Long range correlations in quantum gravity

Smolin has pointed out that the spin network formulation of quantum gravity will not necessarily possess the long range correlations needed for a proper classical limit, typically, the action of the scalar constraint is too local. Thiemann's length operator is used to argue for a further restriction on the action of the scalar constraint: it should not introduce new edges of color unity into a spin network, but should rather change preexisting edges by +/- one unit of color. Smolin has proposed a specific ansatz for a correlated scalar constraint. This ansatz does not introduce color unity edges, but the [scalar, scalar] commutator is shown to be anomalous. In general, it will be hard to avoid anomalies, once correlation is introduced into the constraint, but it is argued that the scalar constraint may not need to be anomaly-free when acting on the kinematic basis. [S0556-2821(99)07704-8].