Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Linking semantic structure and search dynamics in aging: a cross-network approach
Poster Session C, Thursday, October 1, 10:45 am - 12:45 pm, Wangari Maathai
Jinyi Hung1; 1MacKay medical university
The human semantic system supports language through both the organization of conceptual knowledge and the processes by which it is accessed. However, how semantic structure and search dynamics jointly relate to language performance remains unclear, particularly in aging. In the present study, two complementary semantic networks were constructed to investigate cross-network associations and their contributions to language performance. We hypothesized that semantic networks derived from relatedness judgments and verbal fluency capture convergent yet nonredundant aspects of semantic system, which differentially relate to naming performance. Participants included 18 young adults, 62 healthy older adults, and 23 individuals with cognitive decline. We first examined group differences in each network and then evaluated cross-network associations using theoretically defined constructs indexing network integration, connectivity, and switching, allowing correspondence between structural and search-based metrics. A semantic network derived from relatedness judgments was used to characterize the structural organization of conceptual knowledge, whereas fluency-based networks captured search dynamics based on sequential word transitions. While the relatedness network reflects similarity-based associations among concepts, the fluency network captures the temporal structure of retrieval as participants move between items. Both networks were constructed within the same semantic categories (animal, tool, fruit), enabling direct evaluation of structural and search-related dimensions. Across the two network types, parallel patterns of group differences in semantic structure and search dynamics showed that semantic structure was largely preserved in healthy aging but was characterized by reduced connectivity and increased fragmentation in cognitive decline. Lower transition entropy was also found in cognitive decline, reflecting less flexible search dynamics. Cross-network analyses revealed selective associations between structure and search, particularly for integration and connectivity, suggesting that semantic structure may constrain search dynamics in a selective, domain-sensitive manner. To further investigate how semantic structure and search dynamics relate to language performance, regression-based predictive modeling analyses were conducted. The results revealed that fluency-based measures explained additional variance in naming performance, whereas structural metrics derived from relatedness networks did not provide unique explanatory power. Within the fluency domain, measures indexing search efficiency and flexible transitions, particularly integration and switching-related indices, were the strongest predictors. This pattern was observed consistently across analyses. These findings suggest that semantic structure may contribute to language performance through its association with search dynamics, highlighting a functional dissociation between semantic organization and the processes supporting successful retrieval in aging. This work provides a cross-network framework for linking semantic network organization and search dynamics, offering a new perspective on language variability in aging.
Topic Areas: Meaning: Lexical Semantics,