It is universally recognized that NMR measures the shielding of nuclei from magnetic fields by electrons. The electrons move, producing an eddy current that produces an opposing magnetic field.
It is also fairly widely said that a filled valence shell contains electrons making every possible motion within a certain range of energy. For example the p orbitals are added up using +1 and -1 values for angular momentum, and the electron spins are paired. MO theory seems, in most cases, to have the same paradox.
Now I know you can look into it and find more detailed math for NMR tensors, but at the intuitive level: what is going on here? (Are the filled orbitals being hybridized, slightly, with ordinarily unfilled complex-valued higher-energy orbitals to permit the electrons to take on some degree of net overall angular momentum in relation to the axis of the magnetic field, or is it something else?)