Stochastic resetting — randomly returning a system to some reference state — has become a standard tool in statistical physics. The usual setup assumes complete or near-complete resets: the particle teleports back to the origin, the counter drops to zero, the population collapses. The mathematics is clean because each reset is a fresh start.
Galla considers the opposite limit. Resets are frequent but small. Rather than occasional catastrophes, the system experiences constant tremors. A particle at position x gets nudged toward zero by a fraction s ≪ 1, but this happens at rate λ/s — so the total resetting impulse stays finite as the individual resets become infinitesimal.
The key mathematical result is that this limit produces a diffusion. The discrete random nudges merge into a continuous noise term √(λs)g(x)η plus a deterministic drift −λg(x). Frequent weak resetting becomes indistinguishable from a specific kind of multiplicative noise. This is not a metaphor or approximation in the loose sense — it is a controlled limit with explicit error bounds.
The surprise is what the noise creates. For a two-species system near equilibrium, the resetting-induced noise drives oscillations that the deterministic dynamics alone cannot produce. The power spectrum shows a peak at a characteristic frequency determined by the Jacobian of the system. These are genuine stochastic cycles — patterns born from disruption, not in spite of it. The deterministic approximation (keeping the drift, dropping the noise) misses them entirely.
The multi-particle case reveals something structural. When all particles share the same resetting events (simultaneous small catastrophes), they become conditionally independent given the reset history. Condition on the sequence of tremors, and each particle evolves independently. Average over that sequence, and correlations appear — not from interaction, but from shared disruption. The particles never influence each other. They are correlated because they were shaken by the same hand.
The practical reach depends on how many systems actually live in this regime. Biological populations subject to environmental disturbance, financial portfolios experiencing market-wide shocks, ecological communities hit by recurring minor perturbations — all arguably face frequent weak resets rather than rare catastrophic ones. The diffusion approximation converts each of these from a discrete-event problem (hard to analyze) into a stochastic differential equation (extensively studied). The toolbox of SDE theory — stationary distributions, first passage times, spectral analysis — becomes immediately available.
The deepest implication: catastrophes that are individually negligible can collectively create structure that no single catastrophe could. The cycles exist only in the aggregate. No individual tremor oscillates. The pattern is a property of the ensemble of disruptions, not of any one.