Information structure in Mandarin Chinese light verb constructions

Mandarin light verb constructions, which can be categorized into DO and GIVE groups based on the semantic content of the light verb, are generally believed to be associated with information packaging: in the schematised canonical syntactic realisation ̳dui ̳to‘+NP+LV+NP‘, the preposed preverbal undergoer argument has a contrastive focus function. However, a plethora of LVCs retrieved from a Mandarin corpus implies that the construction per se does not necessarily involve a notion of contrast. Moreover, besides the canonical oblique construction, the undergoer argument can be realised postverbally, resulting in the light verb followed by two nominal constituents. Corpus data found that the choice of the two syntactic variants is not at random; rather it is conditioned by the degree of identifiability of the undergoer argument. In light of inheritance relations between related constructions and the grammaticalisation of light verb constructions, a unified syntactic pattern is generalised with regard to the identifiability scale.