A topological phase is defined by what it doesn't have: no local order parameter, no symmetry breaking, no Landau classification. Its identity lives in global properties — anyon types, braiding statistics, ground-state degeneracy on a torus. Getting in is exotic. Getting out turns out to be surprisingly constrained.
Note (2026-11-07): the abstract assumes the transition is triggered by a single Abelian anyon, which the essay does not say. It finds that the transition theory depends on one additional integer parameter, and that only in generic cases is the theory after the transition related to the original by an Abelian hierarchy construction; in special cases the theory after the transition is gapless, and enriching by a U(1) symmetry can be incompatible with the full transition theory. The link to the fractional quantum Hall effect and the remarks on exotic phases having fewer ways to end are mine.
Cheng and Seiberg study what happens when a single anyon species in a 2+1-dimensional topological phase proliferates — condenses into a new vacuum. The space of possible post-transition theories looks infinite: any field theory could, in principle, emerge. But the topological starting point constrains the destination so severely that the entire transition is characterized by a single integer.
The integer controls which post-transition theory you land in, and the relationship between the original and final theories follows an Abelian hierarchy construction — the same mathematical structure that describes the fractional quantum Hall effect. The apparent freedom of a phase transition collapses to one dimension of choice.
This is topological constraint on destruction. The phase being topological means that even the ways it can die are topologically restricted. A conventional ordered phase can disorder in many ways — different fluctuation modes, different correlation length divergences. A topological phase's exit routes are counted by integers because the allowed transitions must respect the anyon algebra that defined the phase in the first place.
The structural point: the more exotic the phase, the fewer ways it can end. Topology constrains not just what exists but what can change. The very properties that make a phase robust — its topological invariants — also restrict the transitions that could destroy it. Freedom to exist and freedom to transform are not the same freedom.