On the origins of grammar: from structural complexity in auditory sequences to syntactic structures

在语言处理的理论语言学、神经语言学和认知心理学领域，一个尚未解决的问题是，人类处理句法结构的独特能力是如何出现的，从而支撑了语言能力（LF）的特殊性。此外，最近对句法理论和神经认知方面的研究指出，在非语言序列和人类句法处理中实现的结构规则可能具有功能对等。当处理序列和语法时，这种共同进化可能存在于并且仍然依赖于额纹状体脑系统。此外，这个大脑系统在其标记时间属性的能力中可能负责听觉语言处理中预测时间和顺序线索的同步。然而，还有一个未解之谜，那就是如何从额纹状体系统的时间和排序能力中理解句法能力的出现，包括递归计算和分层语言结构的处理。目前的项目旨在实现与这些开放问题相关的重大进展。这将通过建立一个跨学科的研究团队来实现，为跨语言和跨物种的实验工作和建模方法提供了独特的机会。特别是，我们将结合健康参与者的行为、电生理（ERP）和功能性磁共振成像（fMRI）实验，患有额纹状体紊乱（未治疗（早期）和治疗（进展中）帕金森病）的患者，计算模型和早期额纹状体功能紊乱的创新动物模型，以确定人类语法的计算起源。

A still unresolved issue in the fields of theoretical linguistics, neurolinguistics, and cognitive psychology of language processing is how the unique ability to process syntactic structure emerged in humans, underpinning the specificity of Language Faculty (LF). Furthermore, recent research on theoretical and neurocognitive aspects of syntax is currently pointing towards possible functional equivalence of structural rules implemented in non-linguistic sequences and syntactic processing in humans. Such co-evolution may reside in and still rely on a fronto-striatal brain system when processing sequences and syntax. Furthermore, this brain system in its capacity to tag temporal attributes may be responsible for the synchronization of predictive temporal and sequential cues in auditory language processing. However, one remaining mystery is to understand both the emergence of syntactic capabilities, including recursive computation and the processing of hierarchical linguistic structures, from the temporal and sequencing capacities of a frontostriatal system. The current project aims at realizing significant advancement in relation to these open issues. This will be possible by setting up an interdisciplinary research team offering the unique opportunity to combine cross-linguistic and cross-species experimental work and modelling approaches. In particular, we will combine behavioural, electrophysiological (ERP), and functional magnetic resonance imaging (fMRI) experiments in healthy participants, patients with fronto-striatal disruptions (unmedicated (early) and medicated (in progression) Parkinson’s disease), computational modelling, and an innovative animal model of early fronto-striatal function disruption in order to identify the computational origins of grammar in humans.