Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Discourse-level thematic context modulates lexical retrieval: Behavioral and neural evidence from younger and older adults
Poster Session C, Saturday, September 13, 11:00 am - 12:30 pm, Field House
Katarina Antolovic1, Jessica Marutani2, Fatima Medrano1, Yan H. Yu3, Mira Goral1,4, Valerie L. Shafer1; 1The Graduate Center, City University of New York, 2Hunter College, City University of New York, 3St. John's University, 4Lehman College, City University of New York
Aging induces changes in language processing and production, including alterations in predictive processing and increased word retrieval failures (e.g., Barresi et al., 2000). Prior research has focused on retrieval failures arising from semantic interference induced by categorical word relations (e.g., dog and horse). However, retrieval in more ecologically valid settings, including discourse, may involve co-activation of thematic relations, which are context-dependent word associations (e.g., "butterfly" and "shovel" are exemplars of "garden"). Thematic context arguably requires both the formation of ad-hoc categories representing the theme and the maintenance of contextual information. These processes may be differentially affected by aging, as older adults show decrements in contextual processing and maintenance (e.g., Federmeier et al., 2010). The mechanisms underlying how thematic relations impact retrieval are poorly understood. Competing theories suggest that thematic context may either aid (e.g., Dark Side Model Revision, Oppenheim & Nozari, 2021) or interfere with (e.g., Swinging Lexical Network Hypothesis, Abdel Rahman & Melinger, 2009; 2019) retrieval. To determine how thematic context influences retrieval and whether the effect of context is age-invariant or not, we tested 24 younger (M = 28.5) and 23 older (M = 70.35) adults using a modified Blocked Cyclic Naming Paradigm (BCNP). Participants listened to a narrative establishing a thematic context (e.g., day at the beach) to simulate a discourse-like context. They then named thematically related or unrelated images in two block configurations: related items followed by unrelated items (related-first blocks), and vice-versa (unrelated-first blocks). After naming, participants recalled the narrative content to assess maintenance. We examined both behavioral responses during initial picture naming and across five repetitions, as well as neural responses using electroencephalography time-locked to image onset. We found that participants experienced behavioral facilitation from thematically related items in related-first blocks. However, when a delay was introduced and participants named unrelated items first (unrelated-first block), they instead experienced interference from thematic relations. Interestingly, younger adults experienced greater interference from related items than did older adults, but this effect diminished with repeated retrieval. Prior studies using the BCNP have identified an enhanced anterior N2 response to unrelated images relative to categorically related items (e.g., Janssen et al., 2011; 2015). However, in our study, N2 amplitudes varied by block order. For both groups, N2 amplitudes were more negative for related items in related-first blocks and for unrelated items in unrelated-first blocks, suggesting that the mechanisms underlying retrieval are sensitive to thematic context. Moreover, the enhanced interference effect for younger adults may indicate that older adults did not maintain the narrative content to the same extent, which is further supported by our finding of poorer recall of narrative content by older adults. In sum, neither the Dark Side Model nor the Swinging Lexical Network Hypothesis were fully supported. Instead, we suggest that working memory modulates the relationship between thematic context and retrieval in discourse settings. We explore how this evidence of age-related decay of thematic context may have implications for prior findings of weakened predictive processing and poorer coherence in discourse for older adults.
Topic Areas: Meaning: Lexical Semantics, Control, Selection, and Executive Processes