本研究通过三个线上阅读（词词呈现、部分部分呈现和整句呈现）测量反应时间的实验，旨在探讨基于频率标准的词串结构能否作为一个整体在语言处理中进行存储和提取。受试包括英语本族人、高水平和低水平英语学习者。结果发现，本族人和高水平英语学习者拥有相似的语言处理机制，能够利用词串结构的整体性更有效地处理实验句中的词串结构，而低水平英语学习者则不能。作者同时用研究结果对二语学习者的隐性词汇知识作了探讨。

The present study uses three self-paced reading experiments （word-by-word presentation, portion-by-portion experiment and sentence-by-sentence presentation） in order to explore if Lexical Bundles, the highly recurring word sequences, can be stored and accessed as a whole during language processing. The subjects include native English speakers and second language （L2） learners of high proficiency level and low proficiency level. The results show that the native speakers and L2 learners of high proficiency level possess a similar processing mechanism in that they can process the sentences with lexical bundles faster than their non-lexical bundles counterparts. L2 learners of low proficiency level, however, do not possess that mechanism. The author also relates the results to the study of L2 learners’ implicit lexical knowledge. Compared with the explicit knowledge, the implicit knowledge plays more important role in language processing. The unique role of lexical bundles in language processing makes it imperative to teach L2 learners those high frequency lexical bundles.