已有研究考察了母语和二语阅读的普遍性和特异性的神经机制，但是到目前为止，母语和二语阅读脑网络的普遍性和特异性神经机制是否受到两种语言间书写系统的相似性的调节，以及在不同年龄发展阶段是如何变化的，母语阅读和二语阅读大脑网络随着发展是更加汇聚还是更加分离，还缺乏系统的研究对这些问题进行考察。因此本论文主要关心汉英双语者母语和二语阅读脑网络的普遍性和特异性的发展过程。我们结合元分析和实证研究的方法考察母语和二语阅读脑网络的普遍性和特异性，研究既选取了不同年龄阶段被试，探索母语和二语阅读脑网络普遍性和特异性的发展，又选取了阅读困难儿童对母语和二语阅读脑网络的普遍性、特异性及其发展进行进一步验证。

研究一考察了双语者阅读母语和二语的普遍性和特异性的神经机制，并探索两种语言书写系统的相似性是否会对母语和二语阅读脑网络的普遍性和特异性有调节作用。研究采用元分析的方法，对40篇以双语者为研究对象的文献进行了分析，所选取文献的实验范式为阅读相关的任务(语音，正字法或者语义任务)，被试既进行了母语阅读，又进行了二语阅读。研究发现，左侧额下回，左侧辅助运动区，左侧楔前叶，左侧梭状回和右侧枕中回为母语阅读和二语阅读普遍性区域，左侧额中回和右侧梭状回后部在母语阅读中表现出更强的激活，左侧顶下小叶，右侧楔前叶和左侧脑岛在二语阅读中表现出更强的激活。当将双语者进一步分为母语和二语书写系统相似以及母语和二语书写系统不相似两种类型的双语者时，发现母语和二语书写系统的相似性可以对阅读普遍性和特异性的脑网络进行调节。然而更为有趣的是，两种类型双语者在第二语言阅读加工中，左侧顶下小叶均会表现出更强的激活。也就是说不管母语与二语书写系统是否相似，顶下小叶在双语者的第二语言阅读加工中都发挥了重要作用。

研究二考察了母语和二语阅读脑网络的普遍性和特异性在不同发展阶段是如何变化的。为了更适于进行不同年龄阶段的双语者大脑神经活动的比较，研究二采用被动阅读任务，采集了30名儿童，27名青少年和30名成人阅读母语和二语的功能性磁共振成像数据，考察母语和二语阅读脑网络的整合随着发展的差异。首先我们基于激活或表征文字稳定性指标，考察母语和二语阅读网络在空间位置相似性上随着发展的变化。结果只在以表征文字稳定性为指标进行计算时，发现了母语和二语阅读脑网络在空间位置相似性上表现出儿童组要显著小于成人组，符合随着年龄发展两种语言的网络变得更加汇聚的假设。这暗示了从文字符号到意义的映射建立发生在阅读发展的早期阶段，两种不同书写系统对阅读网络的作用随着发展逐渐减弱，表现为母语和二语阅读脑网络的特异性减弱，普遍性增强。然而这种汇聚只体现在这种空间位置相似性上，当我们进一步考察表征母语和二语的模式相似性的发展时，并没有发现三组之间存在组间差异，也就是说两种语言的激活模式相似性没有显著发展效应。这可能与该研究采用被动阅读任务有关，因此在接下来的研究中，我们采用主动阅读任务对这些结果进行了进一步验证。

研究三采用主动阅读任务，采集了32名儿童和32名成人的神经成像数据，对研究二的结果进行验证。研究三从激活空间位置相似性，表征两种语言的模式相似性以及大尺度脑网络相似性几个方面展开，考察在发展过程中儿童和成人母语和二语阅读脑网络的普遍性和特异性变化趋势。研究发现，对于激活空间位置，母语和二语的阅读脑网络从儿童到成人逐渐汇聚，且在核心阅读脑区上，表征母语和二语的模式相似性随着发展也更加相似。最后对于大尺度脑网络这一层面，成人较儿童更为相似，体现在成人母语和二语阅读的模块化归属及大脑功能连接模式上。此外，在研究三中，我们观察到左侧顶下小叶既表现出了语言的主效应，又表现出了发展的主效应，在主动任务中左侧顶下小叶表现出特异于第二语言阅读的特征，这不同于在被动阅读任务中的发现。

研究四考察了阅读困难儿童母语和二语阅读是否存在普遍性或特异性缺陷，以及母语和二语阅读之间的普遍性和特异性与阅读困难之间的关系。同样为了更适于在不同类型儿童之间进行对比，研究四使用被动阅读任务，采集了20名正常发展儿童，13名母语阅读困难儿童，14名二语阅读困难儿童和13名双语阅读困难儿童的神经成像数据。结果发现二语阅读困难儿童在左侧缘上回延伸到颞中回的位置表征第二语言存在缺陷，且这种缺陷是特异于第二语言阅读困难的，但没有发现特异于母语阅读困难的神经缺陷。当进行基于感兴趣区分析时，结果发现，特异于母语阅读的感兴趣区(左侧额中回)在母语阅读困难儿童阅读母语的过程中不能稳定表征母语；特异于二语阅读的感兴趣区(右侧楔前叶和左侧顶下小叶)在二语阅读困难儿童阅读二语的过程中不能稳定表征二语。此外，研究发现阅读困难儿童和正常发展儿童在能够稳定表征母语和二语的脑区空间位置的相似性和表征模式的相似性上存在差异，相比于阅读困难儿童，正常儿童表征母语和二语的脑区在空间位置上更加汇聚，且正常儿童的阅读网络表征母语和二语的模式更加相似，体现在左侧梭状回和左侧额下回这两个阅读加工的普遍性脑区上。

最后，研究五对儿童进行第二语言训练，进一步考察二语经验是否会改变母语和二语脑网络之间的相似性。研究发现，二语经验不会单独调控母语和二语阅读脑网络的普遍性和特异性，但是使得第二语言特异性脑区顶下小叶发生可塑性改变。且在训练前后左侧顶下小叶神经模式的改变与行为成绩的提升存在正相关。最后，训练使得左侧顶下小叶和左侧楔叶之间的功能连接增加，这可能反映了从形到音的大脑活动的增强。

综上，本论文考察了母语和二语阅读脑网络普遍性和特异性是否受到书写系统的调节，随发展的变化趋势，与阅读困难的关系及其调控机制。研究发现，母语和二语阅读脑网络的普遍性和特异性受到两种语言之间的书写系统相似性的调节，然而左侧顶下小叶稳定地表现出在第二语言阅读中更高的激活。随着年龄的发展，母语和二语阅读脑网络的普遍性增强，体现在母语和二语阅读脑网络在空间位置的相似性，表征文字的模式相似性以及大脑网络的相似性这几个方面。母语和二语阅读脑网络的普遍性发展机制受阅读能力影响，阅读困难儿童母语和二语阅读脑网络的相似性小于正常发展儿童。此外，母语阅读困难儿童和二语阅读困难儿童存在特异性的神经缺陷。在儿童群体中，左侧顶下小叶已表现出对第二语言阅读的特异性。二语经验不会单独调控母语和二语阅读脑网络的普遍性，然而会使得左侧顶下小叶表现出可塑性改变。本论文以母语和二语阅读脑网络普遍性和特异性为视角，从不同层面揭示了母语和二语阅读脑网络普遍性和特异性的发展，以及顶下小叶在第二语言阅读中的重要性。考察双语儿童第二语言的习得可以让我们从发展的角度考察语言特异性的机制，以及在发展过程中两种语言之间的交互关系。这些为理解儿童母语和二语阅读网络的发展提供了新视角，为第二语言阅读的神经机制丰富了证据。

Previous research has examined the neural mechanisms underlying the universality and specificity of the first and second language reading, but whether the universality and specificity of brain networks for the first and second language reading are modulated by the similarity between the writing systems of the two languages, and how they change during different developmental stages, as well as whether the brain networks for the first and second language reading converge or diverge with development, lacks systematic investigation. Therefore, in this study, we focused on the universality and specificity of brain networks on Chinese-English bilinguals for the first and second language reading and their development. We used a meta-analysis and empirical research to examine the universality and specificity of brain networks for the first and second language reading, and selected participants of different ages to explore the development of the universality and specificity of brain networks for the first and second language reading. We also further validated the universality, specificity, and development of brain networks for the first and second language reading in children with reading impairment.

In the first study, we investigated the neural mechanisms underlying the universality and specificity of the first and second language reading in bilinguals, and explored whether the similarity between the writing systems of the two languages modulates the universality and specificity of brain networks for the first and second language reading. Therefore, we used a meta-analysis method and selected 40 studies on bilinguals. The selected studies used reading-related tasks (phonological, orthographic, or semantic tasks) in which participants perform both first and second language reading. We found that the left inferior frontal gyrus, left supplementary motor area, left precuneus, left fusiform gyrus, and right middle occipital gyrus are the universal brain regions for the first and second language reading, while the left middle frontal gyrus and the posterior part of the right fusiform gyrus showed stronger activation during the first language reading. The left inferior parietal lobule, right precuneus, and left insula showed stronger activation during the second language reading. When we further categorized bilinguals into two types according to the similarity of writing systems between their first and second languages, we found that the similarity between writing systems can modulate the universality and specificity of brain networks for reading. However, interestingly, the left inferior parietal lobule showed stronger activation during second language reading in both types of bilinguals, indicating that the left inferior parietal lobule plays a vital role in second language reading processing in bilinguals, regardless of whether the writing systems of the first and second language are similar.

In the second study, we aimed to investigate the universality and specificity of the neural networks underlying the first and second language reading, and how they change across different developmental stages. To enable comparisons of brain activity in bilingual individuals of different ages, a passive reading task was used to collect functional magnetic resonance imaging (fMRI) data from 30 children, 27 adolescents, and 30 adults as they read in their first and second languages. We examined how the convergence of neural networks for the first and second language reading changes with development, based on measures of activation or representation stability. We found that only when using representation stability as a measure did we observe that the spatial similarity of neural networks for the first and second language reading was significantly lower in the child group than in the adult group. This suggests that the mapping from written symbols to meaning occurs in the early stages of reading development and the effect of two different writing systems on the reading network gradually weakened with the development, as manifested by the decreased specificity and increased universality of the first and second language reading brain network. However, this convergence was only observed in spatial similarity, and when we further examined the development of pattern similarity in the representation of the first and second languages, no significant differences were found between these three groups. This suggests that the similarity of activation patterns for both languages does not show significant developmental effects. Therefore, this may be related to the use of a passive reading task in this study. Therefore, in the next study, we will further validate these results in an active reading task.

In the third study, we employed an active reading task and collected neuroimaging data from 32 children and 32 adults to validate the results of the second study. This study examined the universality and specificity of the first language and second language reading brain networks in children and adults, focusing on several aspects including activation spatial similarity, pattern similarity in representing the two languages, and large-scale brain network similarity. We found that the first and second language reading brain networks gradually converged from children to adults in terms of activation spatial location. In addition, the pattern similarity in representing the first and second language became more similar with development in the core reading brain regions. Finally, regarding the large-scale brain network level, adults were more similar than children in terms of modular assignment and functional connectivity patterns for both the first and second language reading. Furthermore, in the third study, we observed a main effect of language and a main effect of age in the left inferior parietal lobule, which showed the specificity for second language reading in the active task, unlike the findings in the passive task in the second study.

In the fourth study, we investigated whether there are universal or specific deficits in reading both of two languages in children with reading difficulties, and to explore the relationship between the universality and specificity of language networks and reading difficulties. To facilitate cross-group comparisons, passive reading tasks were administered and neural imaging data were collected from 20 typically developing children (TR), 13 children with reading difficulties in the first language (PC), 14 children with reading difficulties in the second language (PE), and 13 children with reading difficulties in both languages (PB). Results indicated that children with reading difficulties in the second language exhibit deficits in the left supramarginal gyrus extending to the middle temporal gyrus, which are specific to reading difficulties in the second language, however, there were no neural deficits specific to reading difficulties in the first language. When performing region of interest (ROI) analysis, it was found that the region specific to the first language reading (the left middle frontal gyrus) exhibited neural deficits in the PC group reading the first language, while the regions specific to the second language reading (the right precuneus and left inferior parietal lobule) exhibited neural deficits in the PE group reading the second language. Furthermore, differences were found between children with reading difficulties and typically developing children in terms of the spatial similarity and pattern similarity in representing the two languages. Specifically, typically developing children exhibited greater convergence of brain regions involved in both two languages, and their patterns of neural representation for both languages were more similar, particularly in the left fusiform gyrus and left inferior frontal gyrus, which are universal brain regions involved in reading processing.

In the final study, we recruited children who received second language training to investigate whether the training would alter the similarity between the brain networks for the first language and second language. The study found that second language experience did not regulate the universality and specificity of the first and second language reading brain networks alone. However, the second language training induced plastic changes in the left inferior parietal lobule, which showed specificity for the second language in our study. The changes in the neural patterns of the left inferior parietal lobule before and after training were positively correlated with improvements in behavioral performance. Furthermore, training led to increased functional connectivity between the left inferior parietal lobule and the left cuneus, which may reflect enhanced brain activity during the process of orthographical-to-phonological mapping.

In summary, this study investigated whether the universality and specificity of the first and second language reading brain networks are modulated by writing systems, their developmental trends, their relationship with reading difficulties, and the underlying regulatory mechanisms. It was found that the universality and specificity of the first and second language reading brain networks are modulated by the similarity of writing systems between the two languages. However, the left inferior parietal lobule consistently exhibits higher activation in second language reading. The universality of the first and second language reading brain networks strengthens with development, as evidenced by the spatial location similarity, the pattern similarity of character representation, and the similarity of brain networks in both the first and second language reading. The development of the universality of the first and second language reading brain networks is affected by reading ability, with children with reading difficulties showing less similarity in their first and second language reading brain networks than typically developing children. Moreover, there are specific neural deficits in children with reading difficulties in the first language and children with reading difficulties in the second language. The left inferior parietal lobule already showed specificity in second language reading in children. Second language experience did not independently regulate the universality of the first and second language reading brain networks, but it does induce plasticity changes in the left inferior parietal lobule. This study depicted the development of the universality and specificity of the reading brain network in the first and second language from different levels, and revealed the importance of the left inferior parietal lobule in the second language reading. Investigating the acquisition of second language in bilingual children allows us to examine the mechanisms of language specificity and the interaction between the two languages from a developmental perspective. It provides a new perspective for understanding the development of children's first and second language reading network, and enriches new evidence for the neural mechanism of second language reading.