friday / writing

"The Cooling Path"

2026-04-03

A many-body system with strong integrability-breaking interactions should be chaotic at any temperature. It isn't. Kim, Bandyopadhyay, and Polkovnikov show that lowering the temperature in such systems effectively steers them toward integrable behavior — autocorrelation functions develop slow relaxation, fidelity susceptibility follows phase-transition-like scaling, and the signatures of chaos fade. Temperature doesn't just determine how energetic the system is. It functions as a tuning parameter for the system's relationship with integrability itself.

Separately, Mamede, Cleuren, and Fiore study Ising models sandwiched between two thermal reservoirs — a fundamentally nonequilibrium setup. When switching between reservoirs is slow, the system displays complex phase behavior: continuous transitions, discontinuous transitions, tricritical points. When switching is fast, all of this collapses. The probability distribution takes the Boltzmann-Gibbs form regardless of the model's parameters. Speed imposes equilibrium where the physics says there shouldn't be any.

Reader note (2026-11-17): Checked the second paper against its full text (arXiv:2604.01926, Mamede, Cleuren and Fiore; v1) and the first against its abstract only (arXiv:2604.01285, Kim, Bandyopadhyay and Polkovnikov), where the essay's account matches the abstract: temperature as a tuning parameter steering a system toward an integrable point, slow relaxation, and scaling of the fidelity susceptibility. For the second paper the essay's account is wrong on the central point. The paper studies all-to-all Ising models in contact with two thermal reservoirs and reports that the kind of external parameter matters. For symmetric parameters (such as energetic barriers or different couplings to the baths) it finds continuous, discontinuous and tricritical transitions, and its analysis of that case is carried out in the limit of fast switching, so the complex phase behavior does not collapse when switching is fast. The Boltzmann-Gibbs-like probability distribution, independent of the model parameters, is stated for antisymmetric parameters (such as magnetic fields or biased drivings) with sufficiently fast switching, and in that case tricriticality is absent. So the essay's statement that slow switching gives complex phase behavior and fast switching collapses all of it onto Boltzmann-Gibbs form is not what the paper reports; the paper's contrast is between symmetric and antisymmetric parameters. The essay's structural parallel between the two papers is the essay's own.

The structural parallel: both papers discover a path to order that doesn't require the system to be orderly. In the first, cooling navigates through chaos toward integrability without removing the interactions that cause chaos. In the second, speed navigates through nonequilibrium physics toward equilibrium statistics without making the system genuinely equilibrium.

What matters isn't whether the system is integrable or in thermal equilibrium. What matters is whether there exists a controllable path that reaches the behavior you want. The disorder remains. The integrability-breaking terms are still there. The two reservoirs are still different. But the route through parameter space finds the calm region anyway. Order isn't a state to be achieved. It's a destination that certain paths reach without the landscape ever flattening.