The Digital Humanities Coursebook

An Introduction to Digital Methods for Research and Scholarship

Johanna Drucker

• Publisher: Routledge
• Year: 2021
• ISBN:
  9781000364538
• Edition: 1
• Topic: Sprachen & Linguistik
• Subtopic: Bibliotheks- & Informationswissenschaft

The Digital Humanities Coursebook provides critical frameworks for the application of digital humanities tools and platforms, which have become an integral part of work across a wide range of disciplines.

Written by an expert with twenty years of experience in this field, the book is focused on the principles and fundamental concepts for application, rather than on specific tools or platforms. Each chapter contains examples of projects, tools, or platforms that demonstrate these principles in action. The book is structured to complement courses on digital humanities and provides a series of modules, each of which is organized around a set of concerns and topics, thought experiments and questions, as well as specific discussions of the ways in which tools and platforms work. The book covers a wide range of topics and clearly details how to integrate the acquisition of expertise in data, metadata, classification, interface, visualization, network analysis, topic modeling, data mining, mapping, and web presentation with issues in intellectual property, sustainability, privacy, and the ethical use of information.

Written in an accessible and engaging manner, The Digital Humanities Coursebook will be a useful guide for anyone teaching or studying a course in the areas of digital humanities, library and information science, English, or computer science. The book will provide a framework for direct engagement with digital humanities and, as such, should be of interest to others working across the humanities as well.

1 Digital humanities overview

1a What is digital humanities?

Definition

Digital humanities work is done at the intersection of computational methods and humanities materials.¹ The research materials may be analog or born digital, and the integration with computational methods depends upon decisions at every stage of a project’s design. A few frameworks for how digital humanities research is structured are provided here at the outset. First is a general description of the components of digital humanities projects. This is followed by discussion of the fundamental activities in the lifecycle of digital work.

Components and activities of digital humanities

The components of digital humanities work can be expressed in this way: MATERIALS + PROCESSING + PRESENTATION. In this formulation, digital humanities projects begin with materials (images, texts, maps, three-dimensional models, sound and media files, or any combination of these) that are central to the research project. These materials are subject to computational processing (data mining or statistical analysis). The outcomes are organized in a presentation that may be web-based or offline, depending on the needs and goals of the project. A project can be digital without being online, but all online projects are based on digital files and computational processes. Here they will be explored in detail. Keep in mind that that term digital refers to information in binary form while computational refers to processes performed by algorithms.

The fundamental activities that take the materials through processing and into presentation are mediation/remediation, datafication/modelling, processing/analytics, presentation/display, and sustainability/preservation. Understanding the ways in which each aspect of these activities engages technical as well as ethical considerations helps shape responsible work in digital projects. This assists innovation that brings humanistic values to bear on technological procedures. Technological instruments are never value-neutral but seeing precisely where and how values affect implementation requires understanding the workflow.

Components and workflow

The components of a digital humanities project workflow involve methods that are likely to be unfamiliar to traditional humanists. They need a bit more explanation:

• MATERIALS have to be put into digital formats so that they are computationally tractable. This requires mediation/remediation and data modeling, creation of files, development of metadata for description and use, and other decisions that will have implications for intellectual property, access, sustainability, and use. The way in which materials become digital matters.
• PROCESSING uses computational methods and tools. Processing is often the black box of the digital humanist’s work, since much of it can be performed with off-the-shelf tools whose workings may be invisible or incomprehensible to the user. Having some understanding of these processes allows critical engagement.
• PRESENTATION often makes use of online platforms such as blogs, WordPress sites, or those specifically designed for humanists, like Omeka or Scalar. These may be hard to customize and may force a project to conform to a particular argument structure. Project-specific designs involve work and skill. Digital projects often have off-line outcomes in the form of publications or reports. Format has its own force in structuring the impact of a project and reinforcing its argument. Even elements like background color, font choice, and graphic style affect how research is received.

Engaging with any of these components requires striking a balance between ease of use and critical understanding. Digital methods can be learned gradually and build in complexity.

These components and the activities on which they depend are the foundation of all digital humanities projects, from the simplest to the most complex. At each stage of the discussion, issues relevant to implementation and design are combined with their critical implications. The decisions about how materials are made computationally tractable in digital formats have long-term consequences just as the use of processes by the researcher has implications for the way results are understood and interpreted. A presentation hides as well as shows many of these decisions, and documentation is an essential feature of responsible research work in digital humanities so that other researchers can recover the decision-making process that shaped the project. A healthy combination of hands-on, practical, how-to-do-digital-humanities, and critical skepticism is required throughout, particularly if values and methods central to the humanities are to find their way into the technologically driven activities. [See: Exercise #1 Analyzing a project.]

Fundamental activities

The activities fundamental to the design and implementation of research have been mentioned in the previous section.

Mediation/remediation

This work involves either making analog materials—such as maps, manuscripts, archaeological remains, specific editions, or primary documents of historical or recent events—available in a digital format or creating and using born-digital materials. Taking information in one form into another is the activity of remediation. The information in one medium (paper, film, stone, etc.) is remade in digital form. This creates information in a file format that a computer can process. Born-digital materials are also highly mediated and decisions about their formats are crucial to their use as well. Digital humanities projects often engage with both analog materials and born-digital ones, but all are managed in the form of digital assets.

Datafication/modeling

This is the work of abstracting discrete values from a phenomenon or artifact. These values may be expressed in numbers or texts and are necessarily a reduction of complex materials into a form for computation. With data we can automate processes of sorting, counting, comparing, or making statistical assessments. Materials or phenomena of almost any kind (such as the height of churches, the length of edited segments of film, the number of persons in a social network, or a list of names or places) can be turned into quantifiable and/or discrete data. The practice of datafication depends upon modeling. What are the assumptions that are built into deciding what can be extracted from an artifact, text, or other object of research? What terms can be used to represent or describe an object? Data modeling will be discussed in depth ahead. It plays a major role in the intersection of values and biases and the interpretative work central to the humanities in dialogue with technological methods.

Processing/analytics

This involves the automation of counting, sorting, or analyzing through computational processing. Every computational process involves intellectual models, no matter how automated its operation. To some extent, research questions are shaped by what can be done using automated processes. For example, if a map is generated using only two-dimensional data, its flat image cannot present the experiential understanding that comes from walking through the territory. The arguments that can be made in any process are constrained by its specifications, as will be explored in more detail in discussions ahead of visualization, data mining, network analysis, and other automated processes. The concept of analytics encompasses many aspects of feature extraction (deciding what is relevant in a file) and then processing those features through some activity of contrast or comparison. As in the case of data, these processes contain cultural assumptions that value certain elements of the cultural record over others. In addition, as processes become more sophisticated and engage machine learning and artificial intelligence (AI), they often reinforce biases built into their model.

Presentation/display

The presentation of results often takes digital form, sometimes in an online environment, as visualizations, maps, diagrams, stories, articles, or exhibits, and sometimes in analog or hybrid formats. Every research presentation is structured according to a narrative that organizes the display. Even repositories and collections present an argument or story through the structure of their interface. The simplest interface design embodies decisions about the hierarchy of what is important and what is not, what is to be revealed and what is concealed, in the presentation of research.

Sustainability/preservation

This must be taken into account from the initial design of a project. Initial design decisions will depend on the institutional site in which the project is conceived and will be managed, on the amount of resources and expertise available (costs and labor), and on other factors specific to the project (such as the intellectual property involved, issues of privacy and protections, and regulations to which the work is subject). Sustainability also involves consideration of ecological costs and human rights issues central to digital labor in a globally networked environment. All digital projects are costly to produce and maintain.

Computation and the humanities

Remediation/mediation, datafication/modeling, processing/analytics, presentation/display, and sustainability/preservation are the fundamental activities of digital humanities research. Each involves a series of decisions and tasks that will be described in the sections ahead in more detail. These activities make it possible for automated algorithmically driven processes to operate on humanistic artifacts in a digital format. Similar methods are used in other fields and many are drawn from fields far from the humanities. Much of the earliest digital humanities work had its roots in linguistics, where statistical and computational methods were first applied to keyboarded texts. But computational methods are largely discipline agnostic. We can use the concepts just described (e.g. remediation or analytics) to do work in biology, ecology, history, geography, performing arts, or economics and political work as easily as in the study of novels, poetry, or philosophy.

This recognition should not be used as an excuse to forget that the humanities have their own intellectual traditions. An ongoing question for digital humanists is whether or not these traditions, and their methods, have found a place in the field. Are computational techniques, many of which depend on quantitative methods from the natural and social sciences, appropriate for humanistic work? Digital tools and platforms have greatly augmented humanities work, and in this book, we will be concerned with these as an extension of humanities methods, rather than engaging digital techniques simply as a method in themselves. The question of what cannot be done with such methods—or how they are based on certain assumptions they enact in their use—will return frequently.

Tools and platforms change, but the principles of digital practice have become clearer and more codified over time. Therefore, this book takes a pragmatic approach: how to introduce the basics of hands-on practice in combination with critical principles in a way that can be transferred as tools change. For instance, if you know how to structure data to analyze a network, that data can be put into any platform (Gephi, Cytoscape, Tableau etc.) to generate a visualization. The difference among platforms is that they generally move towards greater ease of use, but greater complexity (more functions), over time. They become harder to customize or understand.

The underlying activities just noted—remediation, datafication, processing, presentation, and sustainability—remain central to digital work and methods in the humanities and other fields. The chapters break these activities into more specific tasks, so they will be referenced throughout in the sequence that follows the workflow from materials through processing to presentation stages. Topics such as mapping, visualization, managing intellectual property, and so on are subsets of the basic activities and will also be addressed within the chapter sequence as it follows workflows.

Debates

Much debate about the dynamic between technological methods and humanistic ones has been generated over the last twenty years, but in essence, the question is whether the capacities for judgment that include ambiguity, contradiction, cultural specificity, and other qualities that involve human beings in their individual as well as collective understanding of the value of expressions and artifacts can be adequately carried out by automated processes (Gold and Klein 2019). Or not. The answer is of co...