Bell [Helv. Phys.Act. 48 1975] described an idealized test of the projection postulate in order to illustrate the ambiguities in the experimental predictions of quantum theory. An approach to the implementation of such tests is outlined here, based on the interesting logical relationship between the projection postulate and the no-superluminal-signaling principle. Gisin's theorem [Helv. Phys. Act. 62 1989] shows that the no-signaling principle implies that any description of projection at the elementary level must be nondeterministic. The lack of determinism means that information cannot be instantiated in a reproducible and transmittable form in isolated microsystems. The implied limitation on information allows one to interpret the no-signaling principle in purely physical terms as a prohibition of superluminal information transmission. This prohibition, in turn, imposes tight constraints on the connection between projection and elementary processes, leading to a specific hypothesis about how projection effects originate. The hypothesis serves as the basis for the experimental proposal.