Novels in the News: The Reprinting of Fiction in Nineteenth-Century Newspapers
Northeastern University, United States of America
The Viral Texts project (http://viraltexts.org) addresses the practice of cutting and pasting in the nineteenth-century press broadly, uncovering a range of widely-circulated news, “information literature,” poetry, and miscellany that constituted the bulk of newspaper content during the nineteenth century. However, our broad lens has obscured the circulation of fiction, though the newspaper was an essential vehicle this genre. Whether through the publication of short stories, serialized novels, or excerpts, or through the extended quotations that appeared in reviews, the newspaper was many nineteenth-century readers’ primary fiction medium. This paper describes a new Viral Texts experiment using existing collections of nineteenth-century fiction as seed corpora, enabling us to identify only newspaper reprints drawn from the genre. This paper outlines initial experiments using the Wright American Fiction archive (http://webapp1.dlib.indiana.edu/TEIgeneral/welcome.do?brand=wright) of nearly 3,000 American novels and story collections published 1851-1875 (also the period of greatest strength in our collected newspaper corpora) to identify the stories or novel chapters that circulated most widely in newspapers, and to analyze how these viral fictions remap the intersections of literary culture and mass media in the period.
The approach we outline here accounts for what fiction suffused everyday reading in the period as well as how those texts were used. Sometimes entire stories or novels would be reprinted in the press, but more often particular chapters or excerpts circulated unmoored from their source. Newspapers sometimes embedded segments from fiction within other kinds of texts (e.g. sermons, speeches) or to emphasize particular editorial points. Such uses are particularly salient from a literary-historical perspective, as they offer insight into which characters, scenes, or themes most resonated with editors and readers, or what kinds of cultural narratives emerged around texts: sometimes aligned with their canonized perception but sometimes not. For instance, preliminary experiments indicate that while chapters twelve, thirteen, and sixteen of Harriet Beecher Stowe’s novel, Uncle Tom’s Cabin, were reprinted thirty to thirty-one times, other chapters might only be reprinted once or twice. The most reprinted chapters of Uncle Tom’s Cabin include what would become the novel’s most iconic scenes—George and Eliza Harris’ armed stand against slave catchers, for instance, or Uncle Tom’s horrific beating—but also shorter scenes of slavery’s horrors, such as a child taken from his mother at auction. A close examination of these and similarly “viral” chapters illuminates the popular interests of the general public—both the elements of fiction they found compelling and the issues they used fiction to address. Analyzing canonical texts also points us toward similar patterns that emerge in non-canonical texts—Female Life Among the Mormons, for example, or A Crown from the Spear—which early experiments show were reprinted far more often than more familiar literature. To think across our corpus, in closing we will describe topic modeling and vector space analyses of reprinted chapters’ content that helps surface the themes and textual features that drove readerly interest in popular newspaper fiction.
The Odors of Mediation: A Case Study in 3D Printing
University of Maryland, College Park, United States of America
Digital media studies has long relied on techniques of vision, from its roots in cinema studies to contemporary strategies of data visualization. However, the field's recent orientation toward big data, infrastructural systems, and planetary ecology have challenged this focus on vision, as scholars grapple with the project of rendering such large scales within the human perceptual field. How might digital media studies tuned to an infrastructural or environmental key benefit from expanding its sensory toolkit, particular by attending to the "lowest" or most "chemical" sense: smell? Smell is massively understudied and under-theorized in the media studies tradition, even as it's unavoidably present in media writ large, from the fetishized smell of old books to the ozone outgas of an overheated motherboard. In this paper, I sketch a methodology of smell for media studies and apply it to a single digital technology: the 3D printer. In doing so, I ask how smell enriches and refines the analysis of technical objects and digital cultures. 3D printers heat and extrude filament according to digital shapefiles, forming objects ranging from small trinkets to prosthetic limbs. Even more so than its actual functionality, the 3D printer has become one of the more potent cultural symbols of techno-innovation and the "maker" movement writ large. In operation, 3D printers release a range of noxious gases ranging from the pleasant-smelling to the absolutely wretched. I argue that these smells serve a dual function, both as toxicities that require management with ventilation and as material signifiers of a participation in a broader techno-culture. In particular, smell reveals this culture as not solely confined to affectively "cool" Western maker culture, but the racialized manufacturing practices displaced into the developing world now returning to the West in miniature through the 3D printer. Attending to smell then connects material specificity to affective experience, in turn becoming a useful sense for engaging questions of pollution and toxicity that dog research into media manufacturing.
Mapping an Archipelago of Influence
Rutgers University, United States of America
Archivepelago is a project visualizing the transmission and translation of notions of sexuality and gender by mapping networks of queer writers and artists (and early sexologists), bringing into relief the communities that developed through these networks. The project draws upon finding aids and biographic data, charting connections between these figures ranging from their correspondence to the works dedicated to and translated by one another. The project is intended to act as a resource for the public to understand the forces underpinning queer diaspora while encouraging scholars to rethink our conceptions of artistic influence beyond the misogynistic, heteronormative notions presented 35 years ago in Harold Bloom's The Anxiety of Influence.
In addition to drawing upon scholarly work on archipelagoes (Manuel Guzman, Michelle Stephens, Island Studies Journal), this project draws upon Digital Humanities projects visualizing and thinking about archived correspondence such as the Republic of Letters and Olive Schreiner Letters Online as well as conversations about recovering and amplifying history via Digital Humanities projects. This presentation will serve as an overview of the project and reflect upon its current phase (development of a graph database and prototyping) while inviting audience discussion regarding methods, theory, and how scholars can better understand and model queer literary influence.
Ultimately, Archivepelago’s main features would be 1) interactive network maps demonstrating relationships in terms of correspondence, translations of one another’s works, and shared demographic or psychographic characteristics such as religious affiliations; 2) interactive geographic maps depicting the migration of individual writers and artists, demonstrating emerging communities in metropolitan centers such as Paris, NYC, Algiers, and Mexico City; and 3) an online exhibit outlining key concepts drawing on well-known moments and figures (e.g., Langston Hughes, Gertrude Stein).
From Document to Data: Prosopography and Topography in the Tax Rolls of Medieval Paris
Johns Hopkins University, United States of America
This poster presents on the experimental stages of a project to create a digital edition of the tax rolls of medieval Paris. Levied on the city by King Philip IV between 1292-1300, and again in 1313, these documents provide a wealth of information about the individual people, demographics, and topography of the city. However, despite their importance, they have never been systematically published or digitized, and all analysis between them must currently be done by hand. This project, inspired by the Henry III Fine Rolls Project (finerollshenry3.org.uk) and the Map of Early Modern London (mapoflondon.uvic.ca), proposes to render the seven extant rolls in TEI/XML, where personal, occupational, topographical, and financial data are dynamically brought together and opened up for searching, cross-referencing, mapping, tabulating, and exporting. Although the entries in the rolls themselves seem simple at first glance—organized by parish and consisting at most of a name, occupation, and amount of tax owed—actually encoding them prompts a variety of questions for the inexperienced editor. Many of these relate to the conventions of digital editions: whether the edition should reflect the layout of the manuscript; how to handle corrections, errors, and abbreviations; and how to approach uncertainty—especially when identifying individuals between multiple tax rolls. Others relate to underlying technologies and frameworks: including detailed relationships (familial, occupational, geographical) is desirable and adds interoperability within the semantic web, but also means working with complex ontologies and data models. This poster explores these questions and others, and engages with the early concerns and pitfalls of creating a digital edition.
Reimagining Romance: "The Legend of Korra" Critical Fandom Practices
Northeastern University, United States of America
Current fan studies research merges sociopolitical arguments and rhetoric studies, complicating how fans engage with texts and the values of these engagements. Scholars have articulated what “critical fandom” practices (Lothian) look like in terms of identity and representation (Ebony Elizabeth Thomas & Amy Stornaiuolo; Paul Booth; andre carrington). Texts like Fifty Shades of Grey, originally a Twilight fanfiction, demonstrate fans’ desires to explore diverse types of intimacies and relationships; the published Fifty Shades of Grey novels, as other fanfictions have done, reinscribes heteronormative notions of intimacy, sexuality, and gender. This presentation examines how fandom literacy practices can resist harmful hegemonic narratives around race, sexuality, and gender.
Online fan communities such as Archive of Our Own (AO3)–a fanfiction database–prioritize fan interpretations of a text over the text itself, which makes these spaces important to study and celebrate when thinking about critical fandom practices. I focus on one fandom in particular: The Legend of Korra (TLoK), a popular Nickelodeon children/young adult television show. What makes TLoK a crucial cultural text is that the main character, Korra, is a bisexual woman of color, which already resists mainstream narratives around race, sexuality, and gender. I have collected and scraped over 8,000 TLoK fanfictions published on AO3 with approval from the Organization for Transformative Works. Using this corpus, I will examine fan representations of intimacy, sexuality, and gender by cross referencing computational text analysis with the “Ratings” and “Relationship” metadata. For example, how is intimacy portrayed in General Audience as opposed to Mature? In TLoK fan community, the average General Audience fanfic word count is around 3000, while the average Mature fanfic word count is around 20,000; this discrepancy suggests that there are generic patterns that fanfiction authors follow. By pairing these Ratings with the Relationship tags chosen for the texts, I will also examine what patterns occur in Mature texts that center around particular relationships and identities; for example, how is the relationship between Korra/Asami (two women) portrayed as opposed to Korra/Mako (a woman and man)? By doing this work, I hope to create a methodology that centralizes fan literacy practices in order to demonstrate how critical fandoms reimagine intimacy and romance.
What’s in a Face? Examining Historical Trends Through the Faces of a Mass Media Publication
1SUNY Polytechnic Institute, United States of America; 2Chapman University, United States of America
The role of popular print media in broader social, political, and cultural landscapes has been widely studied in the field of digital humanities. Many of the existing, large-scale studies, however, have focused on text analysis rather than image analysis. This paper describes an interdisciplinary project that analyses images of human faces from the complete Time magazine archive. Our project combines quantitative and qualitative methods to study the images, as well as design methods to make our results broadly accessible. Our research team includes two faculty members, a postdoctoral researcher, and undergraduate and graduate students.
The first section of the paper provides an overview of our project: our goals, a description of our methodology, and results to date. In the second section, we discuss our rationale for focusing on Time magazine: it has been a mainstay publication for many decades, and has a well-documented corporate history. While it is widely held in libraries across the US these holdings are primarily in the form of microfiche, which limits public accessibility. The final section outlines the unique aspects of using human faces as a conduit into the Time archive as well as into the broader social-historical context. For example, exploring changes in representations of gendered or racialized faces can capture both Time’s coverage and how it may have been a driver of events.
We attempt to shed new light on how and why to study this kind of archive, and what such research can offer both academics and the general public in terms of understanding the role of a publication like Time magazine in broader socio-political contexts.
Designing a Community Based Digital Archives Project and FYE Course Module
University of The Bahamas, Bahamas, The
I plan to present the early planning stages of a developing a community-based digital archival project parts of which are embedded in a first-year required composition course. In each semester of the Writing and Rhetoric course, research projects were assigned that aligned with themes being targeted for possible development of digital archives: Junkanoo and sites of cultural memory. The community digital archives project team includes an English professor (PI), Librarian (Co-PI), Assistant Librarian, Technology support staff (2), and a local practitioner (member of the Junkanoo group) who is also a professor of Chemistry. Community stakeholders have been identified and met with including the owner of a defunct Junkanoo museum (and its contents), the leader of a Junkanoo group and one of the originators of the festival on the island, and the chair of the local Junkanoo Committee. The local newspaper (which does not have digital archives) has also been identified as a possible partner. Community members and leaders of Junkanoo have expressed an interest and desire to preserve the history of the Junkanoo festival on the island and to help educate the youth and general community about its value as a cultural tradition given recent decline in interest and participation among the younger generation. Fall 2018 was the first iteration of student research projects to determine the interest and viability of developing archives in these thematic areas and identify possible collaborators. Students were required to use primary sources held in the university library special collections, the public library, the local newspaper (paper) archives, or conduct interviews with relevant persons in the community. These research papers were then used as sources for students to publish blog entries on Wordpress websites for each topic/theme. Models, resources and input for proceeding with this project are sought as members of the project team are new to digital humanities approaches. In addition, although both the PI and Co-PI are Caribbeanists, both are foreigners, which presents additional challenges to a community-based project, and planning for sustainability.
The Digital Studies 101 Website: Developing and Using An Un-Textbook
1Georgetown University, United States of America; 2University of Mary Washington, United States of America
When designing a course, one early decision instructors make is what text or texts to assign. Whether textbooks, individual essays, videos, or other materials, assigned texts provide students with a content introduction before class discussion, present an additional perspective on a topic, and offer sources for later reference, among other potential benefits. However, textbook come with other limitations and deleterious effects, including cost (Moxley; Munson) and ideologies (Welch; Johnson-Eilola). In an era of rising tuition, fees, and student debt, educators (Smith & Casserly; Morris-Babb & Henderson) and states have argued in favor of open-source textbooks as one means of mitigating higher education financial impacts to students. Long before our own state developed a law addressing textbook costs and encouraging the adoption of low or no-cost open education resources, our department created a common website resource for all Digital Studies (DGST) 101 instructors and students. We found using the digital medium to communicate about digital studies topics not only a cost-savings measure, but an ethical approach to digital education for instructors, students, and the subject matter itself.
Using ethics as our touchstone, our panel discusses the historical and on-going development of our DGST 101 web resource, the benefits and challenges of an institutionally-created resource for the instructor-content developers, the advantages and concerns for students, and discuss how instructors and students use the website within the context of DGST 101 courses. Throughout the presentations, the panelists present a discussion of ethical impacts of balancing academic freedom across multiple instructors, developing sustainable content attuned to a rapidly changing digital context while negotiating issues such as supporting commercial, ad-based resources, long-term maintenance questions, issues of digital permanence/digital ephemera, and web labor issues.
Speaker 1 addresses the history of, intent, and development of the website. The speaker describes how the website matches the programmatic ethos.
Speaker 2 addresses the development of resources for the website and an instructor perspective on using the website.
Speaker 3 describes student uses of the website and the ethics behind those uses surrounding student choice, digital accessibility, and opportunities for discussion about fair use. While the term “academic freedom” generally appears in discussion of faculty research inquiry and instructional content development and presentation, we find the Digital Studies 101 website reconfigures academic freedom for students. Rather than students all focusing on the same materials, students choose from a curated list of web materials to develop their own expertise and to develop their own projects.
As Digital Humanities begins to more critical interrogate DH pedagogy, we hope that this model can provide an inspiration and model to use in a variety of formats and digital approaches.
Minority Representation in the Foreign Language Classroom: Teaching Languages Through Digital Engagement
Huson TiIllotson University, United States of America
In the texts used for Spanish Basic Language programs, cultural conversations often reinforce negative stereotypes and/or further marginalize minority groups. These texts ostracize students while also giving a skewed, Eurocentric/mestizo, inaccurate and monolithic view of the Spanish-speaking world. In my classes, I have changed this by employing monolingual voices of protest and change into the classroom. We studied Costa Rica, for example, but through the lens of the Afro-Costa Ricans fighting to be recognized on their census and exploring their daily reality in their native country.
My proposed 15 minute digital presentation would discuss methods for foreign language instructors to effectively use primary sources in their classes to build interactive digital databases their students can use to engage the culture of the target language through the minority lenses most textbooks lack. This is imperative, as students then can recognize the similarities f the struggles and landscape of all the Americas.
Creating a Spatial History of London’s Public Drinking Fountains
Rice University, United States of America
As Emma Jones notes, “Everyday demands for water in the city have greatly shaped the design, use and experience of our built environment” (78). In 1859, temperance advocates and social reformers began constructing public drinking fountains across London, hoping to offer the poor unpolluted water and a free alternative to alcohol. Soon they also built cattle, dog and sheep troughs to address animal welfare needs. Drawing from the detailed records of the Metropolitan Drinking Fountains and Cattle Trough Association (MDFCTA), I analyze how a private charity developed an important part of the public health infrastructure in Victorian London. By mapping the locations and attributes of these fountains and troughs, and by combining that information with other data sources such as economic and health records, we can examine what factors influenced their placement, including population size, level of poverty, rates of disease, and proximity to churches and parks. The location of fountains and troughs seems to reflect the MDFCTA’s desire to reach large numbers of people and animals, its pragmatic decisions in negotiating with government officials, the influence of donors, and its associations with the temperance, parks, and animal welfare movements. This project demonstrates how researchers can employ Geographical Information Systems (GIS) and data visualization along with archival research to discern patterns in the development and decay of urban infrastructure.
This work is supported by the Rice University Humanities Research Center’s Spatial Humanities Initiative.
Jones, Emma M. Parched City. John Hunt Publishing, 2013.
Viaggio alla ricerca della conoscenza: Dalla faccetta R alle faccette R4 dei FAIR PRINCIPLES per l'identificazione del Digital Cultural Heritage
University of Bari Aldo Moro, Italy
L’identificazione del Digital Cultural Heritage, riconosciuto ufficialmente dall’UE nel 2014, passa necessariamente dalla risoluzione delle rilevanti criticità relative a conservazione, stabilità, sostenibilità, fruibilità e riusabilità delle risorse digitali nello spazio e nel tempo.
Diventa urgente e indispensabile evolvere l’approccio corrente al digitale, oggi ancora inteso esclusivamente quale mediatore di valorizzazione di patrimoni culturali analogici, verso una sua ridefinizione quale facies culturale identitaria della contemporaneità.
Allo scopo, una questione di primo livello da affrontare con urgenza è l'affidabilità delle risorse digitali e, in particolare, dei metadati.
Quasi tutte le collezioni digitali oggi fruibili, siano esse digital libraries, o archivi digitali, o audio-video, sono condizionate pesantemente dall’inaffidabilità, essendo incoerenti, non interoperabili, non preservabili proprio perché le risorse sono state generate con scarsa attenzione ai metadati e ai loro contenuti descrittivi.
Per rendere più affidabili i dati digitali, proponiamo che la R: Re-usable dei FAIR Principles sia ridefinita quadruplicandola in R4: Re-usable, Relevant, Reliable and Resistant. Questi requisiti, infatti, conferirebbero ai dati digitali il valore di Cultural Heritage, in quanto li renderebbero sostenibili e permanenti
L’approccio metodologico sopra delineato è stato applicato sperimentalmente nel progetto di digitalizzazione dell’Archivio storico della casa editrice G. Laterza & Figli, in pubblicazione nella Puglia Digital Library della Regione Puglia. I risultati mostrano che la distinzione tra computational artifacts culturali e prodotti digitali “di consumo” risiede nei metadati descrittivi, e in particolare nella correttezza delle proporzioni tra:
- configurazione quantitativa: è il rapporto bilanciato tra scelta e quantità di elementi e attributi della struttura del metadato ed esaustività di informazioni/conoscenze sulla risorsa e sul suo ciclo di vita da fornire nelle descrizioni;
- configurazione qualitativa: è la scelta equilibrata del livello informativo/cognitivo da conferire a ciascuna descrizione e all'insieme delle descrizioni che rappresentano la risorsa e il suo ciclo di vita, mediata in relazione alle possibili variabili delle esigenze di conoscenza degli utenti sia contemporanei che futuri.
Lo schema di metadati è stato co-creato avendo quale riferimento lo standard METS-SAN, integrato con metadati presi da altri standard basati su ontologie e linguaggi sia semantici che concettuali a completare la struttura definitiva.
Topics and Terms: Differential Vocabularies in Composition/Rhetoric Doctoral Dissertations
University of Pittsburgh, United States of America
In this poster presentation, I identify vocabulary used to discuss related subjects in several thousand doctoral dissertations in Rhetoric, Composition and Writing Studies (RCWS) – a field whose hybrid and shifting names illustrate the risks of assuming easy transparency of terms. Drawing on LDA topic models of full-text dissertations in the field, followed by similarity clustering among topics, I examine both divergent associations of terms shared across content clusters and synonyms that associate distinctively with different topics within clusters. Consideration will be given to the role of institutional histories and methodologies in contributing to the way these vocabularies are distributed.
This project follows on calls for research into graduate student identity and training in RCWS (Brown et al.) and into more empirical grounding for claims about the nature of the field (Haswell; Phelps and Ackerman), including more grounding in digital research (Johnson).
College Your Way: The Language of Marketing in Contemporary Commerce and Higher Education
Grinnell College, United States of America
This poster comes from a collaborative team: one of us is a Professor of English with an interest in quantitative analysis of textual data; the other is an undergraduate English major and Neuroscience concentrator with work experience in our institution’s offices of Communications and Admissions.
We have compiled a dataset of text from college and university websites in five institutional categories: community colleges, regional colleges and universities, and national colleges and universities. Our investigation explores relationships between institutional type and public-facing marketing language. This approach reveals how concepts such as exclusivity, success, and choice migrate across market sectors. How, we ask, does the great menu of higher education in the United States present its options to prospective students? And moreover, how does the language used by these institutions present their core values?
Some institutions speak of quality; some speak of excellence. Across the board, the websites present a narrow vision of the academic fields doing research. The institutions of two sectors are much more likely than the others to speak of what “you,” the student, will do as an undergraduate, and the schools of one sector were more markedly more likely to speak of diversity. Drawing on these and other cases, our poster presentation will address patterns in the marketing of higher education that have been difficult to perceive from our positions within the academy.
Towards Computational Analysis of Survivor Interviews about Holocaust and Massacre
1Syracuse University, United States of America; 2Concordia University, Canada; 3Nanjing University, China; 4University of Western Ontario, Canada
Eyewitness accounts of survivors of past mass violence are valued for their ability to educate and commemorate, to bring perpetrators to justice or counter denial, for societal reconciliation, and to contribute to social or intergenerational regeneration (Field, 2006). Personal stories are also valued for their moral or ethical force. According to Kay Schaffer and Sidonie Smith, this retelling is “one of the most potent vehicles for advancing human rights claims.” (Schaffer & Smith, 2004, p.1). Often, these stories become known to the public in the form of survivor interviews conducted by historians.
These survivor interviews are also valuable research data. Traditionally, qualitative content analysis and hermeneutics are the main methods in analyzing these interview data, and the focuses are a better understanding of that historical period and discovering the impact of the mass violence on survivors and their families. With the fast development of computer and information technologies, digital environments are now central conduits for the global circulation of these stories, which allows first-person testimony to be increasingly used in human rights research and advocacy. The amount of interview data increases significantly, which may make it impractical to solely rely on traditional qualitative methods to analyze them. In addition, information technologies may be leveraged to foster proposing new research questions that require the computational analysis of the interview data and contribute to the relevant research communities by bringing in new research perspectives and theoretical insights. To illustrate this aspect, we will discuss two ongoing projects in this panel that aim at developing information technologies to afford new research perspectives on survivor interviews about the Rwanda Genocide and the Nanking Massacre.
Rwanda Genocide Survivor Interviews –These interviews were collected for the Montreal Life Stories project (2006-12), for which Prof. Steven High was Principal Investigator. Five hundred Montrealers displaced by large-scale violence were interviewed using the “life story” approach, resulting in more than 2500 hours of video-recorded interviews. We have been developing tools to foster information retrieval with the data, identify interviews of similar topics, and analyze tensions in the interview process. We leverage machine learning, natural language processing, and information visualization techniques in these activities.
Nanking Massacre Survivor Interviews – Survivors of the Nanking Massacre were interviewed multiple times in the last eighty years: 1946, 1986, 2000, and 2017. In this project, we explore new research angles in the analysis, such as the change of interview topics over time to reflect the different historical period and the effects of translation and language analysis tool in understanding the interviews.
In this panel, we discuss these two survivor interview research initiatives and progress, focusing on how the data characteristics and research contexts interact with the choice and development of computational technologies. In other words, the primary goal in this panel is to discuss with the audience on issues related to how to leverage computational techniques in analyzing sensitive interview data like survivor interviews. The secondary goal is to report our progress in the two interview projects seeking feedback from the audience.
Visualizing Uplift: Women of the Early Harlem Renaissance (1900-1922)
Lehigh University, United States of America
“Women of the Early Harlem Renaissance: African American Women Writers, 1900–1922,” aims to digitize and annotate a limited array of primary texts, mainly poetry, and present these materials as a digital archive in the Scalar platform. The project aligns with what Kim Gallon has referred to as a “technology of recovery,” which is one of the core principles bridging African American literary studies and the digital humanities. The project uses Scalar’s visualization and tagging structures to explore stylistic, thematic, and social relationships among a small group of writers, as well as to explore the conversations these writers were having with established writers and editors. Several key themes have begun to emerge as the project has developed. The first of these is the confrontation with American racism, which impacted African American communities intensely in the 1910s; and these poets document the emergence racial violence and the response to that violence. Second, these poets show how the role of African American motherhood was evolving in the early years of the twentieth-century, in part because of the stresses of raising children in a racist society. Third, the black Church became strongly connected to the movement for social justice at this time; and several, if not most, of the writers in this archive explored Christian themes in their anti-racist writing.
For this poster session, I will demonstrate the site I have been developing and give attendees an overview of the site's use of semantic tags and visualizations built around those tags (focusing on the three themes mentioned: racism, motherhood, and the black Church). I will also demonstrate I am developing -- such as geotemporal representations of this community of writers.
Documents to Data: The Evolution of Approaches to a Library Archive
Princeton University, United States of America
In Digital Humanities we speak of moving from “documents to data.” In many projects, this is literal, a process of extracting information or turning text into tokens suitable for computational analysis. For the Shakespeare and Company Project, it entailed a conceptual shift from thinking of archival materials as texts to be encoded and described, to thinking of them as data to be managed in a relational database.
This project is based on the Sylvia Beach papers, held at Princeton University, which document the privately owned lending library in Paris frequented by notable writers of the Lost Generation. Materials include logbooks with membership information and lending cards for a subset of members with addresses and borrowing histories.
This poster will present the history of a multi-year project in three phases, each with benefits, difficulties, and stakes. The evolution of the project demonstrates the development of our thinking as a team as we moved toward a public-facing site designed for a broad audience. In the first phase, we encoded content from the library using TEI/XML, an approach commonly employed for documentary editing. The choice of TEI/XML fit the initial aims of the project, but even rich transcription did not offer the opportunity to fully connect the people, places, and books referenced. Consequently, the second phase was dedicated to designing a custom relational database to model the world of the library by explicitly surfacing different types of connections. The third phase required migrating data from the TEI/XML to the relational database, a lengthy process that exposed inconsistencies in the encoding, but also gave us an opportunity to eliminate redundant and unsynchronized information. The conversion process highlighted the benefits and the difficulties of both systems in pursuing similar research questions. A TEI corpus and a relational database both support querying and making connections, but a database is designed for explicit connections, which makes it easier to identify and group member activities with individual people across multiple data sources. Both approaches require technical expertise, resulting in barriers to non-technical team members working with the data. We found the relational database to be more inclusive for project members: we built and progressively refined a web-based interface that was easier to use than oXygen XML editor, and provided on-the-fly data exports in familiar formats such as CSV. Team members could then do their own analysis (without learning query languages such as SQL or XQuery) and, as a result, had more meaningful engagements with the data.
To illustrate the history of the project, this poster will include sample images of the archival material. It will provide a diagram that maps the transition of the data from its location across multiple XML documents to a relational database. The poster will present examples of the data work enabled by and insights gained since conversion to a relational database. Finally, it will include visuals from the public-facing web application now in development which will eventually provide researchers and the public access to the world of this library.
Models of Influence: Analyzing Choreography, Geography, and Reach Through the Performances of Katherine Dunham
1University of London, Royal Central School of Speech and Drama, United Kingdom; 2The Ohio State University, United States of America
This poster session focuses on the ways in which we model and interpret influence in the project Dunham’s Data: Katherine Dunham and Digital Methods for Dance Historical Inquiry. The overarching project explores the kinds of questions and problems that make the analysis and visualization of data meaningful for dance history, pursued through the case study of 20th century African American choreographer Katherine Dunham. We have thus far manually curated datasets representing 10 years of her performing career from undigitized archival documents, with the goal of representing over 30 years of her touring and travel. At ACH, we focus on models of what we describe as traces of “influence” in and around dance touring, and reflect on the development of scalable digital analytical methods that are shaped by approaches to embodiment from dance, critical race theory, and digital cultures. Modeling influence offers means to elaborate ephemeral practices of cultural transmission in dance, and the dynamic relations of people and places through which Dunham’s diasporic aesthetic developed and circulated, in dance gestures, forms, and practices.
The poster session will address three core challenges in our work to date on representing and better understanding influence around Dunham’s extensive career: 1) how tours build on each other over the years to open up new touring destinations, thus expanding Dunham’s geographic reach; 2) how Dunham’s travels inform the content of her repertoire in terms of movement vocabulary and style; and 3) the impact of performers coming into and out of the company, and the dance information they bring with them and take when they leave. The poster session focuses on the choices made in statistical, spatial, and network analysis to develop historical arguments regarding the patterns and implications of Dunham’s company repertoire and travels. In so doing, this presentation offers an important set of tools for demonstrating a choreographer’s legacy in an ephemeral medium.
Visualizing Citations in Digital Humanities Quarterly's Biblio
Northeastern University, United States of America
In 2017, a team of researchers and developers at Northeastern University secured a Digital Humanities Advancement Grant from the NEH to continue the development of Biblio, a centralized bibliographic resource for Digital Humanities Quarterly, ADHO's international, open-access journal. In addition to streamlining the process of encoding article bibliographies for DHQ’s authors and editorial staff, the project intends to open up DHQ’s archive for citation analysis, via an interface on the journal’s website. As it is currently being developed, this interface will query dynamic citation data from Biblio’s BASEX database via an XQuery API, rendering it in XHTML for in-browser viewing, as well as making the data available for export in formats that include XML and JSON. Moreover, I am working on a part of the interface that will allow users to visualize, using d3, the field landscape of digital humanities as represented (or not) in the journal’s networks of citation.
The proposed poster reports on preliminary citation analysis research I am conducting, as part of the Biblio grant, towards imagining the possibilities for this interface to map the epistemic “geographies” of digital humanities. This research builds on work in digital humanities that attempts to visualize networks of digital humanities research and scholars in the field’s publication record. It also imports methods from a number of fields — including bibliometrics/scientometrics, science and technology studies, and writing and rhetoric — used to cluster and visualize knowledge domains and the borders drawn between them. Most importantly, it draws on the recent work of other digital humanists using data visualization with a feminist orientation, or for the purposes of social justice, in order to imagine possibilities for what these scholarly networks could be, in addition to describing them as they currently are.
The visualizations I will present will compare networks of co-citation in the DHQ corpus to the corpus of citations of DH journals indexed in the Web of Science (WOS). I will draw these networks using bibliometric parsing and network analysis packages in python, detecting communities within them with techniques that include Louvain Modularity clustering and HDBSCAN clustering. I seek to represent, here, some of the traditions of DH scholarship that we narrate in our field’s many historiographies; I further hope to surface potentially absent traditions and attend to the representational shortcomings in the work that DHQ authors cite, which tends to be Western and even North-American centric. The eventual goal of my dissertation project is to show how visualization can be used to represent citation and citation analysis as methods we use to do what Julie Thompson Klein, in her book Interdisciplining Digital Humanities, calls epistemic "boundary work.” The interface we are building will be one means of performing that work, with the goal of transforming the field’s landscape. On behalf of the Biblio team, I will be looking for feedback from conference attendees on what they would like to be able to see and do with the interface we are building.
Between Advocacy, Research, and Praxis: A Critical Reassessment of the Open Access Discourse
University of Maryland, United States of America
Within the field of digital humanities, open access is often considered a virtue that upholds practitioners to strive for making primary resources, scholarly monographs, and other educational materials publicly available. A case for adopting open access is often justified by the fact that research may be publicly funded in the first place, or in an effort to keep the humanities enterprise relevant beyond academia. What is seldom discussed, however, is that open access is a historical concept, bearing particular ideological biases. Kimberly Christen, for instance, critiques the access discourse, regarding the culturally appropriate access protocol she adopted for the content management system Mukurtu. In her “Does Information Really Want to be Free?” (2012), Christen argues how even a seemingly benign concept such as “public domain” has been used to exploit indigenous communities, further perpetuating the legacy of global colonialism. Taken up on the call of Christen to critically investigate the open access discourse, my poster examines how such concepts as “piracy” and “gatekeeping,” too, are not as black-and-white as open access advocates might have one believe. Using the Digital Frost Project as a case study, the poster invites my fellow special collection librarians, editors of electronic editions, digital humanities project coordinators, grant officers, and others, to consider how the otherwise prolific open access discourse may be ill-equipped to foster collaboration necessary for the development of an online platform. In addition to the unattended exhibition, I will deliver multiple 10-minute long walkthroughs of my argument during the poster presentation session.
Limits of the Syuzhet Package and Its Lexicons
Texas A&M University, United States of America
There is always a risk when conducting digital analyses, since programming tools often contain flaws. Therefore, it is important to consider the limits and usages of different programming tools to avoid faulty methods. Sentiment analysis has historically focused on product reviews, such as those of movies, hotels, cars, books, and restaurants, rather than on sentiment in literature. The Syuzhet package, however, is aimed at providing a proper tool for sentiment analysis in literature. Syuzhet is dictionary-based, drawing upon the Syuzhet, Bing, Afinn, and NRC lexicons. In my poster, I display several example analyses of literary texts using Syuzhet in order to reveal the limits of the package and its lexicons. For instance, Syuzhet draws upon simplified vectors of sentimental words in order to analyze the emotional flow of texts, creating subjectivity problems with the four lexicons employed in the package. In addition, Syuzhet does not properly deal with negators, amplifiers, de-amplifiers, and adversative conjunctions/transitions. Lastly, I demonstrate how to use the settings in Syuzhet in order to avoid creating faulty results. By showing the limits and usages of Syuzhet and the lexicons it draws upon, I can detail what digital humanists should consider when conducting sentiment analysis and contribute possible improvements for the package and lexicons.
Keywords: TextMining, OpinionMining, SentimentAnalysis, Syuzhet, Lexicons
"Make an Idol of the Hoe": Tools in 19th-Century American Garden Literature
Stanford University, United States of America
“Personal garden writing” (PGW) evolved in the 19th century as a niche genre of literature that intimately connected the individual reader to instructional horticulture. PGW intertwined a great deal of opinion with its instruction, offering readers easy insight into a writer’s worldview. Often this worldview advocated community in the garden, as PGW writers prided themselves on making the garden accessible to laypeople, but their reliance on small gardening tools might have either served or counteracted that purpose: while tools barred economically and physically disadvantaged classes from the garden, they also united common home gardeners in their attempts to find “pleasure or profit or health.” Using digital humanities tools to analyze language about tools in PGW could help quantify the role that gardening tools served in American community-building in the 19th century.
A 5-work sample of PGW was used to understand the valence and frequency of language surrounding tools in this genre. After the texts were digitized, they were searched to identify the tools mentioned. The texts were then cleaned with various tm_map functions in R, and then 214 Key Words in Context (KWICs) were collected to identify all “tool mentions” across the texts. Two kinds of sentiment analysis were performed on these KWICs, using the bing and nrc lexicons, which depict the positive / negative valence and more detailed emotions of input text, respectively.
Bing sentiment analysis shows that the language about tools in personal garden writing is more often and more strongly positive than negative, with the most positive KWIC having a Bing score of 4 and the most negative KWIC having a Bing score of -3. The positive scores of KWICs for ”tool mentions” suggests both PGW writers’ embrace of tools in the garden and their tendency toward positive language in general. NRC sentiment analysis shows that the sentiment most strongly associated with tools in PGW is trust. This result speaks to the compacts of the garden: between workers and plants, or writers and readers, exist covenants to protect the land and its laborers.
PGW demonstrates how, as symbols of trust, tools built community in American gardens. In 19th-century search for kinship, gardeners were right to “make an idol of the hoe.”
University of Pennsylvania, United States of America
Vector-space representations of meaning have become central to modern natural language processing techniques. But there is still not much theoretical justification for their success. Why can some aspects of meaning be captured in these linear structures? Perhaps more importantly, what aspects of meaning cannot be so captured?
This poster presentation will summarize some preliminary research that may help answer these questions. Rather than regarding word vectors merely as useful outputs of a black box learning algorithm, we can regard them as measurements of the possible infinitesimal change in meaning that a word could undergo. This way of thinking about vector-space semantics can be translated into a precise mathematical form, which can be used to derive new algorithms for generating word vectors. After introducing this theoretical framework, I will briefly describe a simple algorithm based on it, and show its performance as compared to other familiar word embedding models.
Finally, I will offer a few reflections (and invite feedback) on the broader implications of these ideas. In addition to providing a richer richer theoretical basis for understanding the behavior of word vectors, this framework may suggest new connections between quantitative and intuitive modes of reading. To show one possible direction for future work, I will discuss a way to reframe the concept that Jacques Derrida calls “iterability” in terms of linguistic infinitesimals, repetition, and nonlinearity (in the strictly mathematical sense). This reframing may not be perfectly true to Derrida’s ideas, but will, I hope, show unexpected overlaps between these seemingly very different paradigms of reading.
Using Webscraper.io to Scrape Digital Humanities Websites
Penn State University
Omeka CMS is a system that allows users to organize a wide range of digitized artifacts and related metadata on the web and present them in collections, exhibits, or a combination of both. While Omeka is beneficial for creating digital projects, it lacks a robust technical infrastructure for exporting public data in an accessible way, either through API’s or metadata exports. This project involved methods of scraping information from Digital Humanities projects built in Omeka. The information harvested included the title of the project and its overall description, in addition to the names and description of any collection or exhibit. Other pertinent information about creating the project was also extracted, including organizational affiliation and contributors. In this poster presentation I will discuss Webscraper.io, a Chrome extension, which is used primarily for extracting information from e-commerce sites, but can be utilized for any published website. This browser extension uses the underlying HTML code generated by Omeka and the browser’s developer tools to select different elements on the web page and then make a copy of the information into a CSV file. This allows Digital Humanities scholars to curate a metadata collection of web-based projects. This type of web scraping does not require any administrative access to projects, allowing users to scrape websites that no longer have such access.
20-30 minute poster presentation
The most relevant groups are those within the Digital Humanities community who use or administrate online content management systems. Librarians may find the scraping methodology important for curating online content.
“Not Exceeding Ten Miles Square”: A History of Washington DC’s Rectangular, Nondescript, Document Boxes
University of Maryland, United States of America
In this poster proposal, I exhibit my ongoing research from my dissertation work on early document storage containers used at libraries and archives. My current work focuses on knowledge infrastructures (an ACH 2019 topic of interest)—my research sits at the intersection of media archaeology and critical information studies. I draw on theorists writing on media storage including Wendy Chun, Shannon Mattern, and John Durham Peters. In my project, I present an untold history of cardboard storage containers adopted in the World War II period at US libraries and archives, including the Library of Congress and National Archives. Specifically, I focus on those units that originate in the Washington DC area, which were the most successful box designs and were adopted rapidly at nearby federal records centers. I build on my previous research, which traces the rise of paper storage boxes as a replacement for steel archival boxes (due to wartime steel shortages and high costs). I visually present patent drawings of the Woodruff File container, designs for the paper Hollinger document box, and screen-captured images of the Library of Congress developed BagIt digital container. Across these three containers, I unpack the affordances of each design that led to their success. I argue that above all, a more economical storage container was needed in each case, where the box manufacturing site’s close proximity to the seat of federal government and an abundance of records centers only further advanced each box’s adoption.
Technological Histories in Software Studies
University of Louisiana at Monroe, United States of America
Scholars in software studies tell varied and wide-ranging technological backstories to explain the impact of computing on society. Since 2009, the MIT Press Software Studies series has provided a reasonable, though not necessarily representative, sample of the field with eight single- and multiple-author books that address different topics in software studies from a variety of approaches. While the series' stated focus is on “today’s software culture,” its authors cite technologies from various periods of time to illustrate differences and relationships that characterize our current cultural situation. In this poster, I analyze and visualize the chronology of technologies represented throughout the series to identify how that body of scholarship crafts technological histories of computing. My analysis and visualization demonstrate several patterns in the chronologies of the series. The authors assign definite dates to certain technologies, often properly named examples of hardware and software. The authors more ambiguously position other technologies, like television, as predecessors or successors to the properly named technologies or exemplars of periods of technological development. By overlaying the specific dates and less specific chronological relationships assigned to technologies across the series, my visualizations show differences in the amount and specificity of consideration given to different technologies, the number of technologies clustered in certain periods of time, and the historical relationships between technologies in the books. In identifying how technological history is constructed over the course of the series, I argue that software studies can focus on technological influences on computation that receive inconsistent scholarly consideration in the humanities.