Simin Meykadeh; Werner Sommer; Seyed Amir Hossien Batouli
Abstract
The cerebellum is linked to all the key regions of the language control network. Currently, the cerebellum is recognized to be involved in the networks that handle grammatical aspects. Clinical and neuroimaging studies have confirmed cerebellar contributions to grammar processing. The present study intended ...
Read More
The cerebellum is linked to all the key regions of the language control network. Currently, the cerebellum is recognized to be involved in the networks that handle grammatical aspects. Clinical and neuroimaging studies have confirmed cerebellar contributions to grammar processing. The present study intended to investigate the activity of the cerebellum in alternating L1-L2 processing in balanced bilinguals. We selected 35 Turkish-Persian bilinguals (21 women) who had learned their second language at the age of 7. Based on the Bilingual Dominance Scale, there was no significant difference between the high proficiency levels of the participants in L1 (Turkish) and L2 (Persian). Participants carried out an auditory grammaticality judgment task in an alternative language-switching paradigm while fMRI images were acquired using a standard protocol. Combining a whole-brain and regions-of-interest (ROIs) approach, we examined event-related fMRI during syntactic processing. Following the identification of the activity of the bilateral cerebellum at the whole-brain level according to the Harvard-Oxford Atlas in FSL, percent signal change was extracted per participant as an intensity measure in the cerebellar region and statistically analyzed in SPSS. The results indicate a right hemispheric superiority in bilingual language processing, confirming that the right cerebellum is more involved in language control. Furthermore, bilinguals have shown stronger activation for L1 as compared to L2 in the cerebellum, substantiating the reversed language dominance effects.IntroductionThe cerebellum, located at the back of the brain beneath the occipital lobes, contains approximately 80% of all brain neurons, but constitutes only approximately 10% of brain volume. Despite the fact that this brain region was previously known as a nervous system for movement control, many studies have confirmed that the cerebellum plays an important role in behavioral, sensory, and cognitive functions, including non-motor language functions. Hemispheric cerebellum asymmetry of functional activation during language processing is also reported. Due to the role of the cerebellum in language processing, we examined its contribution to morphosyntactic processing.The main research questions are as follows:RQ1. To what extent is the Cerebellum involved in the processing of grammatical agreement by balanced bilinguals?RQ2. Does the left and right cerebellum act differently for the simultaneous processing of the L1 and L2?To answer the research questions guiding this study, a bilingual task with an alternating language-switching paradigm was developed. In this task, brain imaging was performed using event-related fMRI while the participants listened to a total of 128 sentences in two Turkish and Persian languages.Literature ReviewUsing normal participants, Kovelman et al. (2008) examined 11 Spanish-English bilinguals and 10 English monolinguals during a syntactic judgment task. Bilinguals received their bilingual exposure before age 5. Monolinguals were presented with 40 English sentences and bilinguals were presented with 40 English and 40 Spanish sentences. Based on their findings, although the activity of the cerebellum was detected in both bilinguals and monolinguals, bilinguals showed a stronger effect in the cerebellum as compared to the monolinguals. Given that no neuroimaging study to date has examined the pattern of brain activity within the same individuals in Planum Temporale, we aimed to contribute to the literature about morphosyntactic analysis of L1 and L2 in two Subject-Object-Verb (SOV) languages. MethodologyIn this section, the applied methods and procedures including the choice of participants and stimuli followed by a description of the fMRI data acquisition and preprocessing are presented.3.1. ParticipantsTo allow for reliable ROI-based analysis, 36 right-handed and balanced Turkish-Persian bilingual students were recruited to participate in this study. All participants were native speakers of Turkish and learned Persian at school from the age of seven. Participants' language proficiency levels were assessed by the Bilingual Dominance Scale (BDS) and no significant difference was observed between Turkish and Persian (i.e., between L1 and L2) in language dominance.3.2. Materials and ProcedureDuring a bilingual grammaticality judgment task, participants heard 128 test sentences (64 in L1 and 64 in L2, with 50% violation per language) and made their judgment by pressing a button. Stimuli were presented using the Psychtoolbox in MATLAB via headphones. Stimuli were randomized for each condition, but alternated in a fixed sequence for language.3.3. ImagingMRI data were collected in NBML, Tehran, Iran, using a Siemens Prisma 3T scanner with a 20-channel head coil. For each participant, a high-resolution T1-weighted anatomical scan was acquired (TR = 1800msec, TE = 3053 msec, flip angle: 7°, 192 axial slices, slice thickness = 1 mm, field of view (FOV) = 256 mm2, 256 × 256 acquisition matrix, voxel size: 1×1×1 mm). After the anatomical scan, participants underwent a 21.5-min fMRI scan that used a whole brain echo planar imaging (EPI) sequence (TE: 30 ms, TR: 3000 ms, flip angle: 90°, slice thickness: 3 mm, voxel size: 3×3×3 mm, matrix size: 64×64, FOV: 192 mm2, 430 volumes and 45 axial slices per volume).3.4. Data preprocessingProcessing of the fMRI data was carried out using FEAT in FSL. Preprocessing steps included motion correction, slice-timing correction, non-brain removal using BET, spatial smoothing (6 mm FWHM), normalization, temporal filtering (with sigma = 50.0 s), and exploratory ICA-based data analysis. Statistical analyses of fMRI data were conducted using general linear modeling (GLM), as implemented in FSL. Z statistic images were thresholded using clusters determined by Z > 3.1 and a (corrected) cluster significance threshold of P < 0.05. After detecting the Cerebellum activation in the whole-brain analysis, percent signal changes were extracted as an intensity measure in this brain region. All statistical analyses were conducted in IBM SPSS Statistics 26.ResultsIn this section, results are presented starting with the whole-brain findings followed by the detection of cerebellum activity.4.1. Whole-brain resultsWidespread significant BOLD activation was found during the presentation of the sentences of L1 and L2 in the Cerebellum relative to the baseline (Figure 1). Visual inspection of panels 1 and 2 indicates more activity in L1 as compared to L2. Therefore, an ROI-based analysis was performed for both languages in the bilateral cerebellum to determine the activity pattern of the stimuli in this brain area.Figure 1. Whole-brain clusters (dark blue) of BOLD activation for (A) L1 and (B) L2 sentences in the Cerebellum, projected onto surface templates using MRIcroGL software in two experimental conditions including (from left to right) Ungrammatical and Grammatical conditions relative to the baseline. 4.2. Results of the region of CerebellumThe location of the Cerebellum is rendered in Figure 2. A significant main effect of Grammaticality was found, indicating a stronger activation for ungrammatical as compared to grammatical conditions (4.510 vs. 3.712 PSC). There was also a significant main effect of Language, indicating that the L1 conditions generated stronger effects than the L2 conditions (4.312 vs. 3.911 PSC). The main effect of the Hemisphere was also significant with a higher PSC for the right (4.465) than that for the left hemisphere (3.756). A separate t-test of the grammaticality effects per language in Cerebellum showed that it was significant for L1 but not for L2. In L1, post-hoc analysis indicated a right hemispheric superiority in our participants. Figure 2.(A). Location of Cerebellum (in yellow). (B) Box plots of percent signal change (PSC) values for L1 in Cerebellum per hemisphere and condition. *p < 0.05 ConclusionThe present ROI-based analysis has two important findings. First, the grammaticality effect was detected in the right hemisphere, which confirms previous studies in normal (Marien et al. 2014) and patients (Silveri et al., 1994; Marien et al., 1996; Gasparini et al. ., 1999). The most important argument in support of the role of the right cerebellum in the present study is the simultaneous activity of the Pars opercularis, posterior Superior Temporal Gyrus (pSTG) and the right Cerebellum (see also Meykadeh et al., 2021a) which is consistent with the findings of Berken et al. 2016. Berken and his colleague examined the French-English bilinguals during resting-state fMRI and observed functional connections between the left inferior frontal gyrus and the bilateral Cerebellum. Second, the grammaticality effect was significantly stronger in L1 than in L2 in the Cerebellum region. In line with the activity threshold hypothesis (Paradis, 1993; 2001), our participants regarded Turkish (L1) as the base language and Persian (L2) as the guest language during language exchange.AcknowledgmentsThis work was supported by the Cognitive Sciences and Technologies Council of Iran (Grant agreement, no. 7401); a Doctoral Dissertation Grant from the Department of Linguistics, Tarbiat Modares University, a Scholarship Fund (Ph.D. Visiting Scholar Program) from the Iranian Ministry of Science, Research and Technology.
