An exceptionally simple argument against the many-worlds interpretation

It is shown that the superposed wave function of a measuring device, in each branch of which there is a definite measurement result, does not correspond to many worlds as assumed by the many-worlds interpretation, because all branches of the superposed wave function can be observed in our world by protective measurement. According to the many-worlds interpretation of quantum mechanics, each branch of the wave function of a measuring device in which there is a definite measurement result corresponds to each world among the many worlds (see, e.g. Vaidman 2008; Barrett 2011). This means that in one world there is only one branch of the superposed wave function in which there is a definite measurement result, and the other branches do not exist in this world. Therefore, according to the many-worlds interpretation, in every world the whole superposed wave function of the measuring device cannot be measured. If all branches of the superposed wave function of the device can be observed in one world, then they will all exist in this world, which obviously contradicts the many-worlds interpretation. It is unsurprising that the existence of such many worlds may be consistent with the results of conventional impulse measurements, as the many-worlds interpretation is just invented to explain the emergence of these results, e.g. the definite measurement result in each world always denotes the result of a conventional impulse measurement. However, this does not guarantee consistency for all types of measurements. It has been known that there exists another type of measurement, the protective measurement (Aharonov and Vaidman 1993; Aharonov, Anandan and Vaidman 1993; Aharonov, Anandan and Vaidman 1996; Vaidman 2009). Like the conventional impulse measurement, protective measurement also uses the standard measuring procedure, but with a weak, adiabatic coupling and an appropriate protection. Its general method is to let the measured system be in a nondegenerate eigenstate of the whole Hamiltonian using a suitable protective interaction, and then make the measurement ∗Unit for HPS and Centre for Time, University of Sydney, NSW 2006, Australia. E-mail: [email protected]. 1It should be noted that the consistency is still debated. For more discussions see Saunders et al (2010) and references therein.