Neural Mechanisms underlying Different Degrees of Causal Inference in Processing Chinese Word Pairs

Abstract

The present study used ERP technique to uncover the neural mechanisms underlying different degrees of causal inference in processing Chinese word pairs. We asked participants to read and judge the causal relatedness of word pairs with different degrees of causal relation, namely, highly-related, intermediately-related and unrelated causal condition. The results showed that the intermediately-related causal conditions elicited a larger N400 and a smaller late positive component (LPC) than the highly-related causal conditions. These results indicate that processing of intermediately-related causal relation requires more cognitive effort relative to highly-related causal relation, and the causal inference can influence early semantic processing and later integrative processing of an incoming word. The results empirically support the memory-based models of discourse processing generally, at least when readers participate in establishing causal coherence in Chinese word pairs.