Lattice QCD, gauge fixing and the transition to the perturbative regime

Abstract

The standard definition of perturbative QCD uses the Faddeev-Popov gauge-fixing procedure, which leads to ghosts and the local BRST invariance of the gauge-fixed perturbative QCD action. In the nonperturbative regime, there appears to be a choice of using nonlocal Gribov-copy free gauges (e.g. Laplacian gauge) or of attempting to maintain local BRST invariance at the expense of admitting Gribov copies and somehow summing or averaging over them. It should be recognized that the standard implementation of lattice QCD corresponds to the former choice even when only physical (i.e. colour singlet) observables are being calculated. These issues are introduced and briefly explained.