The Origins of Digital Humanities: From Humanistic Computing to a Paradigm Shift

In 1949, a 37-year-old Italian Jesuit priest named Roberto Busa sent a letter to Thomas J. Watson Sr., then president of IBM, requesting access to the company’s latest mainframe computer to index every word in the 13th-century theological summa *Summa Theologica* authored by Thomas Aquinas. Watson’s reported reply—‘We do not do theology’—was followed not by a rejection, but by a shipment of IBM’s first commercial punch card tabulator to Busa’s office in Milan. Over the next 30 years, Busa and his team processed 10.7 million words of Aquinas’s text, producing the 56-volume *Index Thomisticus*, a structured concordance (a complete alphabetical index of every word in a text, paired with its contextual location) that remains a foundational reference for Thomistic scholarship to this day.^1 What seemed at the time a niche technical collaboration between a humanities scholar and a corporate computing lab is now widely recognized as the inaugural project of what we call digital humanities: a field that uses computational tools, digital archives, and analytical methods to ask and answer questions about human culture, history, and expression. This origin story is not incidental: it captures the core tension that has defined digital humanities since its inception: the integration of technical rigor with humanistic inquiry, and the question of how computational tools can deepen, rather than diminish, our understanding of human culture.

The decades that followed Busa’s pioneering work saw slow but steady growth in what was then called “humanistic computing,” a term used to describe the application of computing tools to humanities research, long before the field had a unifying identity or name. Through the 1960s and 1970s, as mainframe computers became more accessible to university researchers, a small but growing community of scholars began experimenting with computational tools to process large textual corpora. Literary scholars produced the first full-text concordances of canonical works, from Shakespeare’s plays to the novels of Jane Austen, allowing researchers to trace word usage and thematic patterns across entire bodies of work in hours rather than years. Historians began building structured databases to organize archival materials, from census records to diplomatic correspondence, enabling large-scale quantitative analysis of historical trends that had previously been impossible to conduct due to the sheer volume of source materials. In China, parallel pioneering work emerged in the late 1970s, when researchers at the Institute of History under the Chinese Academy of Social Sciences began digitizing the *Twenty-Four Histories* (《二十四史》), the official dynastic histories of China spanning more than 2,000 years. By the mid-1980s, the project had produced a searchable full-text database of 24 million Chinese characters, laying the groundwork for later large-scale digital humanities projects focused on Chinese classical texts, ethnic minority histories, and modern Chinese literature.

The 1980s and 1990s marked a turning point for humanistic computing, as the widespread adoption of personal computers eliminated the need for scholars to rent time on expensive mainframe systems, and the rise of the World Wide Web made it possible to share digital collections and research findings with a global audience. This period saw a proliferation of large-scale digital humanities projects, many of them focused on building searchable databases of cultural and historical materials. In the United States, the Brown University Women Writers Project, founded in 1986, began digitizing and encoding rare texts by early modern women writers, creating a structured corpus that would become a model for later gender-focused digital humanities projects. The Corpus of Historical American English (COHA), launched at the University of Wisconsin-Madison in 1991, now contains more than 400 million words of American text published between 1810 and 2000, and is used by linguists, literary scholars, and historians to track changes in American language and culture over two centuries. In the field of history, the China Historical GIS project, launched at the University of California, Berkeley in 1996, digitized and georeferenced historical maps of China from the Han dynasty to the early 20th century, creating an interactive platform that allows scholars to analyze changes in administrative boundaries, population distribution, and trade routes over 2,000 years of Chinese history.

This period of rapid technical adoption was not without friction. Many traditional humanities scholars viewed computational methods with deep skepticism, arguing that the reduction of complex, nuanced texts to searchable data points stripped them of the interpretive depth that defined humanistic inquiry. In a 1987 editorial in the journal *Critical Inquiry*, literary theorist Stanley Fish famously dismissed computational text analysis as “a solution in search of a problem,” asserting that computers were incapable of interpreting metaphor, irony, or the contextual meaning that lies at the heart of literary study. These debates laid bare a core tension that still defines digital humanities today: the balance between computational scale and humanistic interpretation, between technical method and interpretive depth. Proponents of humanistic computing argued that computational tools did not replace interpretation, but rather expanded the scope of what scholars could study: just as the microscope expanded the scope of biological research by allowing scientists to see structures invisible to the naked eye, computational tools expanded the scope of humanities research by allowing scholars to identify patterns invisible to close reading of individual texts.

The formal emergence of digital humanities as a distinct interdisciplinary field can be traced to a series of institutional milestones in the early 2000s. In 2004, the three leading professional organizations for humanistic computing—the Association for Computers and the Humanities (ACH), the Association for Literary and Linguistic Computing (ALLC), and the Society for Digital Humanities (SDH)—merged to form the Alliance of Digital Humanities Organizations (ADHO), the first global umbrella organization for the field, which now has more than 4,000 individual members and 45 organizational members spanning 60 countries.^3 By 2005, leading research universities including Stanford, the University of Maryland, and University College London had established dedicated tenure-track positions in digital humanities, moving the field beyond its earlier status as a niche technical service for traditional humanities departments. A further marker of institutional legitimacy came in 2009, when the Modern Language Association (MLA) added “digital humanities” as an official category for its annual conference, drawing more than 1,000 paper submissions in its first year alone. By 2023, more than 300 universities worldwide offered undergraduate or graduate programs in digital humanities, and the field’s global community included more than 10,000 active researchers spanning literature, history, art history, linguistics, archaeology, and beyond. This institutional growth was accompanied by a shift in the field’s self-understanding: no longer seen as a technical service for traditional humanities, digital humanities was now recognized as a distinct mode of humanistic inquiry, with its own theoretical frameworks, methodological standards, and research questions.

As digital humanities has grown in prominence, it has frequently been conflated with adjacent fields that also use computational methods to study human culture. Three adjacent fields in particular require clear differentiation: computational social science (CSS), digital culture studies, and digital history. Computational social science shares digital humanities’ reliance on computational methods, but its core orientation is toward answering