In the Copenhagen interpretation, this interference is suppressed by measurement, via instantaneous “collapse” of the wave function to a given state.

general introduction to decoherence, in the context of closed quantum systems. Stress the difference with the classical situation in which a single history of the system is realized and is selected by the theory. It is difficult to assign probabilities to histories, in the quantum case, because of interference, mathematically the fact that we work with probabilities amplitudes for states and histories, and not probabilities themselves.

More on decoherence; focus on the histories of spin networks , that is spin foams. Main point: how to assign probabilities to such histories. Importance of decoherence in providing these quantities.

Butterfield: superselection is actually the more general concept and mechanism to be studied, decoherence being one particular way in which such superselection is achieved by coupling with the environment. Speaker agrees, but stresses how this point of view rests to some extent on the general attitute towards quantum mechanics and its relation with the classical world.

Comment by Butterfield: importance of superselection in achieving classicality.

More questions about how the loop quantum gravity case is different or peculiar with respect to other formalisms or other physical systems [less exotic that quantum gravity), with respect to the application and role and interpretation of decoherence. Answer from speaker: indeed, it may be not so peculiar, but the role of decoherence [how superposition is suppressed, how a special set of observables is selected] in this context remains most interesting.

Suggestion from the audience: may be interesting to look at work on “temporal decoherence”.

Butterfield: the main point of decoherence is that when we model the interaction with the environment, the suppression of superposition [diagonalization] is rapid and inevitable; indeed, which is the selected basis and which correlations are suppressed is not given by the theory and is to some extent a choice in modelling of the system. Sudarsky: true, but there exists contexts [cosmology] where no experimenter and no observer can be associated with that “choice”.

Stressing how the way decoherence works is, to some extent, up to us, in the sense that it depends on the choice of what are the degrees of freedom that one treats as environment and which ones one treats as “the system” that should decohere. Some discussion with the audience on how much freedom one really has, or, more bluntly, on whether the whole process is “observer-dependent” and up to the conscious choice of the physicist.