We conducted a retrospective review of fMRI studies of complex syntax, in order to study the stability of the neural bases of mechanisms engaged in syntactic processing. Our review set out rigorous selection criteria of studies which we discuss, including transparency and minimality of the contrasts between stimuli, and the presence of whole brain analyses corrected for multiple comparisons. Seventeen studies with 316 participants survived our sieve. We mapped the 65 resulting maxima onto JuBrain, a state-of-the-art cytoarchitectonic brain atlas (Amunts et al., 2020), and a sharp picture emerged: syntactic displacement operations (a k a MOVE) produce highly consistent results, activating left Broca's region across-the-board and unambiguously; to a somewhat lesser extent, maxima clustered in left posterior brain regions, including the STS/STG. The few studies of syntactic tree-building operations (a k a MERGE) produce a murkier picture regarding the involvement of the left IFG. We conclude that the extant data decisively point to the JuBrain-defined Broca's region as the main locus of complex receptive syntax in healthy people; the STS/STG also are involved, but to a lesser extent.

我们对复杂句法的 fMRI 研究进行了回顾性审查，以研究参与句法处理的机制的神经基础的稳定性。我们的审查列出了我们讨论的研究的严格选择标准，包括刺激之间对比的透明度和最小化，以及针对多重比较校正的全脑分析的存在。有 316 名参与者的 17 项研究在我们的筛选中幸存下来。我们将得到的 65 个最大值映射到JuBrain 上，这是一个最先进的细胞结构脑图谱（Amunts 等人，2020 年），并且出现了一个清晰的画面：句法置换操作（又名MOVE) 产生高度一致的结果，全面且明确地激活左布洛卡区域；在较小程度上，最大值聚集在左后脑区域，包括 STS/STG。对句法树构建操作（又名MERGE）的少数研究产生了关于左 IFG 参与的更模糊的画面。我们得出的结论是，现有数据决定性地指出 JuBrain 定义的布洛卡区是健康人复杂接受句法的主要位点；STS/STG 也参与其中，但程度较轻。