Mina (2019) proposes that a critical theory of technology in writing instruction must supply a framework within which users work and function, demonstrating that technology is always value-laden as well as context-bound. These ideas are supported by my research, which theorizes the critical use of blogwork in first-year composition. Because my students write on blogs, individual student writing samples as well as larger semester-length corpora can be examined. Blogs become small-world networks for students as they read and write to each other, beginning with the course blogroll, which functions as a means of unifying a class.
Students investigate the rhetorical practices of their peers and of themselves by employing various distant reading strategies among the Voyant suite of tools, as certain features of their blogwork offer significant data about their rhetorical choices. Using quantitative means, students learn to inspect their writing to interpret its features and to analyze how their writing has changed. As writing is a form of structured data, learning about its structure via distant reading can be meaningful to student-writers in that the activity removes the social context. Text mining involves the use of computational techniques to derive information from written texts, and a collaborative investigation can encourage social knowing. Social cohesion is hard won in the first-year classroom, as students have limited experience with networks of disciplinary knowledge and limited understanding of writing practices. Their perspective of first-year writing is further skewed by whether or not they self-identify as writers. In the awkward social space of the first-year writing classroom, establishing a fair space for collaboration and influence is challenging but made possible via the class blogwork, which networks students and their writing with surprising equanimity.
The transformative part of this practice is that students prepare the work (i.e. the writing on their blogs) that becomes the object or problem of study. Students can accomplish limited primary research in a semester, so there is a real benefit to having the proposed research in the room and not outside. When students become participant-observers, revisiting from a distance what they have written and what their peers have written, they learn even more. Distant reading enables such analysis by making it simultaneously impersonal and personal. It is a valuable and non-judgmental tool, revealing the patterns with which writers write.
Franco Moretti (2000) theorizes that much can be learned about large corpora of literature by applying distant reading strategies that do not study particular texts but instead aggregate and analyze particular points of data. To date, distant reading is usually accorded to prestige texts: literature (Moretti 2000); academic journals (Palmieri and McCorkle); and rhetoric and composition dissertations (Miller). I propose that a distant reading of student writing – that is the computational consideration of particular aspects of student writing by students– gives students license and authority to opine about what makes for effective writing. It is not practical to have students read and respond to all the writing produced for a class. Distant reading offers students a bird’s-eye view to experience the breadth of the class’ writing. Because student writing on blogs becomes data that can be easily aggregated, distant reading promises that students will be able to access or read in some way student writing for a course
Following are a few ways that students can deploy Voyant to examine student blogwork. It is useful to begin with the default skin to understand the programmers’ purposes and to adapt those to the student investigators’ purposes. At the end of this post, I provide links to Voyant activities and assignments.
After a corpus of student texts is uploaded into Voyant, a user will see its text revealed in five programs in a default skin. (Note: All of the programs in the Voyant’s default skin can be substituted for others.) The Summary program (lower left) initiates students to thinking about texts quantitively. For instance, some students may not have considered that sentence length matters, that it can be a stylistic choice. This average tells them something about their own habits and preferred patterns as well as the class’ and suggests to them that sentence length is meaningful data. The next tabulation, the Readability Index, teases students’ imagination by demonstrating that some features of language are more readily accessible to reader. The category Vocabulary Density challenges writers to think about lexical variety. Vocabulary Density is a measurement of the vocabulary usage relative to the length of the document. Dickerson (2018) explains it can be likened to tabulating “how many words will be read on average before a new word is encountered.” This calculation invites students to consider the efficacy of any word or phrase.
Instead of analyzing how experts write, novice writers contemplate how they write and how effective that writing is. This is a novel conversation, placing student writers in the company of established writers, and it can be a validating enterprise. By setting their writing against itself – early in the term and later – students can observe whether or not their writing has changed and subsequently offer hypotheses about the perceived change. It may also affirm what they find particularly successful about their rhetorical moves.
Mina proposes that one must work to demonstrate the values of a technology and that each technology is constrained by a specific context. The Voyant suite of tools were designed for scholarly work with prestige texts and not for the pedagogical research I am describing. For instance, the upper left default program in Voyant is a word cloud demonstrating word frequency, a familiar visualization for students. This visualization is not useful for student research, as students readily note how repetitive their writing can be. However, a critical theory of technology requires that users interrogate this visualization by exploring what words are displayed and what are not. Users can access the list of stop words under the options tab. A review of what words are not tabulated in the default set up has interesting ramifications. A productive class activity is to have students review the stop words and suggest reasons why this list may or may not offer interesting data points for analysis of student writing. This actiity can help students unpack the logic and objectives of the program to determine if their own inquiry is served
Subsequently, students can reinstate stop words that when added back might yield interesting results under investigation. Students might opt to include coordinating conjunctions (e.g. and, or, but); demonstrative pronouns; personal pronouns; or stance inflections (i.e. boosting or hedging). For instance, once the conjunction “and” is reinstated as a search term, students can track its use across any of the tools displayed, including how it trends over the corpus (upper right) or its links and connections (lower right) across the corpus. Students can determine how individual students or how a class relies on “and” as filler text versus to offer significant supplemental information. By including personal and demonstrative pronouns, students can visualize trends in the class’ use of vague, indeterminate pronouns versus concrete nouns.
Voyant makes it possible for undergraduates to do this research on a class or micro level and on their own writing. These valuable inquiries into student writing happen in the context of a course and contribute to students’ understanding of the ways that language works.
Links to sample Voyant demonstrations, activities, and assignments:
- Voyant Demonstration and In-Class Work: First-year writing course blog: “What do you know about writing?” Spring 2021
- Professional Business Writing Assignment: Course blog: “Getting Ready: Workplace Writing” Fall 2019
- Aull, Laura. “Generality and Certainty in Undergraduate Writing over Time: A Corpus Study of Epistemic Stance Across Levels, Disciplines, and Genres” in Developing Writers in Higher Education: A Longitudinal Study. Anne Ruggles Gere, ed. U Michigan P, 2019.
- Brooke, Robert. “Underlife and Writing Instruction,” College Composition and Communication, (May, 1987), Vol. 38, No. 2: 141-153.
- Dickerson, Madelynn. “A Gentle Introduction to Text Analysis with Voyant Tools.” 2018 https://escholarship.org/uc/item/6jz712sf Accessed Feb 10 2022.
- Lay, Ethna. “What do you know about writing?” Course blog, spring 2021. https://spring2021wsc001.wordpress.com/2021/03/09/inclass-work-of-tues-3-9/Accessed Feb 10 2022.
- Lay, Ethna. “Getting Ready: Workplace Writing ” Course blog, spring 2019. https://gettingreadyworkplacewriting.wordpress.com/2019/11/14/about-project-4-an-introduction/ Accessed Feb 10 2022.
- Miller, Benjamin. “Metadata Heatmaps for Distant Reading” https://cmu-lib.github.io/dhlg/project-videos/miller/Accessed Feb. 10 2022
- Miller, Benjamin. “The making of knowledge-makers in composition: A distant reading of dissertations.” Disssertation. Graduate Center, City University of New York. 2015.
- Mina, Lilian W. “Analyzing and Theorizing Writing Teachers’ Approaches to Using New Media Technologies.” Computers and Composition Volume 52, June 2019: 1-16.
- Moretti, Franco. “Distant Reading.” Digital Scholarship in the Humanities (March 2014) 30(1): 152-154.
- Moretti, Franco. Graphs, Maps, Trees.London and New York: Verso, 2005.
- Mueller, Derek. “Digital Underlife in the Networked Writing Classroom,” Computers and Composition26.4 (Dec. 2009): 240-250.
- Mueller, Derek; Jennifer Clary-Lemon and Kate Pantelides. Try This: Research Methods for Writers. Practices & Possibilities Ser. Fort Collins, CO: CSU Open P, 2022.
- Palmeri, Jason, & McCorkle, Ben. “A distant view of English Journal, 1912-2012.” Kairos: A Journal of Rhetoric, Technology, and Pedagogy (2017) 22(2). http://kairos.technorhetoric.net/22.2/topoi/palmeri-mccorkle/methodology.html Accessed Feb 10 2022.
- Sinclair, Stéfan and Geoffrey Rockwell, 2016. Voyant Tools. Web. http://voyant-tools.org/.
- Sinclair, Stéfan and Geoffrey Rockwell, 2016. [About] page | Voyant https://voyant-tools.org/docs/#!/guide/about-section-license
- Underwood, Ted. “Distant Reading and Recent Intellectual History” in Debates in the Digital Humanities. Eds. Matthew K. Gold and Lauren F. Klein University of Minnesota Press, 2016.