Hierarchical syntactic processing is beyond mere associating: Functional magnetic resonance imaging evidence from a novel artificial grammar

Abstract

Grammar is central to any natural language. In the past decades, the artificial grammar of the AnBn type in which a pair of associated elements can be nested in the other pair was considered as a desirable model to mimic human language syntax without semantic interference. However, such a grammar relies on mere associating mechanisms, thus insufficient to reflect the hierarchical nature of human syntax. Here, we test how the brain imposes syntactic hierarchies according to the category relations on linearized sequences by designing a novel artificial “Hierarchical syntactic structure‐building Grammar” (HG), and compare this to the AnBn grammar as a “Nested associating Grammar” (NG) based on multilevel associations. Thirty‐six healthy German native speakers were randomly assigned to one of the two grammars. Both groups performed a grammaticality judgment task on auditorily presented word sequences generated by the corresponding grammar in the scanner after a successful explicit behavioral learning session. Compared to the NG group, we found that the HG group showed a (a) significantly higher involvement of Brodmann area (BA) 44 in Broca's area and the posterior superior temporal gyrus (pSTG); and (b) qualitatively distinct connectivity between the two regions. Thus, the present study demonstrates that the build‐up process of syntactic hierarchies on the basis of category relations critically relies on a distinctive left‐hemispheric syntactic network involving BA 44 and pSTG. This indicates that our novel artificial grammar can constitute a suitable experimental tool to investigate syntax‐specific processes in the human brain.

Neural classification maps for distinct word combinations in Broca’s area

“…Given the rich conceptual content of the adjective + noun classification, we believe that these findings are in agreement with both correlational and causal analyses of semantic load in this area (Graessner et al, 2021;Hamilton et al, 2009;Hartwigsen et al, 2016). Conversely, the observation that posterior Broca's area contains a significantly larger number of voxels that are informative for the recognition of determiner + noun combinations with reduced conceptual content, appears to mirror syntactic processing effects found in the area for longer structures, regardless of meaning (Bornkessel et al, 2005;Chen et al, 2021;Friederici and Kotz, 2003;Goucha and Friederici, 2015;Zaccarella et al, 2017a). The identification of a set of voxels in the ventral region of BA45 that are responsive to both adjective + noun classification and the semantic task classification fits with the proposal of the anterior Broca's area as a generalised executive semantic processor (Chiou et al, 2018).…”

“…Conversely, when syntactic pseudo-phrases (this flirk) where contrasted with word-list sequences lacking syntactic information (apple flirk), BA44 was conversely found (Zaccarella et al, 2017a). As a matter of fact, an artificial grammar experiment reported that BA44, but not BA45, was maximally involved with the pSTS/STG when the grammaticality of artificial sequences had to be evaluated on the basis of the syntactic relations to be established between the elements of the sequence (Chen et al, 2021). By this account, the major advantage of using simple two-word combination lies in the possibility to focus on the fundamental question of how lexical and grammatical information contribute to the most basic combinatory operations at the very basic two-word level in Broca's area, and to evaluate whether the selective involvement of BA44 and BA45 reflects the processing of distinct lexical categories forming the phrase.…”

Which Cognitive Factors Predict L2 Grammar Learning: Cognitive Control, Statistical Learning, Working Memory, or Attention?

“…This suggests that different cognitive factors might account for the individual variability in different phases of grammar learning, testifying to the plausibility of examining the issue with different factors and different phases of grammar learning into consideration. The findings also lend support to the neurocognitive account of grammar learning ( Skeide et al, 2016 ; Chen et al, 2021a , b ). According to the account, in the initial phase of grammar learning, learners principally rely on the statistical properties of the linguistic sequences to extract the grammatical patterns, resulting in the involvement of SL in the initial phase.…”

“…As proficiency develops, learners would employ merge operation, namely, merging two constituents to form syntactic representation, to further process the grammatical patterns, while in the intermediate and more proficient phase, grammar learning would be a top–down process, with more involvement in the bilateral frontal cortex and dorsal language network. In other words, they might resort to a mechanism for complex syntactic pattern processing and other higher-order cognitive functions such as cognitive control ( Chen et al, 2021a , b ). This suggests that different cognitive factors might play different roles in different phases of L2 grammar learning, which might result in the different predictive power of these cognitive factors during the learning process.…”

Syntax through the looking glass: A review on two-word linguistic processing across behavioral, neuroimaging and neurostimulation studies

“…This is further supported by recent studies showing that the left IFG (Trettenbrein et al, 2020) and the posterior temporal lobe (Matchin et al, 2022) are central regions in processing signed languages, in line with the abstract nature of their linguistic operations. Similarly, recent evidence from artificial grammar paradigms supports the involvement of the left IFG and PTL in abstract categorical processes (Chen et al, 2021;Chen et al, 2019), putting forward the working hypothesis that dissociations between left IFG and pTL in syntactic composition can be traced-the left IFG being specialized in building up hierarchies on the basis of categorical features, the PTL integrating hierarchies with other sources of linguistic information, including meaning. Under this account, the information exchange between the left IFG and PTL would allow to reconstruct the hierarchical dependencies characterizing human language, providing an analysis that will also interface with the semantic system.…”