The present study was conducted using functional magnetic resonance imaging to detected cerebellar areas that were specific to word reading in adult bilingual readers. Comparison of cerebellar activation across different visual processing conditions, we observed that activation in the right lobule VI and VIII was significantly enhanced in the word-reading condition in different languages compared to the other two control conditions (i.e., pseudoword reading condition and picture viewing condition). Additionally, these two regions did not show preferences for the other high-level cognitive task (i.e., the emotional processing task). Furthermore, activation matrix in the right lobule VI was correlated with the phonological matrix and the neural activity was highly consistent across subjects. These findings contribute to a comprehensive study of the role of the cerebellum in reading, deepen our understanding of cognitive models of reading, and provide insights for improving reading ability and intervening in dyslexia.
Specificity of the right cerebellar lobules VI and VIII in reading
To investigate whether there are cerebellar areas specific to reading, three task conditions were designed, including the implicit word reading, pseudoword reading, and picture-viewing conditions. Two cerebellar areas, the right lobule VI and VIII, were selected as the regions of interest. These two regions were reported in our previous studies Li et al. (2022). Moreover, they have identified to be important for reading via meta-analytic analysis (Turker et al., 2023). The results showed that both ROIs activated significantly greater in the implicit reading condition than the other two conditions. Similar results were observed when we used explicit reading condition. To examine the stability of reading-selective areas, we compared activation between different conditions in different languages (Chinese and English). The results revealed that both ROIs demonstrated greater activation in the reading tasks (i.e., the implicit and explicit word reading condition) than the pseudoword task across languages. Finally, we observed that two regions were not significantly activated in other high-level cognitive processing tasks (emotional processing), further consolidating of their specificity to reading tasks.
This finding aligns with the cerebellar reading-selective areas previously identified by our group in child subjects (Li, Marks, et al., 2022), demonstrating stability across age. In a study comparing cerebellar activation in adults and children during reading (Ashburn et al., 2023). Both adults and children activated the cerebellum bilaterally during word processing in an implicit reading task. No differences were found when comparing adults and children. Consistent results across different age groups indicates that the specificity of the cerebellum to reading could be independent of development stages. However, studies on the cerebellum in age-related language processing have found that reductions in cerebellar volume were associated with language learning deficits in older adults (LeBel & D’Mello, 2023). Moreover, the most pronounced decrease in gray matter during the aging process involves the right VI, Crus I, and II cerebellar regions, which coincide with functionally defined language areas in the cerebellum (Stalter et al., 2023). Thus, the relationship between the cerebellum and developmental stages required further investigation.
Notably, in our previous studies, we only compared word reading condition with picture viewing condition. However, these two conditions differ in several key aspects. Word stimuli contain visual orthographic information, which is absent in pictures. Additionally, words carry high-level phonological and semantic information, which are also lacking in pictures. The process of lexical reading primarily involves morphophonemic or morphosyntactic matching. Differences between text and pictures at the visual orthography level may reflect disparities in general language processing rather than specifically in reading processes. To address factors at the general linguistic level, we introduced text-like stimuli known as pseudoword stimuli. These pseudo-words adhere to orthographic rules but lack phonological or semantic content. If a particular region within the cerebellum serves as reading-selective areas for lexical reading, we would expect activation in this region to be significantly stronger when processing pseudo-word stimuli compared to both text stimuli and picture stimuli. As we expected, the right lobule VI and lobule VIII both showed significant activation in the word reading conditions compared to the pseudoword condition, indicating that these two regions were selective to reading rather than generally engaged in all language tasks.
The preference of right lobule VI for phonological processing
Vocabulary reading relies on orthographical, phonological, and semantic processing. Experiment I revealed that the right lobule VI and VIII were specifically involved in the process of vocabulary reading, but it is still unclear what the role of these two areas is in vocabulary reading. Experiment II first conducted pattern similarity analyses in the silent word reading task. The correlation between the similarity across word stimuli and their associated neural activity was calculated. The result showed that the neural similarity matrix in the right cerebellar area VI was correlated with the phonological similarity matrix. No significant correlation was found in the right lobule VIII. We further set up three classical reading tasks, the orthographic task, phonological task, and the semantic task. There results showed that activation in the right lobule VI showed significant inter-subject correlation only in the phonological task. The result was consistent with those of the pattern similarity analysis. Together, these results suggest that the right lobule VI may be more involved in phonological processing during lexical reading, whereas the role of the right lobule VIII in morphosyntactic processing remains unclear.
Functional attribution of the right lobule VI has not been consistent. For example, studies have showing that the right cerebellar lobule VI was implicated in motor tasks, working memory, with increased activity observed in this region when the working memory load is high (Li, Marks, et al., 2022). By summarizing previous studies, a recent meta-analytic study reported consistent activation in right lobule VI in the phonological-related tasks (Turker et al., 2023). Interestingly, semantic processing also activated the right cerebellar lobule VI, consistent with neurostimulation and neuroimaging studies emphasizing its involvement in semantic processing and prediction (D'Mello et al., 2017). In the current study, by combining representation similarity analysis and inter-subject consistency analysis, we found that the right lobule VI may be engaged in phonological processing, contributing to the decoding of the cerebellar contribution to reading.
Language-Specific Effects in Cerebellar Activation
In the current study we examined language effects by comparing activation differences between these two regions in reading tasks and pseudoword processing tasks across languages. We observed a main effect of language both in the right lobule VI and VIII, with greater activation in the English reading tasks compared to the Chinese reading tasks, highlighting its specificity for language processing. This finding offers valuable insights into the neural mechanisms underlying language processing in the brain.
Language difference observed in these two regions may be attributed to the language proficiency. In this study, all participants were native Chinese speakers and non-native English speakers. Compared to Chinese, participants may recruit more neural resources, resulting in greater cerebellar activation. Additionally, English, as a phonetic writing system, emphasizes morphophonetic matching and relies on phonological decoding during processing. In contrast, Chinese, as an ideographic writing system, emphasizes orthography -semantics matching and relies on the holistic processing of character shapes (Tso et al., 2011). As we mentioned before, the right lobule VI could be associated with phonological processing, which may contribute to a greater activation in English word reading than Chinese word reading.
Notably, in Li et al. (2022), we investigated the neural response in the right cerebellar lobule VI to reading tasks involving Chinese characters and English words. However, the results revealed no significant differences in terms of activation between the two languages. These results suggest that activation in the reading-selective areas (i.e., the right lobules VI and VIIIA) may not be sensitive to variations in script type. This discrepancy might be associated with age. Comparing cerebellar activation in different languages across developmental states could be an appropriate extension of the current study.
Implications and Future Directions
In our study, we identified domain-specific activation of right cerebellar areas VI and VIII during a reading task, contrasting with the lack of domain-specificity observed in an emotion recognition task. This finding contradicts the cerebellar coherence transition theory, which posits that the cerebellum engages in higher cognitive processing primarily for motor functions and lacks domain specificity despite activation across various cognitive tasks (Ramnani, 2006; Schmahmann et al., 2019). Our results suggest that cerebellar involvement in reading tasks extends beyond motor functions. Additionally, our findings challenge the cerebellar deficit hypothesis proposed by Nicolson et al. (2001), which links cerebellar abnormalities to dyslexia and emphasizes the cerebellum's role in reading through motor and automation processes. Instead, our study reveals the preferential involvement of right cerebellar area VI in phonological processing, indicating a role in phonological decoding during reading. This finding does not align with the cerebellar deficit hypothesis. Conversely, recent studies have demonstrated the cerebellum's influence on reading through advanced language processing, such as phonology or semantics (D'Mello et al., 2020; Gatti et al., 2020; Li, Marks, et al., 2022; Li, Wu, et al., 2022; Meng et al., 2015), consistent with our results. Therefore, our study adds support to the notion that the cerebellum contributes to reading through various linguistic processes beyond motor functions.
Furthermore, prior studies have primarily examined differences in activation patterns between words and pictures, which has limited our ability to discern between general language processing and reading-specific mechanisms. To address this limitation, our study introduces pseudowords as an additional variable. By analyzing different task conditions, we aimed to determine detect whether the observed reading-selective areas in the cerebellum is specific to reading or part of a broader language processing network. Additionally, our study employs a diverse range of reading tasks and language conditions to investigate whether the reading-selective areas within the cerebellum exhibits stable activation patterns across task and language variations.
This approach enhances our understanding of the consistency and specificity of cerebellar involvement in reading processes. What's more, this study delved into the morphophonemic functions of the reading-selective areas within the cerebellum. We investigated its specific roles by integrating orthographical, phonological, and semantic processing in reading information. Employing a diverse array of reading tasks and neuroimaging techniques, we further dissected and compared activity patterns and temporal sequences of morphophonemic processing. This systematic approach enhances our understanding of the interplay between the cerebellum and reading. The findings of this research deepen our comprehension of the cerebellum's involvement in higher cognitive processing and provide a more precise neural mechanism for future reading investigations and clinical interventions targeting developmental dyslexia.
The present study has several limitations. Firstly, it employed lexical stimuli without considering contextual effects, potentially limiting result generalizability. Future research could employ more intricate experimental paradigms to comprehensively elucidate the neural mechanisms underlying cerebellar reading processes. Secondly, we only recruited adult readers. Continuing the investigation with child readers would be a suitable expansion of the present study. In conclusion, future investigations could delve deeper into the specific functions of the right lobule VIII in morpho-syntactic processing by designing more intricate tasks or incorporating diverse text stimuli.