For forty years, the phase diagram of heavy fermion systems has been organized around a single competition: Kondo screening of local magnetic moments versus RKKY exchange between them. Doniach's 1977 framework places these two mechanisms on opposite sides of a coupling-strength dial. Turn the dial one way and local moments survive, producing magnetic order. Turn it the other way and conduction electrons screen the moments, producing a heavy Fermi liquid. The physics is rich — quantum critical points, unconventional superconductivity, non-Fermi liquid behavior — but the contestants are always the same.
Tuvia, Kang, et al. (2026) show that in β-UTe₃, a uranium-based van der Waals tritelluride, the Kondo effect competes with charge density wave order instead. Scanning tunneling spectroscopy reveals Fano resonances — the fingerprint of heavy fermion behavior — while quasiparticle interference imaging shows the destruction of Fermi surface nesting that would support a CDW. The isostructural rare-earth tritellurides (RTe₃) all exhibit CDW order. β-UTe₃ does not, because its spatially extended 5f orbitals enhance Kondo hybridization enough to suppress the charge instability before it forms.
The same screening mechanism. A different opponent. The Kondo effect doesn't care what order it's destroying — it cares about hybridization strength. When the competing order is magnetic, you get Doniach's diagram. When it's charge order, you get a different diagram with the same underlying physics. The framework wasn't wrong. It was parochial — defined by the opponent rather than the mechanism.
The general principle: when a mechanism has been studied exclusively through one competition, the temptation is to define the mechanism by that competition. Expanding the roster of opponents reveals which features belong to the mechanism itself and which were borrowed from the specific fight it happened to be in.