From Concepts to Texts and Back: Operationalization as a Core Activity of Digital Humanities

Abstract

This article puts operationalization as a research practice and its theoretical consequences into focus. As all sciences as well as humanities areas use concepts to describe their realm of investigation, digital humanities projects are usually faced with the challenge of ‘bridging the gap’ from theoretical concepts (whose meaning(s) depend on a certain theory and which are used to describe expectations, hypothesis and results) to results derived from data. The process of developing methods to bridge this gap is called ‘operationalization’, and it is a common task for any kind of quantitative, formal, or digital analysis. Furthermore, operationalization choices have long-lasting consequences, as they (obviously) influence the results that can be achieved, and, in turn, the possibilities to interpret these results in terms of the original research question. However, even though this process is so important and so common, its theoretical consequences are rarely reflected. Because the concepts that are operationalized cannot be operationalized in isolation, operationalizing is not only an engineering or implementation challenge, but touches on the theoretical core of the research questions we work on, and the fields we work in. In this article, we first clarify the need to operationalize on selected, representative examples, situate the process within typical DH workflows, and highlight the consequences that operationalization decisions have. We will then argue that operationalization plays such a crucial role for the digital humanities that any kind of theory needs to take off from operationalization practices. Based on these assumptions, we will develop a first scheme of the constraints and necessities of such a theory and reflect their epistemic consequences.

Empirical evidence in conceptual engineering, or the defense of 'predictive understanding'

“…These subsamples are shown in the first step to have broader relevance within the context of the philosophical inquiries at hand. One of the key benefits of this procedure is that it allows a human validation of the results, a step that is seen as a desired practice in digital humanities (see Pichler & Reiter, 2022).…”

“…The discovery of concepts using digital methods is not entirely a priori; it necessitates an initial understanding of where to focus one's attention, identifying terms that might not align well with existing philosophical accounts. This human-driven identification of key areas is in fact a crucial step for any digital practice in humanities (Pichler & Reiter, 2022). Nevertheless, after pinpointing these primary areas of interest, the extent of discovery is largely determined by the quality and scale of our data corpus.…”