Author note (2026-11-02). The paper cited here, arXiv:2603.10972 (“Violating the All-or-Nothing Picture of Local Charges in Non-Hermitian Bosonic Chains”), is by Mizuki Sanatani, Naoto Shiraishi. The name(s) this essay attributes it to (Linkowski, Muth) do not appear among the listed authors, so the attribution below is wrong or refers to someone else's earlier work; I have not worked out which. Found by a mechanical check of each essay's named authors against the arXiv author list.
A quantum chain can have a conserved quantity at range 3 but not at range 4.
Integrability in quantum many-body systems is often treated as binary: a system is either integrable (possessing a complete tower of local conserved charges at every range) or it isn't. The empirical expectation, reinforced by decades of examples, is all-or-nothing — if you find one nontrivial local charge, you expect the full tower.
Linkowski and Muth (arXiv:2603.10972) construct explicit counterexamples. Working with bosonic chains that have symmetric nearest-neighbor hopping and arbitrary on-site terms (including non-Hermitian), they find systems with a 3-local charge but no other nontrivial local charges. More striking: systems with k-local charges for all k except k = 4. The ladder of conservation has a missing rung.
This breaks the standard integrability test. The Grabowski-Mathieu test checks for a 3-local charge and, if one exists, declares the system integrable. But a system can pass this test while having no higher local charges — the 3-local charge is an orphan, not the base of a tower.
The construction is systematic: the authors derive necessary and sufficient conditions for k-local charges to exist in their class of models, yielding an exhaustive classification. The non-Hermitian terms are essential. Hermitian systems in this class do appear to obey all-or-nothing. It's the opening of the non-Hermitian door — allowing gain and loss, PT-symmetry breaking, exceptional points — that permits integrability to become selective.
The structural lesson: conservation laws are not a package deal. The symmetry that protects a charge at one range need not extend to adjacent ranges. The tower can have gaps, and the gaps are physical.
Linkowski and Muth, "Violating the All-or-Nothing Picture of Local Charges in Non-Hermitian Bosonic Chains," arXiv:2603.10972 (2026).