M-theory on a circle gives Type IIA string theory. M-theory on an interval gives heterotic string theory. The compactification geometry determines which string theory emerges in ten dimensions. The standard compactifications produce supersymmetric theories.
The paper considers M-theory on S¹ ∨ S¹ — the wedge of two circles, which is a figure-eight: two loops sharing a single point. This is a singular space — the wedge point is not a manifold point — but the singularity is mild enough that the compactification is well-defined.
The result: Type 0A string theory. This is the non-supersymmetric cousin of Type IIA, obtained by projecting out all spacetime fermions. Type 0A has a tachyon, no supersymmetry, and a doubled set of Ramond-Ramond fields compared to Type IIA. It's the most basic non-supersymmetric string theory, and its M-theory origin has been unclear.
The figure-eight compactification provides that origin. The two circles contribute independent winding sectors, and the identification at the wedge point implements the projection that removes fermions. Supersymmetry is broken not by adding anti-branes or turning on fluxes but by the topology of the compactification space itself. The figure-eight is the simplest topology that breaks supersymmetry through geometry alone.
Non-supersymmetric string theories are less studied because they're unstable (the tachyon), but their M-theory origins constrain their structure. The figure-eight origin is the simplest possible — one singular point, two circles — and the simplicity of the geometry maps to the simplicity of the resulting string theory.