Nullstein
- 313
- 202
It might as well become a success story, like quantizing classical electrodynamics. Nobody can tell at this point in time. At least, it's a reasonable approach, no less reasonable than quantizing classical electrodynamics. But it has to be done well. Sure it may fail, but it's worth a try.haushofer said:An intermediate step, like quantizing Fermi's theory of the weak interactions :P
They have no immediate physical meaning. What you do in LQG is to start from the vielbein formulation of GR, where the basic variables are the frame field and the spin connection. This is reasonable, one needs to do it anyway to allow for the inclusion of spinor fields. Then you make a 3+1 split, which is also reasonable if you want to obtain a Hamiltonian formulation. But then you go ahead and form new variables by adding the spin connection to the extrinsic curvature of the spatial slices. How is this a reasonable physical quantity? It's like adding apples and oranges and only accidentally works in 3 dimensions (because the adjoint representation of ##SO(3)## is equivalent to the defining representation). Morover, one introduces a new parameter (the Immirzi parameter), which classically cancels out, but remains important in the quantum theory. With these new variables, many equations simplify or become more elegant. The theory then looks like a Yang-Mills theory with additional constraints, but at the cost of having had to add apples to oranges in an early step of the calculation.haushofer said:Can you e laborate on these canonical variables and what's odd about them?