Through scalp measurements of the electrical activity of the brain (event-related potentials, or ERPs) recorded while subjects verified the truth of sentences relating exemplars and categories (e.g., ALL DOGS ARE ANIMALS), inferences were made about aspects of semantic processing that were not directly reflected by overt responses. In particular, it is suggested that a negative ERP component that peaks about 400 ms after the onset of the sentence predicate (i.e., N400) is sensitive to structural aspects of semantic memory. The amplitude of this component was modulated by the relatedness of the subject and predicate terms, as well as the hierarchical level of both these terms, but was not sensitive to the truth value of a sentence.
No takes yet. Share an insight, caveat, or question.
Kounios et al. (1992) studied this question.
Synapse has enriched one closely related paper. Consider it for comparative context: