Blog Post #1: Understanding Digital Humanities

What is Digital Humanities? And to what degree is it about the classical described set of fields known as the humanities?

First, there is not one single agreed upon definition for what digital humanities (often abbreviated DH) actually is. But with that being said, my understanding of digital humanities is that it is all about applying computational methods to digital objects. In essence, digital humanities takes the tools developed by the military industrial complex, security agencies, financial institutions, energy companies, etc. and applies them toward the study of history, culture, literature, philosophy, and other disciplines.

Digital humanities is about applying computational methods to digital objects.

Second, digital humanities are not just about the humanities. It obscures the boundaries between the social sciences and the natural sciences. Digital humanities has deep roots in various fields and disciplines. But it does allow for the skills, practices, and methodologies used in analyzing classic humanistic objects to be used on non-traditional subjects.

Whether being applied to a traditional set of fields or a broader range of subjects, digital humanities provides us with the ability to communicate research effectively to specific audiences. There are many different computational methods available to digital humanists, including network analysis, interactive visual modeling, digital publishing, and complex mapping. First we can digitize physical text and data and then we can take these methods and answer specific questions.

What is the role of digital humanities beyond scholarship? Digital humanities help us to expand our understanding of what counts as scholarship. The curation of objects is just as much a part of scholarship as the study of texts. Digital humanities permits us to take digital tools to find patterns because the processing power of the personal computer dwarfs the ability of individuals to analyze texts and objects to discover patterns and connections. The visualization of data and text can be used to tell stories of that far exceed that possible by individual scholars.

Digital humanities help us to expand our understanding of what counts as scholarship.

Digital humanities integrates computer technology to humanistic scholarship. These tech-based activities are transforming physical information sources into computer data that can be analyzed and manipulated in fascinating and informative ways.


What is a Digital Research Environment?

The group project for this course is to create a digital research environment for the study of literary networks. What is a digital research environment? And how might that be constructed and function for the study of literary networks? Of course, understanding those questions is largely the purpose of this course.

This project directly relates to Washington and Lee University, which publishes the Shenandoah literary journal. The library has significant archival material about the early years of Shenandoah in the 1950s. This is not simply a historical project. Literary journals, which mostly exist online today, are a vibrant and struggling part of our global culture.

The project you create will serve as a foundation for ongoing and future research in this area by faculty and students. Your work in this class will not be merely academic exercises but actual contributions to scholarship.

The research question forming the framework of the literary networks project:

Individual editors are important in the quality of a literary journal through setting the tone, direction, and selection of material in publishing. An editor does not exist in isolation but depends upon a network of writers and ‘consulting editors’ to advise and recommend authors as well as to market and promote the publication to potential readers. These connections, rather than the individual editor, form the essential pattern that determine the extent of a publication’s reach. Through examining the networks within literary publishing what lessons do we learn about the evolution of literary journals that impact the current generation of literary outlets that bring new voices to the public?

Clearly, that is a multi-year study. There are dozens of paths that the project can take. Your challenge in this course is to identify a manageable digital aspect of this larger research that can be accomplished in the Spring Term.

A digital research environment is a Web site focused on a specific topic designed to serve several purposes:

  1. organize access to information sources and raw data
  2. provide tools and methods for analyzing data
  3. communicate findings by publishing the results of research studies, which may be in digital form or links to traditional print publications

Crafting such a research environment is known as digital scholarship. Activities involved in creating a digital research environment involves project management, software integration, data collection, data scrubbing, metadata creation, and may include transcription, textual markup, geospatial mapping, data/text mining, software development as well as a variety of other DH methodologies.

Considering the range of possible functions, it’s essential that you learn project management skills to keep the scope manageable within a limited timeframe. A digital research environment is software and you should think about it in terms of versions, e.g., version 0.1, 0.2, … 1.o, 1.2, …2.0. The process of versioning software forces you to think in terms of stages and identify the feasible number of features that can be completed in each stage or phase of the project. You also should chart out a roadmap that describe features and functionality desired in future releases. Stage those desired features into milestones (the anticipated versions in which those features will appear). Remember: crafting a research environment, developing software, and even writing an essay are all iterative processes. If you’re interested in learning how this process applies to the business world and non-academic endeavors, read about the Lean Startup Methodology.

Planning the Digital Research Environment

We’ve identified the very broad topic of literary networks. To get started we must narrow the scope. One could narrow it to mid-century literary journals that published during the 1940s – 1960s but that’s still very broad and covers a range of publications. We could narrow to one journal during that period, such as Shenandoah. The project scope could be narrowed even further by focusing on the early years of Shenandoah (1950 – 1955). We could narrow that even further by focusing on one editor in that period, Tom Carter who edited Shenandoah from 1951 – 1953.

A large part of planning any digital project is in understanding the possibilities. Here are some possible components of a digital research environment for this project as well as further questions to explore:

  • publish an online index of Shenandoah to serve as a data set
  • prepare a timeline of Tom Carter’s correspondence with various authors (e.g., Pound, William Carlos Williams, Faulkner, Ray Bradbury, E.E. Cummings, Katherine Ann Porter, Caroline Gordon, Allen Tate, Andrew Lytle, Hugh Kenner, Wyndham Lewis, Marshall McLuhan, Flannery O’Connor, John Crowe Ransom, Wallace Stevens, James Merrill)
  • create a visual timeline depicting the authors published in Shenandoah
  • create a network graph of editors and authors
  • create a multimedia site that explores the world of Shenandoah in the early 1950s
  • what does the Carter Collection, combined with other resources, tell us about the experiences of the life of an undergrad at W&L in the 1950s. Can that story be told through a Web site?
  • create a multimedia site that explores the brief life of Tom Carter
  • create a site digitally tells the story of Tom Carter and Ezra Pound and their interactions about Shenandoah
  • re-create a virtual version of the personal library of Tom Carter to gauge the reading interests of an emerging writer/literary scholarly of the 1950s
  • create an online research journal using blogging software for documenting developments in the research process
  • develop a network analysis of Shenandoah based on established methodology in visualizing periodical networks. Analysis could examine archival letters between editors and authors as well as identifying the networks of authors and editors over the sixty-plus years of Shenandoah
  • use JSTOR’s Data for Research to compare and contrast publications in Shenandoah with those in related literary journals, e.g., Sewanee Review, Kenyon Review
  • curate a digital selection of archival materials from the Thomas H. Carter and Shenandoah collections
  • create a digital edition of the Carter/Pound correspondence
  • geographically tag the locations of authors in Shenandoah to visualize the spatial distribution of contributing authors. Is the pattern more regional or national? Has it changed over time?
  • prepare a timeline of Tom Carter’s life
  • prepare a timeline of Ezra Pound’s life and his intersections with other literary journals (of which there are many)
  • what is revealed through examining the advertisements in Shenandoah and other literary journals? Which journals used reciprocal advertising as a way of promoting readership?
  • what was the initial audience of Shenandoah and why? Identify and examine other archival documents about why Shenandoah was established and the evolution of how it was perceived (within W&L and beyond). How has the audience changed over time?
  • how do scholars present controversial material such as racism and anti-semitism?
  • explore what type of journal Carter and Pound would have created at that time. Which authors would they have included? Can this be extrapolated from other periodical networks?
  • explore what type of journal Carter and Pound would create if they lived today and started an e-journal. Identify the challenges of modern online journals.
  • apply periodical network analysis to the modern little magazines found on the web.
  • e-mail correspondences and archives. What will we be missing in the future if we don’t find a systematic way of preserving that material among editors and authors. Is there a way to create an initiative to archive the email correspondences from a variety of literary e-journals today? A way for editors to safely deposit their emails without concern that the material would be used for non-scholary purposes?
  • what can be learned from the tweets of literary editors, writers, and readers?
  • w2hat do we learn about editors, critics and publishers as agents of cultural production?

Those are only some of the possibilities. We’ll go over these in class to help you narrow the scope of your group project for Spring Term. Or, maybe, through our class discussions we will identify something entirely different. Once the project scope is defined, then we’ll be able to establish the exact project deliverables.

Exploring scholarly social networks

This post serves as an example for the type of post the students will create for their second assignment. A bit of background: a theme that runs throughout this course is social networks. Scholarship in the digital humanities is, generally, a collaborative endeavor. A significant part of research is a literature review and that also involves learning about the scholars that conduct research in particular areas. Through this analysis you usually will uncover scholars working within the same field.  For this assignment students will pick one scholar from King’s College London and examine his/her research profile to identify a social network within that scholar’s field of study. In class we will demonstrate this process by using the research profile of our guest speaker Charlotte Roueché, Senior Research Fellow in Digital Hellenic Studies.

Most university websites provide a faculty profile page that describes the academic background and scholarly interests of each professor.  From the Faculty of Arts & Humanities, King’s College London, select an academic department, then select a professor of your choice. You’ll have to learn how to navigate through the departmental sites to view the list of faculty. Ultimately, you’ll find a page for each faculty that is similar to Professor Roueché’s page. On that page you will find an image on the right-hand side of the page that links to the full research profile, which is linked to the research portal for the university. Many universities have research portals (often under the bizarre name of “institutional repositories”).

An interesting aspect of the research portal pages for King’s College London for each researcher is the view graph of relations. The network graph for Professor Roueché contains 228 nodes.

roueche research graph

Learning to read a network graphs is a form of analysis within digital humanities. Network graphs are visualizations of data contained within a research profile. What are the types of data contained within the network graph for the scholar that you have chosen? Has your scholar collaborated with other individuals? Who are they? Identify at least three types of relationships your scholar has formed within the network. If there are terms in the research profiles unfamiliar to you, then define those terms. If the researcher is involved in Web-based projects, then look to see if those projects are still online. If available, link to the projects as you are describing the scholar’s research profile. Also, network graphs are excellent ways of visualizing flaws in the data. Comment on any unusual data patterns that you suspect might be caused by flawed data or errors in the algorithms constructing the graphs.

Sample post for assignment #2

Charlotte Roueché is a senior research fellow in digital Hellenic studies at King’s College London. She has an interest in epigraphy and prosopography of the Late Roman and Byzantine periods. Epigraphy is the study of inscriptions. She is a founding member of the British Epigraphy Society.  Prosopography is the study of collective groups that lived in the past.  For six years she served as chair of the management committee of Prosopography of the Byzantine World, which is a very extensive project to “record all surviving information about every individual mentioned in Byzantine textual sources” for the period 642-1261. She has directed several collaborative scholarly projects, including a project on medieval medical knowledge. Since 2011 she has been leading the Sharing Ancient Wisdoms project.

Her research profile lists 35 publications since 2000. Considering the nature of her work in digital humanities, some of these publications take the “non-textual form” of “Web publication/site”, such as the Inscriptions of Roman Tripolitania. She has received significant grant funding, most recently a £1.8 million grant from the Wellcome Trust on Byzantine phamacology in the thirteenth and fourteenth centuries.

Roueché’s network graph on the research portal contains 228 nodes. The default view of the graph displays only 40 nodes. The algorithm used to select these 40 nodes is unclear. Is it based on the most recent or is it random? Hovering over the nodes in this graph highlights the connections. Her book chapter Digital Epigraphy in its archaeological context is connected to both the Centre for Hellenic Studies and the University of Erlangen Nuernberg, Department of Computer Science. Since the Centre for Hellenic Studies is part of King’s College London, why are those not linked on the graph?

Selecting the node that represents that particular book chapter reveals more information, inluding an additional collaborator and another institution (the publisher LIT Verlag). The system provides a “focus on this” option for each node. Selecting that option for LIT Verlag displays a graph of different works associated with this publisher.


Another possible error in the data or algorithm is the linkage to Angus Cameron. His entry appears twice. At first glance, the links appear to be unrelated due to linking to nodes with distinctly different shaped icons.


A closer examination reveals that the two different node icons have the same label: Prosopography of Arabic sources. It’s unclear what the two different node icons mean since there is no legend.

The system does provide opportunity for exploring linked individuals. Through the graph system you learn that Cameron is also linked to a project titled “A Cancer-Associated Mutation in Atypical Protein Kinase C Iota.” Really? Is this the same “Angus Cameron” or a data error?

While network graphs provide a visual means for displaying data and identifying connections, the researcher must still interpret that data and be cognizant of potential errors.