The Role of Object-based Attention in Semantic Working Memory

Ting Guo ^1* Leyue Ye ¹ Jinfeng HUANG ² Zhihan Xu ^1*

• ¹ Ningbo University of Technology, Ningbo, Zhejiang Province, China
• ² Shenzhen University, Shenzhen, Guangdong Province, China

Recent studies have shown that in visual tasks, bindings are temporarily stored and processed within working memory (WM), while semantically-related working memory processing often interacts with more permanent long-term memory (LTM). Specifically, we divide semantic information into two categories: grammatically connected sentences that benefit from semantic and syntactic integration through LTM, and isolated word lists that lack these meaningful structural connections. Through grammar chunking in working memory, sentences typically show enhanced memory performance compared to individual words, known as the sentence superiority effect. The current study aimed to investigate how object-based attention resources contribute to the processing of semantic working memory in second language (L2) learners, and further explore the impact of object attention task load on the sentence superiority. We used the Duncan task as an interference task to examine whether object-based attention load would affect the sentence superiority effect in L2 leaners. Results showed that connected sentences were more resistant to attention interference compared to word lists, suggesting different mechanisms of attention resource deployment in semantic processing. These findings reveal important insights into how linguistic context influences the relationship between attention resources and working memory.