Testing the weak NP analysis of gapless bei sentences in Mandarin Chinese: Implications for the affectedness typology of passives

Motivated by the lack of a consensus on the basic judgment data on gapless (or indirect exclusive) bei passives in Chinese, in this study we used quantitative methods to test the weak NP analysis, proposed by Lin (2009, Licensing "gapless" bei passives. Journal of East Asian Linguistics 18(2). 167-177), which is constructed on informally collected acceptability judgment data. Three experiments were designed to obtain a comprehensive body of judgments that followed the rigorous data collection standards as advocated by Gibson and Fedorenko (2013, The need for quantitative methods in syntax and semantics research. Language and Cognitive Processes 28(1-2). 88-124). The experimental results showed that despite acceptable indirect exclusive bei sentences observed in the literature, the generation of such bei passives is limited. The results also cast doubt on the weak NP hypothesis for gapless bei passives given that the test items all received low acceptability irrespective of the presence of the alleged weak NP licensor. In addition, the experimental results revealed that the retained object in indirect bei passives cannot be licensed by any pragmatically significant relation, such as a direct kinship relation manipulated in the experiments. Taken together, the experimental findings were shown to draw implications for the affectedness typology of passives. We argue that the continuity approach (Shibatani 1985, Passives and related constructions: A prototype analysis. Language 61(4). 821-848; Washio 1993, When causatives mean passive. Journal of East Asian Linguistics 2. 45-90) fares better than the core syntax/semantic approach (Oshima 2006, Adversity and Korean/Japanese passives: Constructional analogy. Journal of East Asian Linguistics 15. 137-166) to capturing the affectedness patterns of passives in Korean, Japanese and Chinese.