Synopses & Reviews
Implementing a novel method for identifying idiolectal co-selections and taking the UNABOM investigation as a case study, this book evaluates the effectiveness and reliability of using the web for forensic purposes, since forensic linguists often use the web as a tool in forensic authorship analysis cases. A notable example includes the American trial of the Unabomber, in which the web was searched for idiolectal co-selections, and claims made about their distinctiveness based on the number of search results.
In order to assess whether the web can reliably be used in this way, two experiments are described in which the web is used to attribute documents to their authors, both when texts produced by candidate authors are, and are not, available for comparison. With considerable unreliability in the results- both between different search engines and over time - this book argues that more caution is required in using the web as a corpus for forensic purposes.
About the Author
Samuel Larner is a Lecturer in Forensic Linguistics at the University of Central Lancashire, UK. His research interests are primarily in forensic linguistics, specifically authorship analysis and deception detection. Samuel is a Fellow of the Higher Education Academy and teaches both undergraduate and postgraduate modules in linguistics, contributing in particular to an innovative and interdisciplinary MSc in Emotions, Credibility and Deception.
Table of Contents
1. Introduction: The UNABOM Investigation
2. The Web as Corpus and Authorship Attribution
3. Attributing Documents to Unknown Authors Using Idiolectal Co-selection and the Web
4. Attributing Documents to Candidate Authors Using Idiolectal Co-Selection and the Web
5. The Suitability of the Web for Producing Forensic Evidence