Abstract

I argue that the rôle for the cerebellar cortex is in the generation of sensory predictions, not motor sequences. This proposal may explain the allometric relationship described in Braitenberg et al.'s target article. I also point out that the parallel beam organisation may have a nontemporal basis