How do nonlinear terms in Schrödinger-Newton gravity create superluminal spreading compared to linear quantum mechanics?

The superluminal spreading you're asking about hinges on a subtle distinction: the *apparent* faster-than-light signal propagation that arises from the nonlinear self-gravitational term in Schrödinger-Newton theory, and whether it violates relativity in practice.

The nonlinear mechanism

In linear quantum mechanics, a localized wave packet spreads at a rate set by the uncertainty principle—high-momentum components diffuse outward. The Schrödinger-Newton equation adds a gravitational self-interaction: the particle's own mass curves spacetime (or equivalently, creates a potential that couples back to the wave function). This nonlinear feedback can alter the spreading rate [AG-2024.02-1644].

Specifically, one proposed framework introduces a negative kinetic-energy term that reverses the spreading: as an object's mass exceeds about 10^−23 kg, wave packets *collapse* rather than spread [AG-2024.02-1644]. The nonlinearity thus acts like a mass-dependent damping mechanism.

The apparent superluminal signalling problem

Because the Schrödinger-Newton equation is deterministic and nonlinear, it initially appeared to allow faster-than-light bit transfer—a feature forbidden by relativity's no-signalling principle. However, recent work has challenged this conclusion by quantifying what "superluminal signalling" actually means operationally.

The key insight is that while the wave function's *support* (the region where it's nonzero) can technically expand faster than light, the *probability of successfully encoding and extracting a superluminal message* drops sharply with system size and mass [AG-2025.12-539]. In other words: yes, there is spreading faster than the light cone, but the effect is so weak for macroscopic objects that you cannot use it to communicate. Furthermore, the underlying relativistic parent theory—Einstein's semiclassical gravity equations—respects causality perfectly; the Schrödinger-Newton equation is merely its non-relativistic approximation [AG-2025.12-539].

A complementary result shows that if you carefully preserve the linearity of the *diagonal* density-matrix elements (the observable probabilities), you can construct a nonlinear Schrödinger equation that avoids signalling altogether while still producing wave-packet collapse at large mass [AG-2024.02-1644].

Why this matters

The practical upshot is that the Schrödinger-Newton model, despite its nonlinear appearance, may be *more* compatible with the no-signalling principle than you'd expect—and the theory could be tested in the lab (e.g., via optomechanical experiments on massive oscillators) without fear that it predicts causal paradoxes [AG-2025.03-582]. The superluminal spreading is real mathematically but physically toothless.