Textual Positions of Lexical Bundles across Newspaper Genres

ABSTRACT: According to Hoey (2005), words or their combinations may be primed to occur or avoid occurring at the beginning or end of independently recognized discourse units, like the sentence and the paragraph; these are their textual colligations. This study attempted to investigate the textual positions of two English lexical bundles across two newspaper corpora, each including approximately one million words. The genres considered were economic and political news report, recently downloaded from internationally famous newspapers. To this end, Hoey and O’Donnell’s (2008) dichotomy of textual positions was employed for all the analyses. According to this framework, textual positions comprise: first sentence of text, first sentence of a paragraph that is not the first in text, and any sentence that does not begin a paragraph or a text. Furthermore, based on Hyland’s (2008) taxonomy, two lexical bundles which frequently occurred in the two corpora were selected. In addition, the Wordsmith Tools software (Scott, 2010) was used to conduct the comparisons. The frequency and percentage of each bundle was calculated; then chi-square was applied and the results revealed significant generic differences in the use of the bundles with regard to their textual positions.