The hippocampus is necessary for the consolidation of recent memories, but the neocortex is important for the archival storage of these memories. This is borne out by studies in both humans and other animals (Frankland and Bontempi 2005). Replay during sleep, particularly during slow-wave sleep, is believed to be involved in the consolidation process at the neocortex (Dickey et al. 2022; Wilson and McNaughton 1994). Recall by mice of remote memories activates the prefrontal cortex, the frontal cortex, the anterior cingulate, the retrosplenial cortex, and the temporal cortex as evidenced with 14C-2-deoxyglucose (Bontempi et al. 1999), and expression of activity by the genes, c-Fos and Zif268 (Maviel et al. 2004). Initially, the hippocampus is thought to integrate information from distributed and independent cortical modules that represent various features of a task and then these features are fused to produce a coherent memory trace by strengthening connections between areas (Frankland and Bontempi 2005).
Imaging studies in animals have shown that hippocampal activity is suppressed when spatial and contextual memories are being recalled (Bontempi et al. 1999; Maviel et al. 2004), i.e., when stored memories are being combined with ongoing behavior. As well, the retrieval process is believed to utilize non-NMDA/glutamatergic circuits in the hippocampus (Nakazawa and Tonegawa 2004; Riedel et al. 1999). When neocortical memory is inconsistent with a new hippocampal memory, the hippocampus is re-activated to upgrade the stored neocortical memory (Frankland and Bontempi 2005), much like what happens for memory at the cerebellum: when there is a mismatch between the sensory inputs flowing via the mossy fibres and the inputs from the inferior olive, the learning process is reinstated [i.e., the complex spike firing is put outside of the 0.5-2 Hz window to alter simple-spike discharge, Loyola, De Zeeuw et al. 2019; Mukamel, Schnitzer et al. 2009; Tehovnik, Patel, Tolias et al. 2021].
For the sake of simplicity, it is best to think of neocortical memory as ‘sensory’ and cerebellar memory as ‘motor’, with the former having a sensory bias (i.e., visual, auditory, somatosensory, vestibular, gustatory, olfactory, interoceptive) and the latter having a motor bias, represented by proprioception which is the sensory counterpart of ocular and skeletal movement (Chen 2019; Gibson et al. 2004; Tehovnik and Chen 2015). Cerebellar parallel fibres (from the granular neurons) interconnect the neocortex with the cerebellum via descending pontine information from the neocortex and the ascending thalamic information to the neocortex. Such loops of information are engaged for the execution of all behavior (Tehovnik, Hasanbegović, Chen 2024), even though reductionists have disconnected various component of the cerebellum to suggest that the cerebellar nuclei and the brain stem can operate independently of the cerebellar cortex once learning has been finalized (e.g., Miles and Lisberger 1981; Sendhilnathan and Goldberg 2000b). Such a contradiction (not that different from the ‘wave’ vs. ‘particle’ contradiction in physics) can best be appreciated by what happens during the generation of normal behavior: learning is never finalized since the minute you wake up in the morning, the brain is being subjected to new adaptive forces. And if you don’t believe me, try going to work with the exact same thoughts (Hebb 1949, 1961, 1968) as you had yesterday. We can all agree that on each day our brain experiences a newness that alters consciousness while preserving the historical record of self.
You cannot enter the same river twice, but the river remains a river
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication