The hippocampal formation is central to the consolidation and retrieval of long-term declarative memory, memories that are stored throughout the neocortex with putative subcortical participation (Berger et al. 2011; Corkin 2002; Deadwyler et al. 2016; Hikosaka et al. 2014; Kim and Hikosaka 2013; Scoville and Milner 1957; Squire and Knowlton 2000; Tehovnik, Hasanbegović, Chen 2024; Wilson and McNaughton 1994). Subjects that have hippocampal damage have great difficulty narrating stories (Hassabis et al. 2007ab), which can be viewed as a disruption of one’s stream of consciousness as it pertains to retrieving information. The retrieved stories, which are highly fragmented in hippocampal patients (Hassabis et al. 2007ab), are comparable to those evoked electrically by stimulating a single site in the parietal and temporal lobes (Penfield and Rasmussen 1952; Penfield 1958, 1959, 1975). Nevertheless, individuals with hippocampal damage can still engage others verbally, but the conversation is limited in that it is based on declarative memories that are not updated making the hippocampectomized interlocker seem out of touch (Corkin 2002; Knecht 2004). A rapid exchange of speech is dependent on an efference-copy representation, which is mediated through the cerebellum (Bell et al. 1997; Chen 2019; De Zeeuw 2021; Guell, Schmahmann et al. 2018; Loyola et al. 2019; Miles and Lisberger 1981; Noda et al. 1991; Shadmehr 2020; Tehovnik, Patel, Tolias et al. 2021; Wang et al. 2023).

Patient HM, who had bilateral damage of his hippocampal formation, had ‘blind memory’ (much like ‘blindsight’): when asked to name the president of the United States in the early 2000’s he failed to recall the name, but when given a choice of three names: George Burns, George Brown and George Bush he was able to select George Bush (Corkin 2002). Therefore, his unconscious stores of information were intact (which is also true of blindsight for detecting high-contrast spots of light, Tehovnik, Patel, Tolias et al. 2021). As well, HM had memory traces of his childhood (a time well before his hippocampectomy), but the specifics were lost such that he could not describe even one event about his mother or father (Corkin 2002). Although many presume that HM had memories of his childhood, these memories were so fragmented and lacking in content that referring to his childhood recollections as ‘long-term memories’ is questionable.

The idea that the brain becomes less active once a new task has been acquired through learning can be traced back to the experiments of Chen and Wise (1995ab) that were done in the supplementary motor area, Brodmann’s Area 6. Monkeys were trained to associate a visual image with a particular direction of saccadic eye movement, which could be up, down, left, or right of a centrally-located fixation of the eyes. For a significant proportion of neurons studied it was found that the activity of the cells decreased with overlearning an association. At the time of publication this counter-intuitive result was greeted with much skepticism. After reading the paper, Peter Schiller did not know what to make of the result since his results (seven years before) suggested that the supplementary motor area becomes more active and engaged once new tasks are learned (Mann, Schiller et al. 1988).

Years later, Hikosaka and colleagues continued this line of work to show that the diminution of activity with learning was a real neural phenomenon and that the diminished information was channeled to the caudate nucleus (Hikosaka 2019; Hikosaka et al. 2014; Kim and Hikosaka 2013), which is connected anatomically to the entire neocortex such that the head of the caudate innervates the frontal lobes whereas the tail of the caudate innervates the temporal lobes (Selemon and Goldman-Rakic 1985). Hikosaka (2019) has proposed that the memories of learned tasks are archived in the caudate nucleus, whereby new tasks are stored in the head of the caudate and old tasks are stored in the tail of the caudate—perhaps for immediate use by the temporal lobes which if damage disrupts long-term memories even of one’s childhood (Corkin 2002; Squire et al. 2001).

That neurons throughout the brain (i.e., the cortex and subcortex) become less responsive to task execution once overlearned is a well-established fact (Lehericy et al. 2005). We have argued that this diminution of responsivity is the brain’s way of consolidating learned information efficiently, while reducing the energy expended for the evocation of a learned behavior (Tehovnik, Hasanbegović, Chen 2024). We and others (Lu and Golomb 2023) believe that all memories are stored according to the context of the memorization, which requires that a given site in the neocortex that contains a memory fragment such as a word or visual image be networked with other neurons to recreate the context, which we refer to as a declarative/conscious unit (Tehovnik, Hasanbegović, Chen 2024). When someone narrates a story, declarative/conscious units are concatenated in a string much like the serialization of the images of a film and this process involves both the neocortex and the cerebellum (Hasanbegović 2024).

Furthermore, a primary language (as compared to secondary languages) is stored in the neocortex and cerebellum in such a way that any damage to either structure often preserves the primary language while degrading the secondary languages (Mariën et al. 2017; Ojemann 1983, 1991; Penfield and Roberts 1966). All languages are networked separately in the brain (Ojemann 1991): a unique neocortical-cerebellar loop is summoned during the delivery of a speech in the chosen language (Tehovnik, Hasanbegović, Chen 2024). The language one thinks in (i.e., one’s counting language) is the language that is well archived and highly distributed (including areas of the brain that mediate mathematics), thus making the language more resistant to the effects of brain damage.

In conclusion, information stored in the brain is no different from information stored in a university library: the ancient texts are all housed in a special climate-controlled chamber, while the remaining texts including the most recent publications are made available to all students and professors. Indeed, it is our childhood memories that define us and therefore they deserve to be archived and protected in the brain. The details of how this happens will need to be disclosed.