Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).
Location:Salons 4 & 5, Grand Ballroom, Marriott City Center
Reading Against Models: Approaches to Algorithmic Criticism with Poetry
Lisa Marie Rhody
The Graduate Center, CUNY, United States of America
Our reading ecologies-- ecosystems of searching, aggregating, selecting, recommending, sharing, and interpreting texts--are increasingly governed by computation; nevertheless, stories play a pivotal role in how computer and data scientists develop computational methods. When data scientists talk about the need to "tell a story" with their data, they demonstrate the power of narrative to explicate spatial and quantitative reasoning. This talk considers the role of the literary scholar in reading ecologies continually transforming in the face of machine learning. It takes as its point of departure the long-standing history of remediation between language and images and extends a humanistic tradition to the treatment of specific instances of computational modeling with poetry. Through an exploration of stories used to explicate the formalized, procedural rhetoric of algorithms that underlie machine learning, this talk proposes a heuristic for reading against computational models and argues that such skills are increasingly necessary for humanities scholars.
I came to topic modeling not because I thought it could be prescriptive or replace human interpretation but because it efficiently organizes words in a corpus according to the likelihood that they share vocabularies. In my study of “ekphrasis” (poetry about the visual arts), I wanted to question the assumption that the genre “turn[s] on the gendered antagonism” between poets and painters--a dynamic signaled by the ubiquitous use of words such as silent, mute, and still in the genre. If topic modeling could identify clusters of ekphrastic language over thousands of texts, it may help to locate more examples of the genre by women. Conversely “ekphrastic” poems that did not appear in the same topics could provide clues to alternative approaches to the genre. This paper demonstrates how “communities” of topics may point to alternative genre conventions in ekphrasis, while at the same time demonstrating a new heuristic for reading against models that can extend productively to other areas in literary studies.
Drawing connections to recent publications that explore the potential cultural biases and violence of algorithms by scholars such as Cathy O’Neil, Virginia Eubanks, Meredith Brussard, Safiya Umoja Noble, Ed Finn, and Hannah Fry, this talk will explore topic modeling as an activation of what Heidegger calls the hermeneutic circle and demonstrating how the complementary and interrelated activities of human and computer as readers can expose implicit bias in both literary and algorithmic models.
Intellectual History from Below? Applying Corpus Linguistic Tools to the 19th Century British Periodical Press
Université du Québec à Montréal, Canada
The history of political thought tends to focus on “great” authors and certain key concepts, thereby analysing over and over a small handful of texts. However, with the recent digitization of numerous 19th century corpora, other possibilities emerge. The accessibility of those sources enables researchers to grasp the more popular and common political languages used by a great variety of actors. The question then becomes, how to study those corpora?
Drawing on an on-going thesis on the history of the word “democracy” in Britain, this presentation will argue that corpus linguistics offer crucial insights on how to handle and analyse such sources. More precisely, by bringing together corpus linguistics and the history of political thought, one is able to offer an “history from below” that differs from the interpretations obtained through classical methods.
As an example, an analysis of the uses of the word “democracy” in the British periodical press with corpus linguistics tools (AntConc, #LancsBox, Meaning Fluctuation Analysis (MFA)) will be presented. This exercise underlines several elements, such as the polarized uses of the term, the influence of the American example, as well as the disconnection between “democracy” and “people”. For most of the period, “democracy” was mainly used by conservatives to signify the rule of the “populace”, and it is only with the acceptation of the United-States as a stable and legitimate power that the term became synonymous with popular government. Such results both confirm previous historical insights as well as challenging mainstream interpretation of the rise of “democracy” in Britain.
“I Am Not Enjoying This Book”: Oppositional Auto/biographical Writing in Online Infinite Jest Reading Group Blogs
University of Waterloo, Canada
This paper considers the auto/biographical writing of online reading groups dedicated to David Foster Wallace’s Infinite Jest. For this paper, I draw from my close reading of the most successful online Infinite Jest reading group, 2009’s Infinite Summer, as well as my own experiences moderating an online Infinite Jest reading group which I ran in 2016 as part of my Doctoral research. This paper focuses on the auto/biographical writing within these groups that critically intervene into public conversations surrounding Wallace that ignore or suppress critiques of Wallace’s work while praising his “genius” and justifying his canonization. The critiques within the blog posts and comments of these online reading groups are consonant with recent expressions of discontent with Wallace in blogs, Twitter, online journalism, academic discussions, and independent bookstores. These auto/biographical disclosures of “dislike” are “social and political act[s]” (Fuller and Sedo 41) that communicate private reactions to Wallace’s work or experiences with Wallace’s fans/scholars through a public platform in order to expose the uncritical adoption of Wallace into the American literary canon.
Drawing from auto/biography studies and Mass Reading Event studies, I identify two modes of auto/biographical writing: affirmational and oppositional. Affirmational autobiographical writing identifies with Infinite Jest, containing discussions primarily about mental health and addiction. Oppositional autobiographical writing challenges and/or is unable to identify with Infinite Jest, containing personal disclosures about disliking the novel and the novel’s poor representations of race and gender. While affirmational auto/biographies contribute to the public discussions of mental illness and addiction, they also have a tendency to be uncritical of Wallace’s work and dismiss any critiques. My paper traces the “compositional strategies” (Morrison) of oppositional auto/biographies that negotiate the power structures of “liking” or “admiring” the book. These strategies, I argue, maintaine a space for further articulation of critique and reminding fellow participants about the lives of other participants who are not represented within the novel’s narrative that primarily focuses on white, cishetero male characters.