“…The starting point for this progress was the suggestion by Rovelli [8,9] of a procedure for computing the graviton propagator from Loop Quantum Gravity (LQG) [1,10,11] with the dynamics implemented covariantly in terms of a spin foam model. On the calculational side, the key ingredients are a boundary semiclassical spin network state peaked on large spins and an analytic expression for the large spin asymptotics of the spin foam vertex amplitude [12]; on the conceptual side, the framework is the boundary state formalism discussed in [1] and in [13,14,15,16,17], which prescribes how to compute observables in the boundary of a spacetime region with a path integral over the interior region only. IfÔ 1 ,Ô 2 are local boundary geometry observables (such as areas, dihedral angles, 3-volumes or lengths [18,19,20,21,22,23]) acting on a space of spin networks s , then the expectation value for their correlation in a boundary geometry q is given in the boundary state formalism by…”