ABSTRACT
This chapter argues that the analysis for similarity can also be profitably expanded beyond the corpus selection/creation stage into the discourse analysis. It also argue that this aspect is somewhat neglected within corpus and discourse studies. The chapter identifies all articles published in Corpora and International Journal of Corpus Linguistics since 2015 which referred to discourse or stylistics in the title, keywords or abstract and concordanced them for references to same/similar and differen. The classic corpus linguistic entry points would include the analysis of concordances, collocates and keywords/key clusters and key semantic domains. Regarding the measure of collocation, it was essential to keep this the same for the two corpora and so the measure used had to be mutual information because this is what is available within the brigham young university (BYU) interface. There are also tools which facilitate the comparison of collocates, such as Sketch Engines' Sketch Difference and the 'compare' function in the BYU interface.
