Cerebellar sequencing: a trick for predicting the future

Cerebellum. 2015 Feb;14(1):35-8. doi: 10.1007/s12311-014-0616-x.

Abstract

"Looking into the future" well depicts one of the most significant concepts in cognitive neuroscience: the brain is constantly predicting future events. Such directedness toward the future has been recognized to be relevant to and beneficial for many aspects of information processing in humans, such as perception, motor and cognitive control, decision-making, theory of mind, and other cognitive processes. Because one of the most adaptive characteristics of the brain is to correct errors, the ability to look into the future represents the best chance to avoid repeating errors. Within the structures that constitute the "predictive brain," the cerebellum has been proposed to have a central function, based on its ability to generate internal models. We suggested that "sequence detection" is the operational mode of the cerebellum in predictive processing. According to this hypothesis, the cerebellum detects and simulates repetitive patterns of temporally or spatially structured events and generates internal models that can be used to make predictions. Consequently, we demonstrate that the cerebellum recognizes serial events as a sequence, detects a sequence violation, and successfully reconstructs the correct sequence of events. Thus, we hypothesize that pattern detection and prediction and processing of anticipation are cerebellum-specific functions within the brain and that the sequence detection hypothesis links the multifarious impairments that are reported in patients with cerebellar damage. We propose that this cerebellar operational mode can advance our understanding of the pathophysiological mechanisms in various clinical conditions, such as schizophrenia and autism.

Publication types

  • Review

MeSH terms

  • Anticipation, Psychological / physiology*
  • Cerebellum / physiology*
  • Cerebellum / physiopathology
  • Humans
  • Pattern Recognition, Physiological / physiology*
  • Serial Learning / physiology*