Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Semantic Transparency and Chinese Compound Processing: An ERP Analysis
Poster A16 in Poster Session A, Tuesday, October 24, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
This poster is part of the Sandbox Series.
Yihan Chen1, Eleonora Rossi1; 1University of Florida
Compounding is a morphologically highly productive word formation in languages worldwide. However, its processing is still a debated topic in psycholinguistics. Two major theoretical models have been proposed: the full decomposition model (Taft, 2003; Taft & Forster, 1975), which posits that all morphologically complex words are initially decomposed into internal morphemes during processing, and the dual route model (Baayen et al., 1997; Bertram et al., 2000), which asserts that 1) both decomposition and whole-word access take place during compound processing, and 2) the route taken depends on the type of compound. One key feature that may affect compound processing is semantic transparency, which refers to the relatedness/association between the internal constituent and the overall meaning of the compounds (Kim et al., 2019). Several neuroimaging studies have investigated the time course of compound processing with millisecond precision via electroencephalography (EEG) – ultimately revealing mixed findings. Some studies found that transparent compounds have a higher probability of being decomposed, and opaque compounds are accessed instead as whole units – supporting the dual route model. MacGregor and Shtyrov (2013) and Tsang et al. (2022) employed a passive-listening oddball paradigm in English and Mandarin respectively and observed an increased Mismatch Negativity (indicating lexical enhancement) for opaque compounds compared to transparent compounds. In addition, Bai et al. (2008) found a larger N400 for opaque than transparent compounds using an auditory lexical decision task Other studies demonstrate that all compounds undergo full decomposition regardless of their semantic transparency. For example, Wei et al. (2023) used a visual lexical decision task and found that both transparent and opaque compounds induced a stronger Left Anterior Negativity (LAN) compared to monomorphemic words, but there was no difference between the two compound types in both LAN and N400. One factor that may contribute to these inconsistent findings is that these studies treat transparency as a binary feature, without considering the presence of partially transparent/opaque compounds. Therefore, this study aims to investigate if and to what extent the degree of transparency in semantics will impact compound processing. In the proposed study, native speakers of Mandarin will complete a visual lexical decision task in Mandarin while their EEG will be recorded. The stimuli will consist of real compound words and pseudo compound words. Critically, the real compound words will further be categorized based on their degree of transparency into 1) transparent-transparent compounds (e.g., 火山 ‘fire + mountain’ volcano), transparent-opaque compounds (e.g., 笨蛋 ‘stupid + egg’ idiot), opaque-transparent compounds (e.g., 海报 ‘sea + paper’ poster) and opaque-opaque compounds (e.g., 东西 ‘east + west’ thing). We hypothesize that real compounds will elicit a smaller N400 compared to pseudo compounds. Furthermore, within the category of real compounds, we expect to observe the smallest N400 amplitude in fully transparent compounds, the largest amplitude in fully opaque compounds, and an intermediate amplitude in partially transparent compounds. Our findings will shed light on the intricate mechanisms involved in compound processing and contribute to a broader understanding of word storage in the mental lexicon.
Topic Areas: Meaning: Lexical Semantics,