Plagiarism in student’s writing has long been a prolific issue facing many academic institutions. With today’s plagiarists becoming increasingly aware of the limitations of existing detection systems and ‘idea’ plagiarism now something of an academic conundrum, this study firstly seeks to relate linguistic patterns of plagiarism to the computerized textual features utilised in plagiarism detection tools and secondly employs stylometry from authorship analysis to generate a statistical analysis of the literary style. It begins with the premise that plagiarism has already occurred and aims to determine whether the stylistic, structural and semantic choices of writers can detect the direction of ‘idea’ plagiarism and to what extent this grammar-based method is effective. The study used ‘known’ plagiarism writings between two undergraduate students to test whether the original and plagiarised documents were identifiable through the use of both existing and novel methods within the discipline of forensic linguistics. Through qualitative and statistical data analysis, including correlation graphs, the conclusion was reached that ideas can be easily merged with a writer’s own idiolect which renders ‘idea’ plagiarism almost impossible to identify with another writer’s typical language features. Keywords: Forensic Linguistics, Plagiarism, Originality, Authorship, Idiolect.
- There are currently no refbacks.
Copyright © 2014 University of Central Lancashire