Nonlinear response and retarded three-point correlators of charge transport

Core claim

The paper's central claim is that nonlinear charge response in a relativistic kinetic theory is exactly computable: solving the Boltzmann equation perturbatively in the external $U(1)$ source gives a recursion in which the $n$-th order distribution obeys a linear equation with lower orders acting as known sources, and at second order the result is an explicit closed-form, fully retarded three-point correlator $G_R^{0,00}$ at arbitrary frequencies and momenta, with the tensor components $G_R^{0,0i}$ and $G_R^{0,ij}$ given in Appendix A. The author verifies the nonlinear Ward identities (31)–(32) that charge conservation demands, and checks the result against four consistency limits: vanishing in the homogeneous limits $k'\to 0$, $k-k'\to 0$, and $k\to 0$, and the thermodynamic insertion relation $G_R^{0,00} \to -\partial_{\mu_0} G_R^{0,0}$ in the static-homogeneous limit. The analytic structure splits into the two-point branch points $\omega = \pm|k| - i/\tau_R$ replicated on each of the three external legs, plus genuinely new mixed branch points fixed by $D_1^2 - 4D_0D_2 = 0$, a Landau-type threshold where two streaming denominators become resonant for the same particle direction. In the collisionless limit $\tau_R\to\infty$ the correlator acquires an enlarged permutation symmetry among all three legs, which the author interprets as the loss of a causal arrow of time when dissipation vanishes. Because the Boltzmann equation carries no noise sector, the paper concludes that fluctuation-loop corrections to the two-point function cannot be closed from these retarded vertices alone, and that a Boltzmann–Langevin or KMS-constrained effective action would be needed for that task.