Is autism due to cerebral-cerebellum disconnection?

Autism has been linked to thalidomide exposure at 20-24 days gestation. At this stage, the embryo is roughly the size of this 'C', and has yet to develop its brain (except for brainstem cranial motor nerve nuclei). The neuropathology responsible for autism is presently unknown, but whatever it is, it must logically be one that can be induced by such an early occurring brainstem cranial motor nerve nuclei defect. Many mental faculties impaired in autism (such as theory of mind) depend upon the prefrontal cortex. The maturation of cerebral-cerebellar connections, due to oddities in axon development, is vulnerable to pre-existing brainstem nuclei integrity. Many higher cognitions (including prefrontal ones) are dependent upon these links raising the possibility that abnormalities in them might produce autism. I conjecture that impaired cerebral-cerebellar connections, whether caused early, as by thalidomide, or later (including postnatally) by other factors, is the missing neuropathological cause of autism.