Date of Degree


Document Type


Degree Name





Sondra Perl

Subject Categories

Higher Education Administration | Rhetoric


composition; data visualization; disciplinarity; dissertations; distant reading; topic modeling


Combining qualitative coding with original algorithmic and quantitative analyses, this project aggregates and visualizes metadata from 2,711 recent doctoral dissertations in Composition/Rhetoric, completed between 2001 and 2010 (inclusive), in order to establish an empirical baseline of what new and established scholars in Composition/Rhetoric agree upon as acceptable research in the field. I find that both subject matter and methodologies largely collocate within a small number of clusters, but not without cross-over among these clusters, and I call for increased dialogue among schools focusing on these different methods and subjects.

Chapter 1, 'Disciplinary Anxiety and the Composition of Composition,' reviews the history of Composition/Rhetoric's search for a shared research paradigm, including its potential rejection of that goal. Following Derek Mueller (2009), I argue for 'distant reading' (Moretti), through metadata visualization, as a means of keeping abreast of research trends that would be unmanageable through direct reading alone.

Chapter 2, 'From Dissertations to Data: My Exhibits and My Methods,' explains how I obtained, selected, and prepared the 2,711 documents that go into my subsequent analysis.

Chapter 3, 'Mapping the Methods of Composition/Rhetoric Dissertations: A 'Landscape Plotted and Pieced,' ' takes up the question of whether the field has divided along methodological lines, as Stephen North (1987) predicted. After identifying methods used in dissertations based on their abstracts, I describe correlations between dissertation methods and the graduate schools where they are most frequently employed. Most dissertations used more than one method. I demonstrate that, while aggregable and empirical methods have not disappeared, few schools focus on them; dialectical and text-hermeneutic methods are far more common across the board.

Chapter 4, 'Tapping the Topics: What We Study When We Study Writing in Writing Studies,' turns from methods to content. Drawing on a computer-generated topic model of the full text of 1,754 dissertations, I provide evidence both for high-level clustering of topics and for large numbers of dissertations that cut across these clusters. The most common dissertation topics in this sample address the teaching of writing and, in a largely separate cluster, theories of meaning-making.

In Chapter 5, 'Toward a View From Everywhere: 'Disciplined Interdisciplinarity' and Distant Reading,' I reflect on the benefits and limitations of the methods I have used, and suggest directions for future study.

Although it is generally clear to doctoral students preparing to begin dissertation work that they have a number of methods to choose from, and a number of ways to construct and usefully constrain their subject matter, Composition/Rhetoric as a field has not generally speaking kept good track of trends across institutions, with the result that individual dissertation-writers do not know whether a particular method or subject they are considering is common or quirky, cutting-edge or passé. By offering a recent, zoomed-out view beyond the vantage point of any one program, these analyses provide a shared map of where Composition/Rhetoric doctoral research has been, so that researchers, thesis committees, and curriculum-planners can make more informed local decisions about where their research should go next.