Abstracting science: A corpus-based approach to research article abstracts

Abstracts are largely based on rhetorical activity that involves interactions between writers and readers. Although abstracts, belonging to the scientific scenario, as texts, are not necessarily more objective than others, they are simply more effective in hiding subjectivity linguistically (Bhatia, 1993, 1997). The aim of the present research paper is to investigate the way authors assert and present their results to their peers. Data for the present research paper were drawn from a corpus made up of 1,035 abstracts through a corpus-based analysis. Focus was on recurrent sets of constructions and on specific words defined in the study as research process words. The comparative distributional frequencies were the basis for assessing the pragmatic and the rhetorical role played by these words. The overall purpose was to test the hypothesis that the existence of recurrent expressions related to genre specificity and to the language itself.