ABSTRACT
This chapter presents a pilot study of register variation in English with the Corpus of Contemporary American English (COCA). It explains to replicate Biber's 1988 study with Correspondence Analysis. Both the two-dimensional analysis and the interpretation of the dimensions will be supported by the cross-linguistic comparisons. In the primary horizontal dimension, the right side is populated by features such as personal pronouns, colloquial expressions, contracted forms, interjections, colon, question, exclamation and quotation marks. In the secondary, vertical dimension, there naturally would not be classical elements in the Chinese sense. The distributional pattern that find may shed light on the interpretation of the secondary dimension in Chinese. The distribution of the registers on both dimensions shows the distinctness of fiction, which differs both from the extreme literateness of academic writing and the extreme non-literariness of spoken discourse. The chapter describes the alternative statistical procedure of Correspondence Analysis yielded a smaller number of dimensions than Biber's Factor Analysis.
