On Background Independence

Abstract

This paper concerns what Background Independence itself is (as opposed to some particular physical theory that is background independent). The notions presented mostly arose from a layer-by-layer analysis of the facets of the Problem of Time in Quantum Gravity. Part of this coincides with two relational postulates which are thus identified as classical precursors of two of the facets of the Problem of Time. These are furthemore tied to the forms of each of the GR Hamiltonian and momentum constraints. Other aspects of Background Independence include the algebraic closure of these constraints, expressing physics in terms of beables, foliation independence as implemented by refoliation invariance, the reconstruction of spacetime from space. The final picture is that Background Independence - a philosophically desirable and physically implementable feature for a theory to have - has the facets of the Problem of Time among its consequences. Thus these arise naturally and are problems to be resolved, as opposed to avoided `by making one's physics background-dependent in order not to have these problems'. This serves as a selection criterion that limits the use of a number of model arenas and physical theories.