In equilibrium, topological defects in the O(2) model are permanent. A vortex paired with its anti-vortex can annihilate, but isolated defects are trapped — the system lacks a mechanism to drive them out. Break reciprocity, and the defects can heal.
Rouzaire, Pearce, Pagonabarraga, and Levis study the XY model with non-reciprocal interactions — where the influence of spin A on spin B differs from B on A. Non-reciprocity acts like an active force, driving excitations through the orientation field. The dynamics follow a generalized Burgers equation, and tuning the degree of non-reciprocity enables control over excitation trajectories.
The key result: above a certain non-reciprocity threshold, the system relaxes to its ground state. Defects that would persist indefinitely under reciprocal interactions are swept away. The broken symmetry doesn't disorder the system — it orders it, by providing the directional bias needed to push defects toward boundaries or annihilation partners.
This inverts the standard intuition about symmetry breaking. In most contexts, breaking a symmetry introduces disorder or instability. Here, breaking reciprocal symmetry removes metastable traps. The equilibrium system is stuck because its dynamics are too balanced — every pathway toward ground state relaxation has an equal and opposite pathway away from it. Non-reciprocity breaks the balance, creating a preferred direction in configuration space.
The analogy to active systems is exact. Self-propelled particles, living cytoskeletal networks, and neural oscillators all feature non-reciprocal interactions. In each case, the broken reciprocity enables dynamical behaviors — collective motion, pattern formation, phase ordering — that equilibrium systems cannot access. The non-reciprocal O(2) model captures the generic mechanism: asymmetric coupling converts trapped states into transient ones.
Order from imbalance.