Cognitive motivations behind absent agents in passive sentences across Chinese and English

Abstract Passive sentences featuring absent agents are prevalent in both Chinese and English, yet the cognitive motivations behind these agents have not received sufficient attention. To fill the gap, the current study investigates the differences in comprehension difficulty metrics between passive sentences with and without agents in both languages. Using mixed-effects logistic regression and conditional inference tree models, we find that in Chinese passives, the inaccessibility of subjects and the low predictability of subject dependencies pose comprehension difficulty. Consequently, agents are absent to alleviate the difficulty by reducing the memory burden of subject dependencies. In English passives, the inaccessibility of subjects and the heavy memory load of subject dependencies cause comprehension difficulty. Therefore, agents are absent to mitigate this difficulty by eliminating the memory burden of processing the dependencies between verbs and agents. In conclusion, overall efficient comprehension drives the absence of agents in passive sentences across both Chinese and English.