A Corpus-based Evaluation Of Vocabulary In English Textbooks For Secondary School

The present study is an empirical investigation of textbook vocabulary based onCorpus of English Textbooks for Secondary School set up by Schools of Foreign Studies,South China Normal University. Previous studies of textbook vocabulary mostlyconcentrate on the width of vocabulary. So the present study attempts to investigate boththe width and the depth of vocabulary in English textbooks for secondary school designedunder the guidance of New Curriculum Standard. The research questions are:1) How well do English textbooks for secondary school cover Curriculum Words andBasic Words?2) How well do English textbooks for secondary school present common senses andtypical patterns?Curriculum Words here refer to words prescribed by New Curriculum Standard (3398lemma-words in total). Basic Words (2599 lemma-words in total) are set up by the currentresearcher, which come from a balanced combination of three top 2000 high frequencywords and two dictionary definition words.The research data mainly include 1) Corpus of English Textbooks for SecondarySchool, to be exact, it includes a set of student books for both lower secondary school(five books) and upper secondary school (eight books); 2) Referential data, includingword lists from Curriculum Words and Basic Words, and sense and pattem information incorpus-based Longman Dictionary of Contemporary English (4th edition) (henceforthLDOCE). For research question one, the investigation of the textbooks' coverage ofCurriculum Words and Basic Words is mainly through various word list comparisons byWordsmith. For Curriculum Words and Basic Words which do not occur in the textbooks,further analysis is made from word class and semantic feature. For research question two,it first samples 26 Basic Words, which spread the main content word class and variousoccurrences in the textbooks. Then all instances of each sampled word are retrieved byWordsmith and are further compared with the most common senses (top three senses inLDOCE) and typical patterns in LDOCE.The investigation shows that there is still room for improvement for the investigatedtextbooks in both the coverage of Curriculum Words and Basic Words, and also in thepresentation of their common senses and typical patterns. The investigation finds thatthere is about 16 percent of Curriculum Words which do not occur in the textbooks. Furthermore, there is about 14 percent of Curriculum Words which occur only once ortwice. However, in most cases, one or two occurrences are inadequate to present thecommon senses and typical patterns. Further investigation finds that among CurriculumWords, words of level 8 are most poorly .covered. For the coverage of Basic Words,investigation shows that the textbooks have a better coverage but still there is about 22percent of Basic Words which do not occur or occur only once or twice in the textbooks.Semantic analysis shows that they are mostly abstract in meaning.Sense and pattern analysis of the sampled Basic Words suggests that the investigatedtextbooks seem to have a preference for certain word class or senses and thus ignore orunder-represent other also important word class and senses. Analysis also shows thatwords with multiple class are most poorly represented and words with low occurrencesare easily to miss some common senses and typical patterns. Analysis also shows that theinvestigated textbooks tend to ignore or under-represent the abstract meanings.The present study is expected to have three implications: 1) The study reveals thelimitations of the investigated textbooks in both word coverage and also presentation ofcommon senses and typical patterns; 2) The study proposes a new evaluative frameworkfor textbook vocabulary; and 3) the study demonstrates the great significance of corpusapproach for materials development and evaluation.