- #36
julian
Gold Member
- 829
- 326
Hello tom
I have a limited knowledge of the issues of differentiable structure of spatial diffs in LQG. I know that if the valence of the nodes is great enougth that using the smooth diff structure makes the Hilbert space non-seperable (http://arxiv.org/pdf/gr-qc/0403047.pdf):"Indeed, as we show below, the nodes of sufficiently high valence have a surprising “rigidity” under smooth transformation, and this rigidity turns out to be the one responsible for the moduli. Therefore the non-separability of [itex]H_{diff}[/itex] is a bizarre remnant of the initial choice of the smooth category. It is therefore natural to explore the possibility of using a slightly different functional class of fields to start with."I know in the LOST theorem that they consider piecwise analytic structures. This is to avoid the union of two graphs having an infinite number of edges (if piecewise analytic curves intersect at least a countable number of times they will coincide everywhere) - it is crucial that they be piecwise becuase otherwise everything would be determined by the data in an arbitrarily small region (analyticity) and there would be no local degrees of freedom.
I'd be interested to hear more about what you think about the whole issue. Maybe you are right about topology change and diff structure in LQG. Was this not part of the motivation for Thiemann's Algebriac quantum gravity where there is no fumdamental topology or differential structure?
I have a limited knowledge of the issues of differentiable structure of spatial diffs in LQG. I know that if the valence of the nodes is great enougth that using the smooth diff structure makes the Hilbert space non-seperable (http://arxiv.org/pdf/gr-qc/0403047.pdf):"Indeed, as we show below, the nodes of sufficiently high valence have a surprising “rigidity” under smooth transformation, and this rigidity turns out to be the one responsible for the moduli. Therefore the non-separability of [itex]H_{diff}[/itex] is a bizarre remnant of the initial choice of the smooth category. It is therefore natural to explore the possibility of using a slightly different functional class of fields to start with."I know in the LOST theorem that they consider piecwise analytic structures. This is to avoid the union of two graphs having an infinite number of edges (if piecewise analytic curves intersect at least a countable number of times they will coincide everywhere) - it is crucial that they be piecwise becuase otherwise everything would be determined by the data in an arbitrarily small region (analyticity) and there would be no local degrees of freedom.
I'd be interested to hear more about what you think about the whole issue. Maybe you are right about topology change and diff structure in LQG. Was this not part of the motivation for Thiemann's Algebriac quantum gravity where there is no fumdamental topology or differential structure?
Last edited: