This volume follows and expands on Shakespeare, Computers, and the Mystery of Authorship (ed. Hugh Craig and Arthur F. Kinney [2009]). Using methods from quantitative stylometrics that have been applied most often to problems of authorial attribution, Craig and Greatley-Hirsch tackle questions of character, genre, periodization, and repertory in several new case studies of English plays performed between 1580 and 1669. The project offers a number of useful provocations to scholars of early modern drama and theater, and the authors document their procedures rigorously enough that readers with varying levels of technological expertise will be able to understand the methods used and to assess the significance of the findings.

The introduction establishes the authors’ view of the place of computers within the study of early modern drama and articulates their goal of “combining large datasets and computerised statistical analysis to elicit otherwise hidden or only partially glimpsed patterns, which can then inform humanistic judgment and interpretation” (2). Importantly, they foreground the limitations of this kind of study, highlighting the probabilistic nature of their findings and the tendency of data sets curated by humanities scholars to reflect and reinscribe existing disciplinary assumptions. Their task, they posit, is not “to establish new certainties,” but instead “to present principled generalisations” based on studies that would not be possible without the aid of a computer (3); as they put it elsewhere in the book, they aim to provide “challenges to interpretation” rather than ground truths (80). The introduction is followed by a chapter on methods that details the authors’ processes of text selection and preparation and explains their chosen computational techniques—Principal Components Analysis, Random Forests, Delta, and Shannon Entropy—as well as the common measures of statistical significance they use throughout.

The rest of the book presents a series of experiments, the best of which approach questions of long-standing critical interest from new angles. Chapter 2 challenges scholarly attempts to detect a general principle guiding playwrights’ selection of prose and verse, drawing attention instead to the strategic ways writers deployed the contrast between the two. Chapter 3 investigates the speech patterns of recurring characters and of dramatic types; unsurprisingly, the authors find that dialogue is shaped more strongly by characters’ dramatic function than by any sense of, say, Falstaff as a coherent individual across plays. Chapter 4 examines the use of stage properties and finds that prop selection reflects genre more strongly than the influence of factors like author or performance date. Chapter 5 argues that changes in dramatic language occur gradually over time and are not attributable to particular writers or even to the emergence of new generations of writers. Chapter 6 subjects to quantitative scrutiny the emergent repertory approach and rejects the posited influence of playing company on dramatic style. Finally, chapter 7 argues that comedy after 1660 was more continuous with the pre–Civil War tradition than critical narratives of rupture would lead us to believe.

The title's promise to go “beyond authorship” foregrounds tellingly the very question it claims to bracket. The use of techniques from authorial attribution—such as the frequent recourse to analysis of usage patterns for very common words like the and and—means that many of the book's experiments must be designed around or against the methods’ predisposition to detect the author signal. A definition of style that is known to be sensitive to authorship, that is, may not be the best tool for detecting the influence of playing companies on dramatic works, and therefore may not be the most sound basis for rejecting the entire repertory approach. On the whole, however, this impressive project addresses questions of ongoing scholarly inquiry from fresh perspectives and provides new models of corpus curation and experimental design that are likely to inspire future quantitative literary study. The volume's seventy-one figures, sixteen tables, and five appendixes make it possible not only to understand but also to reproduce the authors’ experiments, reflecting a commitment to methodological transparency that should serve as an example for practitioners of digital humanities across disciplines and time periods.