So fuckin' annoying. This presents an implementation of a tool designed to measure text similarity and tease out the difference between semantic (dis)similarity and stylistic (dis)similarity between iterations of a text (where the former generally involves much more word-over-word difference, changing "what is said" to change "the meaning that is conveyed", while great stylistic change can be achieved with only one or two changes). Fine, such worthy, so scholarship. But when you intend only to present the numbers in this regard and not actually convey the nature of the changes between the first and second versions of Fowles's novel, even in terms of presenting the basic data/facts/examples of what changes (let alone doing any literary analysis of how they might affect the reading of the novel), which is surely what the huge bulk of people who might seek out a paper with this title are looking for, and when you do not clearly convey that you don't intend to do that in the title as well as the abstract, you're a dickhead. Paper given at PALA 26, Sheffield. ( )