Abstract:

The left Angular Gyrus (LAG) is a critical brain region for semantic processing in the cognitive neuroscience of language. Increasing attention has been paid to the function of LAG in lexical-semantic processing since its function has not been consistently understood.

Anatomically, the LAG is located at the junction of the temporal, occipital, and parietal lobes, with extensive white matter fiber bundles, determining that it may integrate information across regions. Researchers have proposed the LAG as the “convergence zone” in semantic representation and processing, as it engaged in high-level semantic representation, conjunctive representation of modalities and feature information, semantic relationship representation, and semantic integration processing. It is also involved in various semantic tasks and stimuli types, suggesting a high-level semantic representation function.

Evidence suggests that the LAG may be responsible for connecting and integrating conceptual multimodal features into a complete semantic representation. On the one hand, patients with LAG impairment exhibit multimodal semantic defects. On the other hand, LAG was observed on the semantic contrast for different word categories when manipulating the modality-specific features of words, indicating its function for cross modal conjunctive representation. At larger scales of phrase-level processing, the LAG integrates individual concepts into coherent semantic combinations. Its function of semantic integration also received support from its sensitive activation for semantic relationships. The LAG is involved in the category system organized based on semantic feature similarity and the theme system organized based on the co-occurrence relationship of events or scenes, which reflect the integrated processing of overlapping semantic feature information.

However, there are still debates about the function of LAG in semantic representation and processing. Firstly, the Embodied Abstraction theory of semantic processing proposes LAG as the hub region for semantic processing, while the Hub and Spokes model proposes that the hub region appears to be in the anterior temporal lobe. The mixed theory model recently seeks reconciliation, proposing both the LAG and the anterior temporal lobe as high-level hubs at different semantic processing stages. The LAG is a multimodal brain area that integrates low-level specific modal information while retaining related modality information. In contrast, the anterior temporal lobe is a higher-level amodal brain region that integrates semantic modal information into higher-level abstract representations without retaining modality-specific information. Secondly, inconsistent findings exist regarding LAG's role in semantic executive control processing, necessitating further exploration. Finally, regarding the function of LAG in the default mode network (DMN), opinions differ on whether LAG serves as a hub for semantic processing or is influenced solely by task difficulty. One view is that the DMN's LAG is a semantic processing hub region. Another view denies the semantic processing role of the LAG in the DMN and believes that task difficulty may cause the semantic effect of the LAG. The third multifunctional view is that the LAG may support semantic processing and general task difficulty.

The LAG exhibits broad activation in lexical-semantic processing, and its function is extensively debated. One possible reason is that there are structurally and functionally distinct sub-regions within the LAG, and another possibility is that the LAG serves as a cross-system connector, supporting multiple aspects of semantic processing. Therefore, future studies should comprehensively consider LAG’s anatomical basis and connections with a wide range of brain regions and conduct in-depth and detailed discussions on the function of the LAG sub-region.

Key words: left Angular Gyrus, semantic representation, semantic processing, “convergence zone”