Leibniz has the third take and was probably right, as Herder suggested. James was on that track a century later. Now, yet another century later, the consciousness seems to be lodged in the claustrum and chooses whether to go ahead with motion the body suggests. Usually it says 'not,' which makes free will more of 'free won't.' That is, the consciousness decides to not do or lets the motion proceed. Cartesian duality is out.
Interesting empirical tidbit in this regard -- neurosurgen Frank Vertosik, in his memoirs When The Air Hits Your Brain, speaks of experiments in which most of the brains of housecats were removed, and then the cats returned to their owners. Most of the owners could not tell any difference in the cats' behaviour.
For the nonce, though, this treatment of the mind -- at least as you have described it -- doesn't explain hesitation, confusion, memory, or learning. Is there more to the story?