An Experimental Investigation of Multiple Sluicing in Mandarin Chinese

This paper examines multiple sluicing constructions in Mandarin Chinese (henceforth, MC) experimentally. The acceptability status of such constructions in MC is controversial, and the judgments reported in the previous literature vary. Obtaining experimental evidence on the acceptability status is, therefore, important to advance the research on multiple sluicing in MC. Consequently, the present study conducts two sets of experiments to investigate factors affecting the acceptability of multiple sluicing sentences and the influence of the distribution of shi preceding wh-remnants on acceptability ratings. The results show that multiple sluicing in MC is generally a marked construction. Nevertheless, factors including prepositionhood and specificity have ameliorating effects on the acceptability of such constructions. Moreover, the influence of the distribution of shi on the acceptability ratings is related to the nature of wh-remnants; that is, its presence significantly improves the acceptability of cases of multiple sluicing when it precedes bare wh-arguments. We argue that the observed ameliorating effects on multiple sluicing can be explained by a cue-based retrieval approach to cross-linguistic elliptical constructions. Compared to bare wh-arguments, prepositional and discourse-linked wh-phrases provide cues to facilitate the retrieval of information from antecedent clauses.