Little brain, little minds: The big role of the cerebellum in social development

Abstract

Seminal work in the 1990's found alterations in the cerebellum of individuals with social disorders including autism spectrum disorder and schizophrenia. In neurotypical populations, distinct portions of the posterior cerebellum are consistently activated in fMRI studies of social cognition and it has been hypothesized that the cerebellum plays an essential role in social cognition, particularly in theory of mind. Here we review the lesion literature and find that the effect of cerebellar damage on social cognition is strongly linked to the age of insult, with dramatic impairments observed after prenatal insult, strong deficits observed after childhood damage, and mild and inconsistent deficits observed following damage to the adult cerebellum. To explain the developmental gradient, we propose that early in life, the forward model dominates cerebellar computations. The forward model learns and uses errors to help build schemas of our interpersonal worlds. Subsequently, we argue that once these schemas have been built up, the inverse model, which is the foundation of automatic processing, becomes dominant. We provide suggestions for how to test this, and also outline directions for future research.

Keywords: Autism; Cerebellum; Learning; Mentalizing; Posterior fossa; Schizophrenia; Social; Stroke; Subcortical.

Fig. 1

A. Schematic diagram, modified from Sathyanesan et al. (2019) depicting the timeline of major cellular events in the cerebellum. Note that beyond year 1, human cellular data are scarce. A few structural MRI studies indicate that the cerebellum increases in size through adolescence (Sussman et al., 2016), however, more data are needed to determine the source and meaning of this increase. B. Sagittal views of cerebellar pathways, from a post-mortem diffusion imaging study, imaged at 17–38 weeks (W) and in the adult (Takahashi et al., 2014). High fractional anisotropy (FA) values are indicated with arrows. Many pathways had low FA values until 38 weeks of gestational age. C. HARDI tractography showing development of the cerebellar peduncles. The superior cerebellar peduncle pathway (SCP) is shown in light blue. The middle cerebellar peduncle (MCP) is in green and yellow. The inferior cerebellar peduncle pathways (ICP), shown in pink (Re et al., 2017). Images in B and C used with permission of authors. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

A series of FLAIR scans depicting different posterior fossa tumors in pediatric case studies. A. Axial view in a case that yielded cerebellar mutism following resection (reproduced with permissions from “The Posterior Fossa Society”); B. Axial view in a 16 year old female, presenting with binocularly blurred vision and emotional changes such as “giddiness” (Shah, 2012); C. Sagittal view in a 5-year-old male presenting with cerebellar symptoms (Dähnert, 2011). D. Sagittal view of medulloblastoma in a 2-year old female presenting with ataxia, lethargy, and post-operative cerebellar mutism (Gadgil et al., 2016). Images used with permission of authors. Note that images B-C are reproduced for publication in compliance with Creative Commons licenses.

Fig. 3

A compilation of studies in which adults with cerebellar damage had theory of mind formally tested. Studies were excluded if they did not test a control group (e.g., Lupo et al., 2020; Moriarty et al., 2016; Parente et al., 2013), failed to provide numerical values for the control group (e.g., Garrard et al., 2008), or did not report raw accuracy scores for each group (e.g., Beuriat, Cohen-Zimerman et al., 2022a). Developmental studies were excluded because pediatric populations are rarely administered ToM tests. Studies are grouped by task category, then by article. The y-axis depicts the mean proportion correct on each task for cerebellar patients (light hue) and healthy controls (dark hue). This was calculated by dividing the total number of points possible for each task by the average score for each group. Note that for Roldan Gerschcovich et al. (2011), scores for the RMET were collapsed across trials for faces and eyes to get a composite score for each group. Moreover, only comprehension scores are reported for the ToM task administered in Sokolovsky et al. (2010), as this was the only metric derived from the task that reflected accuracy. Additionally, the emotion attribution scores from this paper are not reflected in this meta-analysis, as group-wise insights are more difficult to glean given the provided statistics. Finally, the averages for the CB group in the Tamaš et al., 2021 paper are collapsed across types of CB pathology.

Fig. 4

Cerebellar internal models and social development. A. The forward model is critical for the early development of social mental models which are used for learning about and predicting other people’s attention, how they move and gesture, their motives, internal thoughts and emotions, and acquiring social knowledge, as illustrated here. B) The inverse model is important for the fluid utilization of social mental models which by adulthood, are mature. This is important for automatic mentalizing, deploying social norms, language pragmatics, and the quick processing of emotions. C) We hypothesize that earlier in life, the forward model will be critical to build social models and map accurate actions to desired outcomes. Later in life, the inverse model will be more widely used so that individuals can seamlessly interact with others. By this time, an individual will have extensive experience with the social world, new types of social situations are infrequent, and thus the forward model is not as crucial as it once was.