Academic journal article Canadian Journal of Experimental Psychology

Semantic Richness, Concreteness, and Object Domain: An Electrophysiological Study

Academic journal article Canadian Journal of Experimental Psychology

Semantic Richness, Concreteness, and Object Domain: An Electrophysiological Study

Article excerpt

Results from previous event-related potential (ERP) studies of semantic richness and concreteness effects have been mixed. Feature production norms have been used to derive one measure of semantic richness, the number of listed semantic features (NOF) for a given concept. Whereas some ERP studies have found evidence for a semantic richness continuum from abstract concepts, to concrete concepts with few features, to concrete concepts with several features, other studies have not. The present study assessed the effects of NOF (within concrete concepts) and concreteness (concrete vs. abstract concepts), on ERP amplitudes and behavioural decision latencies during a concrete/abstract decision task. It is important we also manipulated object domain, which has been found to influence ERP amplitude and topography. High and low NOF concepts were selected from animal and nonliving thing categories and all four conditions were matched on several potential confounds. We show that although decision latencies support a semantic richness continuum, electrophysiological activity does not. Whereas concrete concepts produce a larger negativity than abstract concepts, low NOF concepts are associated with larger negativities than high NOF concepts. We also replicate an increased posterior positivity for processing animal concepts, and report an interaction between object domain and semantic richness such that the NOF effect is larger within animal concepts.

Keywords: semantic richness, word recognition, concreteness, ERP, event-related potential

Humans possess the remarkable ability to comprehend the meanings of words in less than half a second. Electroencephalog- raphy (EEG) has been very useful for understanding how this occurs in real time, which is especially important for constraining models of word recognition and semantic memory organisation. In the current work we address the effects of semantic richness and object domain (animals and nonliving things) on brain activity accompanying single word reading. Specifically, we obtain event- related potentials (ERP) while participants perform a concreteness decision task on a set of concrete and abstract nouns. Our goal is to use our findings to inform current theories of semantic cognition and to extend the recent literature on semantic richness.

Event-Related Potentials and Semantic Memory

The ERP technique, in which an average EEG signal across participants is computed relative to stimulus onset, is well suited for studying online meaning computation because it provides millisecond temporal precision and can be used to analyse small, fast-decaying neural signals invisible to functional magnetic res- onance imaging (fMRI) and positron emission tomography (PET; e.g., Greenwald, Draine, & Abrams, 1996). One particularly rele- vant ERP component for the current study is the N400, which in part reflects a temporally dynamic interaction between the context in which a word occurs and the word's meaning (Kutas & Federmeier, 2000). The N400 time window may encompass at least two separable components. The classic N400 component introduced above is characterised by a central-parietal scalp dis- tribution. In contrast, pictures (vs. words) and concrete words (vs. abstract words) have been shown to produce a more anterior N400-like component (Ganis, Kutas, & Sereno, 1996; Holcomb, Kounios, Anderson, & West, 1999: Kounios & Holcomb, 1994), which has been suggested to reflect activation of different cortical networks underlying concrete versus abstract knowledge. Specif- ically, Holcomb et al. (1999) suggest that the N400 should be divided into a posterior linguistically sensitive N400 that is sensi- tive to all word types, and a frontal imagistically sensitive N400 that is biased toward concrete nouns.

This more frontally distributed N400 effect has been used to study how different kinds of knowledge may be organized in the brain. The primary goal of this approach has been to determine whether computing the meanings of different types of stimuli is associated with separable temporal and/or spatial neurocognitive signals. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.