“…We are led to the conclusion that it is not possible to make a momentum-spin separation of the system and to define a reduced density matrix for spin, as was done in many previous papers on the subject [2, 4, 9, 10, 12, 15-21, 23, 24]. We also show that the use of the Pauli-Lubanski (or similar) spin operators to describe spin measurements, as in [1,5,8,11,13,14,16,22,23], depends on the coupling of the spin to a quantity that transforms as part of a four-vector under the Lorentz transformations in the measuring apparatus. However, we do not know if such a coupling exists in nature.…”