A Report On The Translation Of An Excerpt From Corpus Linguistics And The Analysis Of Sociolinguistic Change

Corpus Linguistics and the Analysis of Sociolinguistic Change: Language Variety and Ideology in Advertising was published by Routledge in 2020.Drawing on a corpus of ads broadcast on an Irish radio station between 1977 and 2017,this book uses corpus tools to reveal the relationship between language ideology and sociolinguistic change by discussing the media’s choice of specific language styles and varieties.This book can provide references for corpus-based interdisciplinary research,and also provide a new perspective for domestic corpus and sociolinguistic research.This translation practice selects the first and second chapters of the book as translation materials,and adopts the mode of "machine translation + post-editing" which means to machine-translate the original text and then post-edit it.On the basis of post-editing,the machine translation errors are marked and classified,the distribution of errors is counted,the causes of errors are analyzed,and post-editing methods are summarized.The survey found that: at the lexical level,term translation and abbreviation translation are the most frequent types of errors,accounting for more than 50%,while non-linguistic lexical errors account for nearly 10%,which also need attention;at the syntactic level,improper sentence structure has the highest frequency,reaching 70%,and non-linguistic translation errors of no-need-to-be-translated sentences account for 12%,which also deserves more attention.The post-editing methods used here include alternative translation,omission translation,supplementary translation,sentence structure reorganization and adjustment of word order.This translation practice takes academic texts as the translation object.The summarized machine translation errors and revision methods may provide references for the use and research of post-editing methods,and may also for optimizing the performance of machine translation systems.