Forensic stylistics in an online world

Abstract

Forensic stylistics is the examination of style in language for the purpose of resolving litigated questions relating to disputed authorship or meaning. Forensic stylistics has been presented as evidence in a wide range of legal cases, mostly concerning the identity of the author of a document. The purpose of this paper is to discuss the use and potential of forensic stylistics in legal cases, with particular attention to analysing online activities. A review of previous legal cases where stylistics has proved significant is followed by a brief review of quantitative stylistics. The remainder of the paper discusses two empirical studies on the usefulness of a proposed simple stylistics technique, which claims 80% accuracy on determining the gender of the author just by counting the frequency of a small number of keywords. It is concluded that this technique only works in some domains of discussion, but does have some potential for initial scanning of large bodies of text.

Notes

1. W C Moore, Typist identification Nebraska State Bar Journal Vol 10, pp 48–52, 1961.

2. American Jurisprudence. A Modern Comprehensive Text Statement of American Law: State and Federal 2nd edn, Vol 29, Lawyers Cooperative, Rochester, NY, 1967.

3. Taken from a fuller list given in G McMenamin Forensic Science International (Special Issue on Forensic Stylistics), Vol 58, ch 5, 1993.

4. A S Osborn Questioned Documents, 2nd edn, Boyd, Albany, NY, 1929 (quoted in McMenamin, op cit, Note 3, ch 6).

5. M W Corney ‘Analysing e-mail text authorship for forensic purposes’ Master's Thesis, Queensland University of Technology, 2003, at http://sky.fit.qut.edu.au/∼corneym/papers/mit_thesis.pdf (last accessed 13 March 2006).

6. S E Fish ‘What is stylistics and why are they saying such terrible things about it?’ in J J Weber (ed) The Stylistics Reader Arnold, London, 1996, pp 94–116 (originally published 1973). See also J Rudman, D I Holmes, F J Tweedie and R H Baayen ‘The state of authorship attribution studies: (1) The history and the scope; (2) The problems—towards credibility and validity’ in Proceedings of the Joint International Conference of the Association for Computers and the Humanities and the Association for Literary & Linguistic Computing (ACH–ALLC'97), Queen's University, Kingston, Ontario, Canada, 3–7 June 1997, at http://www.cs.queensu.ca/achallc97/papers/s004.html (last accessed 13 March 2006) for a brief summary of individual methods applied to forensic stylistics that have been criticized, and T Grant ‘Forensic linguistics references’, 2005, at http://www.le.ac.uk/psychology/tg21/FL_References.html (last accessed 13 March 2006), accessed for a list of cases that have been decided using one of these methods, the CUSUM technique.

7. Stubbs ‘Conrad in the computer: examples of quantitative stylistic methods’ Language and Literature Vol 14, No 1, pp 5–24, 2004, also at http://lal.sagepub.com/cgi/reprint/14/1/5.pdf (last accessed 13 March 2006).

8. S Argamon, M Koppel, J Fine and A R Shimoni ‘Gender, genre and writing style in formal written texts’ Text Vol 23, No 3, August 2003, or http://www.cs.biu.ac.il/∼koppel/papers/male-female-text-final.pdf

9. See www.bookblog.net/gender/genie.html (last accessed 1 December 2005).

10. Stubbs, op cit, note 7.