It is known that the RVB states can support spin-charge separations and its elementary excitations are spinons and holons. But it seems that there are some different possibilities for the nature of the statistics of the spinons and holons. Specifically, (1) spinons are fermions while holons are bosons(Kivelson et al);(2)spinons are bosons while holons are fermions(Read et al);(3)spinons and holons are both semions obeying fractional statistics(Laughlin et al).
Do we now have a commonly accepted view for one of the above three possibilities? Or do the quasiparticle statistics depend on the details(e.g. which kind) of the RVB wave functions?
Furthermore, assume RVB states as the ground-states of the antiferromagnetic Heisenberg model, and when we use $t-J$ model to study the lightly doped Mott insulator, for the possibility (1), it's reasonable to adopt slave-boson formalism, while for the possibility (2), it's reasonable to adopt slave-fermion formalism, right? But for the possibility (3), what kind of slave-particle formalism we should adopt? This part may be related to my previous question.
Thanks in advance.