\end{split} (19) Eq. ( 19 ) identifies the notion of curved massive twistor space in terms of a deformation of the symplectic structure on massive twistor space: \displaystyle\big(\hskip 1.00006pt{\mathcal{MT},\hskip 1.00006pt\omega^{\circ}}\hskip 1.00006pt\big)\,\,\,\xrightarrow{\,\,\,\,\,\,}\,\,\,\big(\hskip 1.00006pt{\mathcal{MT},\hskip 1.00006pt\omega^{\circ}{\hskip 1.00006pt+\hskip 1.00006pt}\hskip 1.00006pt\hskip 0.24994ptd\theta^{\prime}_{\text{KN}}}\hskip 1.00006pt\big)\,. (20) This deformation is via a (1,0) -perturbation \theta^{\prime}_{\text{KN}} at the level of symplectic potential such that the Poisson commutativity of holomorphic spinspacetime coordinates sst-asym is preserved as Eq. ( 11 ). Physically, this encodes that the gravitoelectromagnetic interaction action is supported on the Newman-Janis shifted worldline z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} for the self-dual (incoming positive-helicity bialynicki1981note ; ashtekar1986note ) sector. As remarked in Ref. njmagic , the above recipes are motivated by the practical use of our framework in perturbation theory, so other pathways toward the definition of curved massive twistor space might exist. Googly Formulation Lastly, our final task is to derive the chiral (“googly”) formulation of the Kerr-Newman action. This computes the generic, non-self-dual action described in Eq. ( 5 ) in terms of the complex, primed variables (z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}},y^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}},\lambda_{\alpha^{\prime}}{}^{I}) by considering the map \gamma of Ref. njmagic . To this end, we revive the gymnastics in Ref. njmagic . Define \displaystyle\theta^{\pm}_{1}\,=\,\frac{1}{2}\,\Big(\,{p_{\mu}Dy^{\mu}\pm y^{{\dot{\alpha}}\alpha}\hskip 1.00006pt\Big(\,{\lambda_{\alpha}{}^{I}\hskip 0.50003ptD\smash{\tilde{\lambda}}_{I{\dot{\alpha}}}{\,-\,}D\lambda_{\alpha}{}^{I}\hskip 1.00006pt\smash{\tilde{\lambda}}_{I{\dot{\alpha}}}}\,\Big)\hskip-1.00006pt}\,\Big)+{\iota}_{N}F^{\pm}\,, (21) where the \pm superscripts describe the projections by \frac{1}{2}\hskip 1.00006pt(1\mp i\hskip 1.00006pt{*}) . Then it follows that \displaystyle\theta_{\text{KN}}\,\, \displaystyle=\,\,p_{\mu}\hskip 0.24994ptdx^{\mu}+A_{\mu}(x)\hskip 1.00006ptdx^{\mu}+\frac{{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{i\pounds_{N}}{\hskip 1.00006pt-\hskip 1.00006pt}1}{\pounds_{N}}\,\theta^{+}_{1}+\frac{{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{-i\pounds_{N}}{\hskip 1.00006pt-\hskip 1.00006pt}1}{\pounds_{N}}\,\theta^{-}_{1}\,, (22a) \displaystyle{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{i\pounds_{N}}\Big(\,{p_{\mu}\hskip 0.24994ptdx^{\mu}+A_{\mu}(x)\hskip 1.00006ptdx^{\mu}\hskip-0.50003pt}\,\Big)\,\, \displaystyle\sim\,\,p_{\mu}\hskip 0.24994ptdx^{\mu}+A_{\mu}(x)\hskip 1.00006ptdx^{\mu}+\frac{{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{i\pounds_{N}}{\hskip 1.00006pt-\hskip 1.00006pt}1}{\pounds_{N}}\,\theta^{+}_{1}+\frac{{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{i\pounds_{N}}{\hskip 1.00006pt-\hskip 1.00006pt}1}{\pounds_{N}}\,\theta^{-}_{1}\,. (22b) The difference between Eqs. ( 22a ) and ( 22b )—Kerr-Newman and “Reissner-Nordström {}+{}iy ”—derives \displaystyle\begin{split}\theta_{\text{KN}}\,\hskip 1.00006pt&\sim\,\,{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{i\pounds_{N}}\hskip 1.00006pt\bigg(\,{p_{\mu}\hskip 0.24994ptdx^{\mu}+\frac{{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{-2i\pounds_{N}}{\hskip 1.00006pt-\hskip 1.00006pt}1}{\pounds_{N}}\,\theta^{-}_{1}}\,\bigg)\,,\\ \,\,&=\,\,{\scalebox{1.05}[1.01]{$\mathrm{e}$}}^{i\pounds_{N}}\hskip 1.00006pt\bigg(\,{\theta_{(0)}-i\hskip 1.00006ptp_{\mu}Dy^{\mu}+\hskip 1.00006pt\hskip 0.24994pt\sum_{\ell=2}^{\infty}\frac{(-2i)^{\ell}}{\ell!}\hskip 1.00006ptp_{\mu}\big(\hskip 1.00006pt{\hskip-0.50003pt({\iota}_{N}\hskip 0.50003ptD)^{\ell-2}{\iota}_{N}\hskip 1.00006ptR^{-}{}^{\mu}{}_{\nu}\hskip-0.24994pt}\hskip 1.00006pt\big)\hskip 0.50003pty^{\nu}+\hskip 1.00006pt\hskip 0.24994pt\sum_{\ell=1}^{\infty}\frac{(-2i)^{\ell}}{\ell!}\hskip 1.00006pt({\iota}_{N}\hskip 0.50003ptd)^{\ell-1}{\iota}_{N}\hskip 1.00006ptF^{-}}\,\bigg)\,,\end{split} (23) which leads to \displaystyle\begin{split}\theta_{\text{KN}}\hskip 1.00006pt\,\sim\,\,{}&{}p_{\mu^{\prime}}\hskip 0.24994ptdz^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}-2i\hskip 1.00006pt\smash{\tilde{\lambda}}_{I{\dot{\alpha}}^{\prime}}\hskip 1.00006pty^{{\dot{\alpha}}{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\alpha{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\hskip 1.00006ptD\lambda_{\alpha^{\prime}}{}^{I}+A_{\mu^{\prime}\hskip-0.40974pt}(z)\hskip 1.00006ptdz^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\\ {}&{}+\hskip 1.00006pt\hskip 0.24994pt\sum_{\ell=2}^{\infty}\frac{(-2i)^{\ell}}{\ell!}\sum_{p=1}^{\lfloor{\ell/2}\rfloor}\kern-1.99997pt\sum_{\alpha\in\Omega_{p}(\ell)}\kern-1.99997pt\mathscr{K}_{\alpha}\hskip 1.00006pt\,p_{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\hskip 0.50003pt\left(\hskip 1.00006pt{\begin{aligned} &\big(\hskip 1.00006pt{\hskip-0.50003ptQ_{\alpha_{1}}^{-}\hskip 0.50003ptQ_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p}}\hskip-0.50003pt}\hskip 1.00006pt\big){}^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\sigma^{\prime}}\hskip 1.00006ptdz^{\sigma{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\\ &+(\alpha_{p}{\hskip 1.00006pt-\hskip 1.00006pt}2)\hskip 1.00006pt\big(\hskip 1.00006pt{\hskip-0.50003ptQ_{\alpha_{1}}^{-}\hskip 0.50003ptQ_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p-1}}Q_{\alpha_{p}-1}\hskip-0.50003pt}\hskip 1.00006pt\big){}^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\sigma^{\prime}}\hskip 0.24994ptDy^{\sigma{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\hskip 1.00006pt\hskip-0.50003pt\end{aligned}}\hskip 1.00006pt\right)\\ {}&{}+\hskip 1.00006pt\hskip 0.24994pt\sum_{\ell=1}^{\infty}\frac{(-2i)^{\ell}}{\ell!}\,\bigg(\,{\hskip-0.50003pt\big(\hskip 1.00006pt{\hskip-0.50003ptP^{-}_{\ell}\hskip-0.50003pt}\hskip 1.00006pt\big){}_{\sigma^{\prime}}\hskip 1.00006ptdz^{\sigma{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}+(\ell{\hskip 1.00006pt-\hskip 1.00006pt}1)\hskip 1.00006pt\big(\hskip 1.00006pt{\hskip-0.50003ptP^{-}_{\ell-1}\hskip-0.50003pt}\hskip 1.00006pt\big){}_{\sigma^{\prime}}\hskip 0.24994ptDy^{\sigma{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}}\,\bigg)\\ {}&{}+\hskip 1.00006pt\hskip 0.24994pt\sum_{\ell=2}^{\infty}\frac{(-2i)^{\ell}}{\ell!}\sum_{p=2}^{\lfloor{(\ell+1)/2}\rfloor}\kern-3.99994pt\sum_{\alpha\in\Omega_{p}\hskip-0.40974pt(\ell+1)}\kern-6.00006pt\mathscr{K}_{\alpha}\left(\hskip 1.00006pt{\begin{aligned} &\big(\hskip 1.00006pt{\hskip-0.50003ptP^{-}_{\alpha_{1}-1}\hskip 1.00006ptQ_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p}}\hskip-0.50003pt}\hskip 1.00006pt\big)_{\hskip-0.81949pt\mu^{\prime}}\hskip 1.00006ptdz^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\\ &+(\alpha_{p}{\hskip 1.00006pt-\hskip 1.00006pt}2)\big(\hskip 1.00006pt{\hskip-0.50003ptP^{-}_{\alpha_{1}-1}\hskip 1.00006ptQ_{\alpha_{2}}\,{\hskip-1.00006pt\cdots\hskip 1.00006pt}Q_{\alpha_{p-1}}Q_{\alpha_{p}-1}\hskip-0.50003pt}\hskip 1.00006pt\big)_{\hskip-0.81949pt\mu^{\prime}}\hskip 0.24994ptDy^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}\end{aligned}}\hskip 1.00006pt\right)\,.\end{split} (24) Again, we find the structure that the first entry in each string of P or Q tensors develops the anti-self-dual projection. Eq. ( 24 ) reproduces Eq. ( 9 ) in the self-dual limit. As a friendly reminder, our index notation gde implies \displaystyle(P_{\ell})_{\mu^{\prime}} \displaystyle\,=\,F_{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{1}\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\hskip 0.40974pt;\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{2};\cdots;\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{\ell}}(z)\,y^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{1}}y^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{2}}{\hskip 0.50003pt\cdots\hskip 1.00006pt}y^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{\ell}}\,, (25a) \displaystyle(Q_{\ell})^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}}{}_{\sigma^{\prime}} \displaystyle\,=\,R^{\mu}{}_{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{1}\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{2}\sigma{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}\hskip 0.81949pt;\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{3};\cdots;\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{\ell}}\hskip-0.50003pt(z)\,y^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{1}}y^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{2}}y^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{3}}{\hskip 0.50003pt\cdots\hskip 1.00006pt}y^{\rho{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}_{\ell}}\,. (25b) Eq. ( 24 ) provides the googly formulation of the Kerr-Newman action. To reiterate, it describes the completely general case of non-self-dual gravitoelectromagnetic field configurations but formulates the self-dual and anti-self-dual modes on an unequal footing. It provides a worldline action localized on the holomorphic worldline z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} , the perturbation theory of which manifests the same-helicity spin exponentiation for incoming positive-helicity photons and gravitons. The photon coupling is always linear. Graviphoton contact vertices arise iff the incoming photon carries negative helicity. When examined in the spin length expansion, the first graviphoton contact vertex appears at \mathcal{O}(y^{3}) . Interpretation We end with quick remarks on the physical interpretation by recalling the comments given in Ref. njmagic . Eq. ( 5 ) represents two charged masses joined by a Dirac-Misner string (line defect of magnetic and NUT fluxes). Eq. ( 5 ) represents a dangling anti-self-dual Dirac-Misner string attached to a charged mass. The shape of the Dirac-Misner string happens to be geodetic for each slice of simultaneity defined by the worldline proper time. For the Dirac string part, this is completely justified based on its well-established status as a topological surface operator. By virtue of the transparent decoding of the Newman-Janis algorithm revealed by Ref. nja , we readily realize that this string structure is the precise infrared counterpart of the actual Dirac-Misner string that exists in the ultraviolet description of the Kerr-Newman solution 1 1 1 Here, “infrared” refers to the point-particle effective theory, while “ultraviolet” refers to general relativity. . Namely, it is shown in Ref. nja that the Kerr-Newman solution as a holomorphic saddle in Einstein-Maxwell theory represents a pair of self-dual and anti-self-dual charged Taub-NUT instantons. In conclusion, the holomorphic worldline z^{\mu{\mathrlap{\smash{{}^{\prime}}}{\hskip 0.40974pt}}} is the infrared avatar of the worldline of the self-dual charged Taub-NUT instanton that constitutes the Kerr-Newman black hole. The googly action in Eq. ( 24 ) perturbs around the self-dual sector—the sector in which we, at least, have a full understanding on dynamics in terms of remarkable properties such as the superintegrability due to Eqs. ( 14 ) and ( 15 ) or the same-helicity spin exponentiation. The googly agenda means to approach the complicated dynamics of the Kerr-Newman probe from its exactly solvable part, which should sound reasonable. Note that Ref. gmoov had speculated on a worldsheet structure, although no clear physical interpretation nor explicit formulae such as Eqs. ( 5 ) and ( 24 ) were provided.