Usually, we associate the half-integer spin to fermion, and the integer spin to boson. And there are constraints like the Spin-Statistics relations.
However, in Spin–charge separation or the parton-construction techniques (Schwinger/Abrikosov boson/fermions, see also other webpages), we can separate a spin-1/2 charge-1 fermion to a spinon (with spin-1/2) and a holon (or called chargeon, with a charge-1).
It terms out that a spinon (with spin-1/2) in Spin–charge separation may be:
(1) bosonic spinon (with spin-1/2)!
(2) fermionic spinon (with spin-1/2).
In (1), even a spin-1/2 object called the spinon can be a boson!!!
I wonder what are the constraints on the Spin-charge separation, and how does this tie to the Spin-Statistics relations (in any dimensions, especially in 2+1d and 3+1d)?