Fluency in ENL, ESL and EFL: A contrastive corpus-based study of English as a first, second, and foreign language

本项目采用整体方法来调查英语作为第一语言（ENL）、第二语言（ESL）和外语（EFL）的流利程度。作为语料库分析的数据库，我们将使用国际英语语料库（ICE; Nelson 1996）和鲁汶国际英语口语中介语言数据库（LINDSEI; Gilquin et al. 2010）的几个组成部分。为了系统地评估这三种不同类型英语的使用者是如何建立流利性的，我们将对这些语料库进行分析，寻找可能对说话者的流利性产生影响的语言变量（即fluenceme； Götz 2013: 8-9）。为此，将使用一个基于流利基因的综合分类法，从三个不同的层次来描述流利性：(1)语音产生的时间变量（如运行长度、停顿率、语速等）以及说话者如何使用流利增强策略（如话语标记或预制单元）；(2)不同/几种流利素的（潜在组合）克服规划困难（流利素组块）以及它们在话语中的位置（流利素定位）；(3)流利素与语言外变量（如性别、语言外变量）的相关性；年龄)，可以预测不同类型英语中流利素的类型和使用（流利素偏好）。自动和人工从语料库中提取每个说话人的所有被调查的影响素，并进行定性后处理和消歧，然后进行统计分析。这个过程可以测试不同类型的英语使用者是否比其他人更流利，是否在不同类型的英语中以不同的方式（例如通过使用不同的流利音素）建立流利，是否在不同类型的英语中以不同的方式和在话语中的不同位置克服计划阶段，以及是否可以通过语言外参数（如年龄或性别）来预测流利度和提高流利度的策略。因此，本项目首次对不同类型英语的流利性进行了全面的描述，并将对多样性研究、第二语言习得和应用语料库语言学做出相关贡献。

The present project takes a holistic approach to investigate the fluency of speakers of English as a first language (ENL), a second language (ESL) and as a foreign language (EFL). As a database for the corpus analysis, we will use several components of the International Corpus of English (ICE; Nelson 1996) and the Louvain International Database of Spoken English Interlanguage (LINDSEI; Gilquin et al. 2010). In order to systematically assess how speakers of these three different types of Englishes establish fluency, these corpora will be analyzed for the linguistic variables that can potentially have an effect on the fluency of a speaker (i.e. fluencemes; Götz 2013: 8-9). For this purpose, an integrated, fluenceme-based taxonomy will be used to make it possible to describe fluency on three different levels: (1) the temporal variables in speech production (e.g. the length of runs, pause ratio, speech rate, etc.) as well as how speakers make use of fluency-enhancing strategies (e.g. discourse markers or prefabricated units), (2) the (potential combinations of) different/several fluencemes to overcome planning difficulties (fluenceme chunking) as well as their positions in the utterance (fluenceme positioning), and (3) correlations of fluencemes with extralinguistic variables (e.g. gender, age) that can predict the type and use of fluencemes in different types of Englishes (fluenceme preferencing).All the investigated fluencemes per speaker will be extracted from the corpora automatically and manually and qualitatively post-processed and disambiguated, after which they can be statistically analyzed. This process makes it possible to test if speakers of different types of Englishes are more or less fluent than others, if fluency is established in different ways (e.g. by using different fluencemes) in different types of Englishes, if planning phases are overcome in different ways in different types of Englishes and in different positions in the utterance, and if fluency and fluency-enhancing strategies can be predicted by extra-linguistic parameters, such as age or gender.The present project thus presents the first holistic description of fluency in different types of Englishes contrastively and will thus make a relevant contribution to variety studies, second language acquisition and applied corpus linguistics.