Demonstration of quantum correlations that are incompatible with absoluteness of measurement

Abstract

Exploiting the tension between the two dynamics of quantum theory (QT) in the Wigner's Friend thought experiment, we point out that the standard QT leads to inconsistency in observed probabilities of measurement outcomes between two superobservers: Wigner and his Student. To avoid such inconsistent predictions of QT, we hypothesize two distinct perspectives. The first one is ``absoluteness of measurement (AoM),'' that is, any measurement process is an absolute event irrespective of other observers and yields a single outcome. The other is ``nonabsoluteness of measurement (NoM)'' as the negation of AoM. We introduce an operational approach, first with one friend and then with two spatially separated friends, to test the validity of these two perceptions in quantum theory without assuming the details of the experiment. First, we show that the set of probabilities obtainable for NoM is strictly larger than the set obtainable for AoM. We provide the simplest scenario so far, involving a single quantum preparation and one unitary operation by a superobserver that can demonstrate correlations incompatible with AoM. Remarkably, in the scenario with spatially separated observers, we present a strict hierarchy among the sets of probabilities observed in the following three theories: classical or local realist, quantum theory with AoM, and quantum theory with NoM.