This paper studies whether a strategy called symmetric mass generation (SMG) can remove unwanted “mirror” fermions on a space-time lattice without breaking the symmetries that one wants to keep. Mirror fermions are extra particles that appear in naive lattice versions of chiral theories. The long-standing Nielsen–Ninomiya theorem says that, under broad and natural conditions for free fermions on a lattice, every massless fermion of one handedness comes with a partner of the opposite handedness. The authors ask whether that obstruction can be overcome when strong, non-gauge interactions are allowed.

Their basic idea is to build an effective one-particle lattice Hamiltonian from the interacting theory and then check whether this effective Hamiltonian satisfies the assumptions of the Nielsen–Ninomiya argument. Concretely, they define the effective Hamiltonian Heff as the inverse of the retarded fermion propagator evaluated at zero frequency. If Heff lives on the same compact momentum space, is local enough to have continuous first derivatives, respects the symmetry charges without spontaneous symmetry breaking, and the continuum limit contains only free relativistic massless fermions (no massless bosons), then the Nielsen–Ninomiya conclusion applies: any massless spectrum must be vector-like, i.e., left- and right-handed modes pair up.

A key technical point concerns “propagator zeros.” When interactions gap a mirror pole in the fermion two-point function, the propagator may develop a zero in momentum space. Such zeros come in two kinds. A “genuine” zero would make Heff singular in a way that signals non-unitary behaviour (ghosts). A more benign possibility is a “kinematical” zero, which indicates that the gapped mirror has become part of a massive Dirac particle made from the original mirror plus a new bound-state partner. The authors argue that for local lattice models genuine zeros are unlikely, so SMG is most plausibly realized by forming bound states for each gapped mirror.