THE CORTICAL ORGANIZATION OF SYNTAX
I am currently developing a neurobiological framework for syntax and semantics in the brain, centered on the posterior superior temporal sulcus (pSTS as a core region for hierarchical syntactic representations (Matchin & Hickok, in review). The PTL is well-positioned to connect sensory brain systems to conceptual brain systems. Our model posits a second fundamental organizing principle for semantics and syntax that places the pSTS in an appropriate cortical location. This principle takes inspiration from vision research regarding two fundamental streams of processing: the dorsal stream (how to interact with an object) and the ventral stream (what an object is) (Goodale & Milner, 1992). The pSTS connects these two systems for processing meaning to each other, allowing for people to create complex new thoughts.
Our model also contributes to our understand of aphasia, and makes the prediction that syntactic deficits in fluent aphasia, paragrammatism, will be localizable to the pSTS. For more information, you can watch a talk I gave at the Center for the Study of Aphasia Recovery (C-STAR).
Contrasts of structure and content in MEG reveal distinct activation profiles for anterior and posterior temporal lobe regions
Neuroimaging studies of sentence processing reliably activate a set of brain regions in the frontal and temporal lobes (Mazoyer et al., 1993; Pallier et al., 2011). However, the specific role each region plays in sentence processing, particularly with respect to syntax, remains unclear. Using parallel fMRI and MEG experiments, we have shown that posterior temporal areas are involved in predictive abstract syntactic processing, while anterior temporal areas are involved in conceptual-semantic processing (Matchin et al., 2017, Cortex; Matchin et al., 2018, Human Brain Mapping). This distinction fits well with aphasia research, which has shown semantic deficits due to anterior temporal lobe damage (Hodges et al., 1992) and syntactic deficits due to posterior temporal lobe damage (Wilson & Saygin, 2004). It also connects well with the psycholinguistic literature on sentence processing that has converged on the notion of structural predictions (Levy & Keller, 2013; Lewis & Vasishth, 2005).
The function of Broca's area in sentence processing
Given that fMRI experiments of syntax commonly find activity in Broca's area (Matchin et al., 2017, Cortex; Zaccarella et al., 2015), a prominent claim is that particular parts of Broca's area (the left inferior frontal gyrus) are involved in structure-building mechanisms (Friederici, 2011). One such claim is that the pars triangularis of Broca's area processes a particular syntactic operation called movement, supported by data suggesting selectivity in Broca's area for distance effects for sentences with movement relative to anaphora constructions (Santi & Grodzinsky, 2007). We have shown that this selectivity disappears when these constructions are matched with respect to their predictability by using a construction called backward anaphora (Matchin et al., 2014, Brain and Language). Rather, it seems that Broca's area plays a role in sentence processing through working memory resources that are retuned for syntactic information (Matchin, 2017, Psychonomic Bulletin and Review).