This thread is a direct shoot-off from this post in the Insights thread Against "interpretation" - Comments. I am usually not a big fan of Ballentine, but I tend to fully agree with him on the following issue (taken from this paper, credits to @bhobba) Decoherence theory is a pragmatic approach based on the density matrix, which in the words of John Bell merely works 'for all practical purpose' (FAPP). The problem with the density matrix approach to the measurement problem is exactly that the ontology of the density matrix is never made clear. Multiple theoreticians, mathematicians and workers in the foundations of QM have made the point that the density matrix approach doesn't seem to resolve the measurement problem, but instead merely shifts the burden of ontology from the wavefunction onto the density matrix and so doesn't bring us any further w.r.t. QM's most important foundational issue, i.e. that of resolving the measurement problem. Indeed, the degree of ontology of this matrix is in a sense contingent upon our degree of technological prowess in experimental QM. A somewhat simplified way to describe this issue is to say that a solution which merely works FAPP (e.g. from the viewpoint and standards of applied or experimental physics), is precisely one that need not and generally does not work in principle (e.g. from the viewpoint and standards of foundational and theoretical physics).