Phraseological competence in spoken English: Using corpus methods to inform language testing

该项目解决了英国和国际上语言学习和测试中一个非常突出的问题——L1（母语）和L2（非母语）英语口语中的短语能力（PC）的构建及其在语言测试实践中的应用。本研究从一个新颖的方向切入，将最先进的语料库方法与语言测试研究的最新进展相结合。除了对语言习得和语言测试理论的贡献外，该研究还寻求通过与主要的国际语言考试机构伦敦三一学院密切合作，提高英语口语的评估水平，并将研究结果与语言测试开发者直接相关。人类实时流畅交流的能力在很大程度上取决于PC，也就是说，存储、访问和产生预制语言块的能力，比如多词表达（would like to）或词典语法框架（就X而言）。这种能力是用母语一样轻松无误的方式进行交流的一个关键方面。因此，几十年来，PC在母语和二语使用者的语言学习/使用研究中占据了突出的地位。然而，尽管关注到目前为止，我们对PC的组成部分及其在语言产生和感知中的作用的理解仍然存在重大差距。这些都与几个方法问题有关。例如，大多数关于PC的证据来自基于非语境化语言的实验心理语言学研究；语境化（语料库）研究PC几乎完全是基于书面生产。此外，许多可用的知识依赖于对短语单位的主观识别和分析。

The project addresses an issue of high prominence in language learning and testing in the UK and internationally - the construct of phraseological competence (PC) in L1 (native) and L2 (non-native) spoken English and its application in the language testing practice. The study approaches this topic from a novel direction, combining state-of-the-art corpus methods with the latest advances in language testing research. In addition to the contribution to the language acquisition and language testing theory, the study seeks to improve assessment of spoken English by closely collaborating with Trinity College London, a major international language examination organisation, and by making the findings directly relevant to language test developers. Human ability to communicate fluently in real time depends to a large extent on the PC, that is, the ability to store, access and produce pre-fabricated chunks of language such as multi-word expressions (would like to) or lexico-grammatical frames (as far as X is concerned). This competence represents a key aspect in communicating in a native-like, effortless and error-free manner. As a result, PC has occupied a prominent role in research on language learning/use by LI and L2 users for decades. However, despite the attention given to the topic so far, significant gaps still remain in our understanding of the components of PC and its role in language production and perception. These are related to several methodological issues. For example, most of the evidence about PC comes from experimental, psycholinguistic research based on decontextualized language; contextualised (corpus) studies of PC have almost exclusively been based on written production. Also, much of the available knowledge relies on subjective identification and analysis of phraseological units.