The Emergence of the 吃-passive Construction: Based on the Grammaticalization Path of ‘to Eat’-Type Verbs

This study explores the emergence of the 吃-passive construction based on the semantic features of the verb 'to eat' and its cross-linguistic grammaticalization path.The 吃-passive construction first appears in bianwen 變文 during the Tang dynasty. It was broadly used in the Ming dynasty vernacular novels. Its use drastically plunged after the early Qing. In modern Chinese, 吃-passive exists vestigially only in some lexicons, such as 吃虧 'suffer losses,' 吃苦 bear hardships,' 吃驚 'be startled.'By analyzing the occurrences of 吃 in modern Chinese, it is argued that the core meaning of 吃 is not just 'to eat,' but 'to take,' 'to accept,' or 'to absorb.' Due to its immanent semantic features of 'to eat'-type verbs, 吃 was very prone to be grammaticalized into a passive marker.Kuteva et al.(2019) provided fourteen different origins of the passive markers. 'To eat'-type verbs are one of the common sources of a passive marker cross-linguistically.This study explains the development of the 'to eat'-type verb into a passive marker using the concept of transitivity suggested by Nass(2007). The 'to eat' type verb developed into a passive marker via two possible intermediate stages(i.e. reflexive and anticausative). The grammaticalization path of 'to eat'> (reflexive) > (anticausative) > passive marker was instigated due to the (1) low transitivity of the 'to eat'-type verbs and (2) its inborn nature of "agent affectedness." As the grammaticalization continues, the verb's transitivity gradually weakens, and the degree of agent affectedness increases. Finally, the passive construction shows the minimal degree of transitivity. On the other hand, its subject, whose semantic role is the patient, shows the maximal degree of agent affectedness. In sum, the passive construction is the opposite of the maximally transitive clause.