He gives a survey of several programmes’ proposals for the quantum building blocks of spacetime. He begins with (i) Loop Quantum Gravity (LQG).
Turning to some details:
(i) Loop Quantum Gravity (LQG):
This started off as a canonical quantization of continuum GR reformulated in terms
of first order variables (namely a Lorentz connection and a triad field). The quantization gives a direct sum of Hilbert spaces associated to graphs. For each given graph γ , a generic state is given by a wave function whose argument is an assignment of data to the elements of the graph. In one scheme, these are: group elements ge representing holonomies of the gravitational connection along the links e =1, …,E of the graph (considered as embedded in the spatial manifold), so that the state is ψ(g1, …., gE ). In another scheme: one labels the links with representations of the Lorentz (or rotation) group.
Dynamics is given: (a) canonically, by a Hamiltonian constraint operator defined on graphs; (b) covariantly, by defining histories as graphs considered as embedded in spacetime (spin foams), and assigning amplitudes to histories, and then summing over amplitudes.
Recently the schemes have emphasized: (a) simplices, (b) thinking of the graph as dual to a (usually simplicial) cellular complex; (c) motivating the proposed assignment of amplitudes by considering lattice version of the gravitational path integral. This leads to: