The 'information paradox', as I understand it naïvely, points out that there is a conflict between i) the 'classical' picture of a black hole, in which information passed through the event horizon will be lost after the phenomenon of Hawking evaporation; and ii) the 'quantum' or maybe more accurately put the 'semi-classical' picture in which a black whole seen as a system in quantum-field theory should obey a unitary evolution in its Hilbert space of states, which precludes loss of information.

I have difficulties to see the paradox here for two reasons, one probably not very serious, the other more serious. First reason: the classical (and even the semi-classical) pictures of a black hole are non-definitive as long as we don't know how to combine gravitation and quantum effects into a sound theory. Of course one might reply that this is precisely the point made by the paradox.

More serious reason: the unitary evolution in quantum mechanics holds only as long as no measurement is made on the system. Wave function collapse phenomena (which are non-unitary and even nonlinear) appear to be put under the rug in the statement of the paradox. What happens to a black hole when it is observed (e.g. how decoherence might play a role) ? Is there a discussion of this question and how it affects the information paradox to be found in the literature?