英文摘要

English abstracts

The multifactorial turn of statistical methods in learner corpus research and their applications

LI Yuanke， HE Anping amp; HUANG Lingmin （1）

Learner corpus research is undergoing an expansion in statistical methods， and exhibiting a noticeable shift from monofactorial methods to multifactorial modeling. However， domestic scholars have paid insufficient attention to this trend. This article begins by explicating the drawbacks of monofactorial methods and then dwells on the advantages of building statistical models. Two studies based on the National Matriculation Essay Corpus are then reported. The first study employs multinomial logistic regression to explore the mechanism whereby ten fine-grained syntactic features jointly affect essay scores. The second study utilizes structural equation modeling to examine the complex path relation through which the frequency， association strength， and accuracy of target language bundles produced by essay writers， measured on the basis of COCA， affect essay scores. The authors conclude by arguing that corpus researchers need to enhance their statistical literacy to meet the demands of adequately researching the complexity of language systems and improve the quality of scientific studies.

A multivariate quantitative study on English modal construction from a variationist linguistic perspective

LI Siyu， DAI Yaning amp; MENG Qingnan （14）

This study applies a corpus-based variationist linguistic perspective to explore the major factors influencing the choice of “must” “have to”， and “have got to” and their diachronic evolution in American English using COHA corpus data. The results show that from 1810 to 2009， when expressing the meaning of “necessary to do sth.”， there was a trend whereby “must” was gradually replaced by “have to”. The main factors influencing the selection of these three constructions ordered by importance are “clause tense” “genre” and “year”. In addition， through further analysis of the evolution of these three constructions， the authors argue that there has been a tendency for central modals to be gradually replaced by semi-modals in American English. Part of the meaning of the modal “must” is increasingly expressed by the semi-modal “have to”， while the use of the semi-modal “have got to” is on the decline.

A study of English accusational expressions in trade-conflict texts： A local function perspective

LIU Yunfeng （25）

Unlike the study of the generic function of describing the macro system of language， the study of local functions focuses on the specific function or meaning of language use in restricted texts， which is a developing trend in functional research in the context of corpus linguistics. Using the corpus-driven approach， this study conducts a contrastive analysis of English accusational expressions in China-U.S. anti-dumping texts from 2001 to 2017 with local functional categories and their constituent sequences. The analysis shows that these accusational expressions present regular features and development changes， indicating the differences in the understanding of anti-dumping practices between the discourse communities of China and U.S.， which has applications in China’s response to anti-dumping trade conflicts. Furthermore， the study also demonstrates the feasibility and usefulness of the local functional perspective in revealing subtle features of the text and provides a reference for text analysis.

A comparable-corpus-based study of phrasal complexity in academic writing in applied linguistics

GAO Xia （47）

This study analyzes the use and distribution of 17 complex noun phrases in L1 Chinese graduate students’ theses and L1 Chinese and English expert writers’ research articles in applied linguistics using a corpus-based approach. The results show that： （i） There is no significant difference in the use of 17 complex noun phrases among the three writer groups （ii） the most frequent complex noun phrases used by the three writer groups are the same; （iii） the use of three complex noun phrases is significantly different among the three writer groups; and （iv） L1 Chinese graduate students’ use of five complex noun phrases is significantly different from both L1 Chinese and English expert writers. The use of complex noun phrases is a better indicator of the complexity of academic writing. Pedagogical implications are provided for academic writing and EAP teachers.

Comparison between direct and inverse translations in rendering zhuyi terms in Mao Zedong’s works： A corpus-based investigation

SHI Xinyu amp; HUANG Libo （61）

The frequent occurrence of zhuyi terms in Mao Zedong’s works reflect Mao’s idiosyncratic characteristics in language use and contribute to the construction of his political discourse system. Starting with a classification of zhuyi terms in Mao Zedong Xuanji， the present paper makes a corpus-based comparison between direct and inverse translations of MaoZedong’s works in terms of their methods for translating zhuyi terms to explore the similarities and differences between the two types of translations in dealing with his idiosyncratic language use. The results suggest that neither type of translation displays complete inner-textual regularities， and both show some inter-textual similarities and differences. These findings can be explained from two perspectives： （a） Translated texts reflect a compromise and balance between adequacy and acceptability; （b） MaoZedong’s works and their direct translations occupy a high and authoritative status in China and the world at large. This paper further points out that the direct translation of political documents should ensure correctness， promptness， authoritativeness， and maintain Chinese characteristics and standards based on acceptability.

A review on text analysis in the research of finance and economics

NIU Huayong， DOU Yixuan amp; XIA Xiaoxue （81）

Previous studies in finance and economics have been based on the causal inference of econometric models. With the development of big data and computer algorithms， unstructured data represented by text information can be quantified and applied to research in finance and economics. The linguistic features of text information， such as text readability， intonation， and similarity， have gradually become the focus of scholarly research and quantification， hence， text analysis technology has been applied to research in this field. Starting from the language characteristics of text and different research issues in finance and economics， this paper combs the current domestic and foreign literature， analyzes the relationship between text and financial information， and points out direction for future development of text analysis technology to provide a reference for researchers in related fields.

A study of verb valency patterns used by learners： The case of “agree”

SUN Haiyan amp; NIU Wenshuang （96）

Valency pattern， an achievement of corpus phraseology， clearly shows the syntactic and semantic constraints of lexical items. Thus， they can help learners acquire lexical knowledge. Examining the high-frequency verb “agree”， this study compares the valency patterns used by Chinese EFL learners and native speakers based on corpus data. The results demonstrate that the frequency distribution is significantly different from that of native speakers. Moreover， learners misuse some valency patterns， and the actants they use lack semantic diversity， which indicates that the learners have not fully grasped the valency patterns of “agree”. This study analyzes the reasons for the learners’ misuse of valency patterns from the perspective of mother tongue transfer and classroom input， and suggests that the patterns should be applied to vocabulary teaching through hands-off data-driven learning and a valency pattern dictionary.

A corpus-based study of the discursive construction of China’s population policy in American news coverage

WANG Qin （109）

Combining the Discourse-Historical Approach （DHA） with corpus techniques， the present study built its own corpus and conducted analysis along three dimensions： themes， discourse strategies， and historical and social context. The findings indicate that the American media employed naming， predication， perspectivization， and argumentation strategies in order to construct two opposing parties： positive “selfness” and negative “otherness.” The American media’s attitudes toward China’s population policy underwent a dynamic diachronic change from total denial and criticism， by-stand and suspicion， to sarcasm and reluctant recognition. The present study sheds light on this frontier research methodology and theoretical framework for political discourse studies.