本研究旨在測試Yuan and Zhao 在2005年和Tsimpli and Dimitrakopoulou在2007年關於成年第二語中文之英語使用者對集合代名詞之參數重設詮釋。54名不同語言程度的英文使用者被要求更正含有集合名詞之不合語法的句子；受測者的答案將和中文母語使用者之對照組做比對。實驗結果反駁了Yuan and Zhao的“輸入驅使參數重設”論點。取而代之的是:不可詮釋的語法對內在文法建構與關鍵期之可操控性有關，並支持Tsimpli and Dimitrakopoulou的“可詮釋性假說”。值得爭議的是，藉由延伸不可詮釋之語法特質假定的不可及，一旦不可詮釋的語法特質被選定，如果第二語言也缺乏這不可詮釋的特質它們便不易被省略。

The study sets to test proposals made by Yuan and Zhao (2005) and Tsimpli and Dimitrakopoulou (2007) in relation to the issue of parameter resetting in the interpretation of resumptive pronouns by adult English speakers of L2 Chinese. Fiftyfour English speakers of different proficiency levels were asked to correct sentences involving ungrammatical resumptive pronouns in L2 Chinese and their responses were compared with those of a native speaker control group. Findings of the study argue against Yuan and Zhao’s input-driven parameter resetting account. Instead, they support the Interpretability Hypothesis of Tsimpli and Dimitrakopoulou in assuming that there is a critical period for the accessibility of uninterpretable syntactic features for the construction of mental grammars. It is argued, by extending the unattainability of the uninterpretable features assumption, that once the uninterpretable syntactic features are selected, they become difficult to lose if L2 lacks such uninterpretable features.