COMMENTS ON INFORMATION LOSS AND REMNANTS

The information loss and remnant proposals for resolving the black hole information paradox are reconsidered. It is argued that in typical cases information loss implies energy loss, and thus can be thought of in terms of coupling to a spectrum of ''fictitious'' remnants. This suggests proposals for information loss that do not imply Planckian energy flucutations in the low energy world. However, if consistency of gravity prevents energy nonconservation, these remnants must then be consiered to be real. In either case, the catastrophe corresponding to infinite pair production remains a potential problem. Using Reissner-Nordstrom black holes as a paradigm for a theory of remnants, it is argued that couplings in such a theory may give finite production despite an infinite spectrum. Evidence for this is found in analyzing the instanton for Schwinger production; fluctuations from the infinite number of states lead to a divergent stress tensor, spoiling the instanton calculation. Therefore naive arguments for infinite production fail.