This study brings a synchronic and a diachronic perspective to the study of lexicalisation in Japanese and Chinese, with focuses upon motion and change-of-state constructions. It aims to uncover how the lexical, morphological and syntactic resources of the two languages play essential roles when it comes to deciding which component of events is characteristically rendered and with what preference. Synchronically, the two languages show a certain degree of similarity in the way they render the core schema of an event, such as through verb compounds. The morpho-syntatic level mainly facilities lexicalisation in Modern Japanese and it turns out that Japanese is not a pure verb-framed language: events with path conveyed via verb root, open-scale AP, means/caused/manner V-V, complement relation V-V, and participle complex predicate exhibit verb-framed behaviour, events with path rendered by closed-scale AP, PP suggest satellite framing, and events with path conflated in pair relation V-V present equipollent framing. Lexicalisation occurs at the level of lexical and syntactic interface in Modern Chinese.
The finding brings us to the point that equipollent framing is not valid in relation to Chinese SVCs, as the multi-morphemes in SVCs are not equipollent. This is probably down to the different degrees of grammaticalisation that each constituent has received. Diachronically, Japanese and Chinese have undeniable similarities in favouring a single verb to convey the core schema but meanwhile the two present distinctions in regard to morphology, i.e. prefix, preverb, and incorporated noun. A proposal to treat the intralinguistic and cross-linguistic variation is put forward: it is the distinct grammatical elements and the way they are selected and combined that facilitate the determination of lexicalisation.