In his comment on Emil Mottola and Ruslan Vaulin’s response (Physics Today, November 2013, page 9) to his piece on black holes and information theory (Physics Today, April 2013, page 30), Steve Giddings makes several statements that are simply wrong. From the point of view of quantum mechanics, black holes cannot exist because quantum mechanical evolution does not allow for the destruction of information and because black hole space-times do not provide a universal time.
As noted by Mottola and Vaulin, conflict with the information nondestruction principle can naturally be avoided if event horizons never form in the real world. Although Giddings is correct that this is an unpopular idea in the theoretical-physics community, the more interesting question for the general physics community is, What do astrophysical observations say about the existence of event horizons? As it happens, in contrast with what Giddings implies in his response to Mottola and Vaulin, at the present time there is no astrophysical evidence that event horizons exist in nature.
It is, of course, incontrovertible that compact objects exist whose size approximates the event-horizon radius predicted by classical general relativity. The pregnant issue is whether matter falling onto the surface of such an object encounters an event horizon where nothing remarkable occurs or whether it encounters a real surface. One might consider distinguishing an event horizon from a real surface by observing whether matter falling onto a compact object produces x rays. In fact, x-ray bursts from compact objects orbiting stars are dramatically smaller when the object’s mass is greater than the maximum mass of a neutron star.1 However, the bursts from neutron stars are due to the interaction of in-falling matter with stationary nuclear matter on the surface, and in no model for compact objects with a mass exceeding the maximum neutron-star mass is it possible for nuclear matter at the surface to be stationary.
In the case of compact objects bounded by a quantum critical surface, nucleons are transformed into leptons and gamma rays when they fall onto the surface. The destruction of baryons at the surface could provide a simple test as to whether the surface is an event horizon.2 The observed spectrum of gamma rays from neutral pions produced by the interaction of cosmic rays with interstellar dust is symmetric about the mean boosted energy of the pions, reflecting the gamma rays’ redshift or blueshift that results from the motion of the pions. However, if the pions are produced at the surface of a compact object, the redshifted component is absent, which results in an asymmetric gamma-ray spectrum—a distinct signature for the absence of an event horizon. The best opportunity for seeing this signal may arise from matter falling onto Sagittarius A*. Although the background of other gamma-ray sources in the central region of our galaxy makes it difficult to see this signature under ordinary circumstances, the Fermi Large Area Telescope may soon be able to detect the feature because of the infall of a gas cloud that is now approaching SgA*.
Matter falling onto the surface of a compact object also heats the object. However, in contradiction with the paper quoted by Giddings3 in his response to Mottola and Vaulin, in the case of a massive object like SgA* the thermal radiation is far below the limits of reference 3 because the specific heat of compact objects with a quantum critical surface is enormously larger than the conventional values the researchers assumed.
