The Digital Humanities Landscape: Paradigms, Practices, and Project Planning

When a literary scholar confronts a corpus of 10,000 nineteenth-century novels, or a historian faces a million pages of diplomatic correspondence, traditional close-reading methods strain against the scale of the material. This is not merely a problem of volume; it is a methodological inflection point that has given rise to the Digital Humanities (DH). At its heart, DH represents a suite of practices that apply computational methods to humanities research questions, and conversely, applies humanistic modes of inquiry to digital technologies. Yet, this field is notoriously difficult to define with precision, not because it lacks substance, but because its identity is deliberately pluralistic and evolving. Definitions often circle around concepts like "collaboration," "interdisciplinarity," and "critical engagement with technology," but these are outcomes rather than foundations. The more productive approach for a practitioner is to understand the underlying paradigms that structure DH work, the common project models that manifest these paradigms, and the practical mechanics of moving from a research question to a feasible project plan. This chapter provides that operational framework. We will move beyond introductory definitions to dissect the fundamental tensions and choices that shape every DH project. You will learn to distinguish between the paradigm of "building" (creating scholarly infrastructure) and "querying" (extracting knowledge from existing resources), recognize which of the four primary project models—archives, tools, analysis, or editions—best fits your research goals, and master the essential skill of scoping a project that is ambitious yet achievable. The chapter culminates in a portable project proposal template and a guided analytical exercise deconstructing existing projects, equipping you with the critical lens needed to design your own work from a position of informed strategy.

Core Paradigms: The "Building" vs. "Querying" Divide

The most fundamental, yet often unstated, division in Digital Humanities practice is between two core paradigms: building and querying. This is not a binary opposition but a spectrum, and most significant projects engage with both to some degree. However, the primary orientation of a project determines its methodologies, its required skill sets, its evaluation criteria, and often its scholarly output format.

The "building" paradigm is concerned with the creation of scholarly infrastructure. This includes digital archives, textual corpora, databases, platforms, and tools. The scholarly contribution here is the resource itself—its comprehensiveness, its accuracy, its sustainability, and its design as a research instrument. The research question is often implicit: "What would it mean to have all of Jane Austen's novels, with every variant spelling normalized, linked to a historical dictionary of the period, and made available via a robust API?" The answer is the built object. The intellectual labor lies in editorial decision-making, data modeling, ontology development, and interface design. Success is measured by adoption, citation, and long-term preservation. A classic example is the "Perseus Digital Library", which began not with a specific hypothesis to test, but with a vision to build a comprehensive, interlinked collection of classical texts and their interpretations. Its scholarly impact is the ecosystem it created, enabling countless other "querying" projects.

The "querying" paradigm, in contrast, uses existing digital resources to ask and answer research questions. The resource is a given; the contribution is the new knowledge or interpretation derived from it. This includes stylometric analysis, topic modeling, network analysis, distant reading, and geospatial visualization. The research question is explicit: "Can we identify previously unknown co-authorship in the Federalist Papers using computational stylometry?" The answer is a finding, supported by evidence and method. The intellectual labor lies in algorithm selection, parameter tuning, statistical validation, and, crucially, the humanistic interpretation of computational results. Success is measured by the persuasiveness of the argument and its contribution to a scholarly debate. A prime example is the Stanford Literary Lab's pamphlet on "The Shape of History," which used topic modeling on a large corpus of historical texts to argue for a shift in historical discourse around 1800. The built object (the model, the visualization) is subordinate to the interpretive claim.

In practice, these paradigms are deeply intertwined. Building a tool like Voyant Tools is a "building" project, but its very purpose is to enable "querying." Conversely, a querying project often necessitates building small-scale, custom scripts or workflows. However, consciously identifying your project's primary paradigm is critical for planning. A "building" project requires long-term sustainability planning, data curation expertise, and often significant upfront funding for infrastructure. A "querying" project requires strong methodological grounding, statistical literacy, and a clear path from computational output to humanistic argument. Misidentifying your paradigm leads to mismatched expectations: a funding agency might reject a "building" proposal for lacking a clear research question, or a journal might reject a "querying" article for not providing reusable data or tools.

> 📖 Print Note: The following sidebar would appear as a formatted text box in the print edition. > > Case in Point: The "Building" of "Building" > The distinction is not merely academic. Consider the "Oxford English Dictionary" (OED) itself. Its first edition (1884-1928) was a monumental "building" project—a systematic, historical dictionary constructed from millions of quotation slips. Its primary contribution was the resource. The second edition (1989), while a revision, was also a "building" project in the digital sense: it was the first major dictionary to be created and stored electronically. The ongoing "OED Online" is a continuous "building" effort, adding new words and revising entries. Now, researchers use this built resource for "querying": linguists query it for semantic change patterns, historians for the first appearance of technological terms. The OED's enduring value lies in its enabling of both paradigms.

The Four Project Models: Archives, Tools, Analysis, and Editions

Moving from paradigm to practice, DH projects typically crystallize into one of four recognizable models. These are not rigid categories but heuristic frameworks that help articulate goals and methods. Understanding which model your work aligns with clarifies your planning process.

1. Digital Archives & Collections: These projects focus on the preservation, curation, and presentation of primary source materials. The core activities involve digitization (scanning, OCR), transcription, metadata creation (using standards like Dublin Core or METS), and the development of a discovery interface. The scholarly contribution is often twofold: making rare or fragile materials accessible, and sometimes re-contextualizing them through new groupings or links. An exemplary model is "The Women Writers Project" at Northeastern University. It is not merely a repository of texts by women before 1830; it is a curated collection with rich semantic encoding (using TEI), providing not just access but new ways to study early women's writing through its structured data. The research question might be: "What can a comprehensive, digitally encoded collection of early women's writing reveal about patterns of publication and readership?" The answer is embodied in the archive itself and the research it enables.

2. Digital Tools & Platforms: These projects create software, algorithms, or platforms that serve a broader research community. The contribution is methodological—pro