Key colligation analysis: Discovering stylistic differences in significant lexico-grammatical units

This methodology-based research proposes key colligation analysis to deals with grammatical relations in a particular dataset at a phrasal level based on statistics, while keyword analysis treats statistically outstanding words rather than phrases, irrespective of grammatical relations or the word class. The method can reveal naturally occurring sequences of subjects and predicates, including verbs, voice, and auxiliaries, if any, so as to detect the subtle stylistic differences which they create between the texts. This is demonstrated by British newspaper editorials, which share the same register and genre but bear distinct characters. The results reveal a diversity of argumentative styles: personality vs. impersonality, involvement vs. detachment, authority vs. intertextuality, for instance. Since quality newspapers are a precious resource for research, it is worthwhile investigating the variations among them for proper linguistic generalisation. The implications of this research are applicable to disciplinary variations