Abstract
A simple proof is given of a recent claim [P. Pearle, Phys. Rev. A 39, 2277 (1989)] that appropriately coupled classical white noise reduces wave packets of macroscopic objects. Three difficulties with the classical mechanism are noted. All three are resolved by passing to the quantum analog. The 2.7-K microwave background is effective in some cases. Quantum noise reduces the density matrix of observables, not the wave function itself. The relative merits and liabilities of these two kinds of reduction, called von Neumann reduction and Heisenberg reduction, respectively, are discussed. In this connection a new type of correlation experiment, devised by Hardy [Phys. Rev. Lett. 68, 2981 (1992)] is used to prove a theorem that says, essentially, that any ontological solution of the quantum measurement problem must admit either parallel worlds of the Everett kind or faster-than-light influences of some kind.