According to Thach et al. (1992) there are three homuncular representations in the cerebellum: one at the anterior lobe that innervates the fastigial nucleus, a second at the posterior lobe that innervates the interpositus nucleus, and a third at the mediolateral lobe that innervates the dentate nucleus. The three nuclei then project to the ventrolateral nucleus of the thalamus in overlapped, topographic register to innervate both M1 (see Fig. 1) and S1, both of which are involved in issuing motor commands (volitional as well as automatic) directed toward the brain stem and spinal cord (Evarts 1966, 1968; Georgopoulos et al. 1984, 1986) and sending via neocortical collaterals (Hasanbegović 2024) an efference-copy signal to the cerebellar cortex and its nuclei (Bell et al. 1997; De Zeeuw 2021; Loyola et al. 2019; Shadmehr 2020; Tehovnik et al. 2021; Wang et al. 2023).
Electrical stimulation of the cerebellum (i.e., either the anterior or posterior cerebellar lobes, see Fig 2 for known cerebellar topography in humans) can suppress body movements evoked electrically from the neocortex and reflexive movements evoked externally in cats, dogs, monkeys, and chimpanzees (Nims and Nulsen 1947; Nulsen et al. 1948; Snider et al. 1947). The suppression is body-part specific and depends on the part of a cerebellar homunculus being activated (Snider et al. 1947). As well, stimulation can inhibit specific muscles (Nulsen et al. 1948).
Wilder Penfield (who studied under Charles Sherrington) is recognized for being the first to use electrical stimulation in the neocortex of human subjects to arrest language functions and his work has been instrumental is disrupting both the sensory and motor aspect of language (Penfield and Roberts 1966; also see Ojemann 1991). Given that the cerebellum has been added to the long list of cognitive centers including its role in language (Schmahmann 1997; see Fig. 3), it would be instructive to see if electrical stimulation of the homunculi of the cerebellar cortex can suppress speaking and writing in human subjects and whether distinctions can be made as to whether the suppression is sensory and/or motor, as has been done for the neocortex (Ojemann 1991; Penfield and Roberts 1966). Finally, by comparing the inhibitory effects of electrically stimulating the neocortex and cerebellar cortex, we will be able to assess the contributions of declarative information (presumed to be stored in the neocortex) and sensory-motor information (presumed to be stored in the cerebellum) to language function (Chomsky 1965).
Figure 1: Homuncular projection scheme between the cerebellar nuclei and M1 via the ventrolateral thalamus of the monkey. From figure 1 of Thach et al. (1992), as adapted from Asanuma, Thach et aI. (1983).
Figure 2: The cerebellar cortex can be divided into three lobes based on humans: an anterior lobe which mediates skeletomotor responses, a posterior lobe which subserves egocentric/spatial body movements, and a mediolateral lobe which is important in eye-hand coordination made in response to allocentric objects (e.g., visual, auditory, and so on) important in language and mathematical skills. For further details see figure 8 and accompanying text of Tehovnik et al. (2021).
Figure 3: Functional areas of the human cerebellum on a flattened representation. Notice the large region dedicated to language in the right cerebellum. Note that projections between a neocortical hemisphere and a cerebellar hemisphere are typically crossed, but at the vermis (i.e., at midline) there is less of a contralateral bias. Modified from figure 4 of Van Overwalle et al. (2023).