The most personal artefact we possess for most of the men and women and children caught up in the criminal justice system of the eighteenth and nineteenth centuries is their words recorded as testimony, their voices. This project uses two sources of testimony (the Old Bailey Proceedings and Hobart Magistrates’ Court records) to identify styles and patterns of speech recorded in court, and then to characterise the individuals involved; and finally to relate changing patterns of verbal self-presentation, to defendants’ life course experiences. The project will use the individuals identified through nominal record linkage to create a subset of people for whom we have two or more records of spoken testimony.
By using Magnus Huber’s recently created Old Bailey Corpus, in combination with an additional body of Hobart records marked up to the same XML schema, a measure of variations in language will be added to data reflecting life course experience, allowing us to explore how ‘character’, or at least verbal self-presentation impacted on life changes. Text reflecting testimony will be extracted from the trials, and subjected to measures of vocabulary, word length and structural complexity, using Voyant Tools.
Additionally, testimony will be processed using a thesaurus-based, explicit semantic approach that will allow direct measures of sentiment. This builds on the work of Tim Hitchcock in collaboration with Simon DeDeo and Sara Klingenstein at the Santa Fe Institute. Automated clustering methodologies, including but not limited to, Bayesian nested sampling and K-means clustering will be used to explore the resulting data. In the process, the defendants’ ability to speak to power in their own voice will form a new measure and data point, through which to understand both the experience of individual defendants, and the factors determining the course of their lives, and relationship to power.