Computing Book Parts with EEBO-TCP

Abstract

This article begins by considering the difference between two kinds of bibliographical study: quantitative bibliography, which treats books as data, and 'anatomical' bibliography, which considers the paratextual and modular components of a book. Because of their methodological differences, it is difficult to reconcile the two approaches, so anatomical bibliography is rarely practiced on a computational scale. We propose a method of doing so using the markup added to early modern English texts by the Early English Books Online-Text Creation Partnership (EEBO-TCP). Keyers have not only transcribed texts, but described them too, adding XML tags that replicate the structure of the book. Among other things, these tags mark 'divisions' in the text, all of which have been assigned 'types' that describe parts of the book with terms such as 'preface,' or 'errata'. The article assesses this dataset by enumerating the division types and evaluating their consistency and scope. It reviews their previous applications, but argues that their full potential has not yet been realised, and offer suggestions for doing so. These suggestions are based on our work on early printed addresses to the reader, and we finish by describing our own method of studying division types with computational methods.