This article investigates whether explicit instruction can affect second language (L2) competence in the domain of English quantifier scope. An intervention study was conducted with L1-Mandarin L2-English learners in order to examine (1) whether these learners can learn inverse scope for the structure on which they are instructed (double-quantifier configuration: A dog scared every man; quantifier-negation configuration: Every sheep did not jump over the fence); (2) whether they can correctly generalize availability of inverse scope from the configuration on which they are instructed to one on which they are not instructed; and (3) whether learners overgeneralize inverse scope to a superficially similar configuration that does not allow inverse scope. Following Schwartz (1993), generalization is taken to be a hallmark of true acquisition (= changes to linguistic competence) as opposed to learning. The results show that learners successfully learn inverse scope for the configuration on which they are instructed, but do not generalize availability of inverse scope to the other configuration. Moreover, learners instructed on double-quantifier configurations overgeneralize availability of inverse scope to island configurations such as There is one dog which scared every man. The findings indicate that in this domain, explicit instruction does not affect linguistic competence.

本文研究了显式指令是否会影响英语量词范围内的第二语言（L2）能力。为了研究（1）这些学习者是否可以学习所指导的结构的反作用域，对L1-普通话L2英语学习者进行了干预研究（双重量词配置：一只狗吓到了每个人；量词否定配置：每只羊都没有跳过篱笆）; （2）是否可以正确地归纳从指示范围到不指示范围的逆范围的可用性；（3）学习者是否将反作用域过度概括为不允许反作用域的表面相似配置。继施瓦茨（Schwartz，1993）之后，泛化被视为与学习相反的真正习得（即语言能力的改变）的标志。结果表明，学习者成功地学习了其所指导的配置的逆范围，但没有将逆范围的可用性推广到其他配置。而且，在双重量化器配置方面受过指导的学习者过分概括了反作用域对岛配置的可用性，例如，有一只狗吓到了每个人。研究结果表明，在这一领域，明确的指导不会影响语言能力。