The analysis of visual and material characteristics of digital representations of historical documents still plays a minor role in computer-assisted methods. Most quantitative methods and the tools used in this context focus on the content and / or writing style of literary and historical texts; bibliographic codes are generally disregarded in analysis and interpretation beyond their representation as digitized images in digital scholarly editions. This paper presents an approach to the quantitative, algorithm-based analysis of visual-material characteristics in a corpus of letters from the period ›Germany around 1900‹ and demonstrates how various automatically determinable values regarding the visually perceptible spatial layout of the letters can be used to derive insights into letter-writing practices.
VCEditor (VisColl Editor) is a digital tool for modelling, visualising, and reconstructing the collation structure of books. This article discusses how tools like VCEditor shape research practices of medieval manuscript scholars. Based on in-depth interviews with the developers and with users, this article details how scholars use VCEditor in their work and, as a consequence, the possible implications for the way manuscript scholarship is conducted. This is taken as starting point for a theoretical consideration in which digitisation and its effects on scholarly material engagement and knowledge production is examined. The findings demonstrate how VCEditor operates not merely as a technical aid, but as a mediating contributor to knowledge production – one whose epistemic significance emerges at the point of interactive engagement between scholar, software, and medieval material.
In topic modeling, analyzing statistical results can be complex and often requires visual interpretation. The Oral History Topic Modeling Dashboard (OHTM Dashboard), a tool developed as part of the DFG-funded project ›Oral-History.Digital‹, offers a solution and is optimized for the specific requirements of life history interviews. Following a methodological introduction to the LDA procedure and a comparative overview of existing analysis tools, the dashboard functions will be presented. The use of OHTM-Dashboard will be demonstrated through a case study analyzing processes of home and settlement in a corpus of 991 interviews from seven different archives.
This article introduces a research data management framework developed within the ERC AdG project ›ARCHIATER: Heritage of Disease: The Art and Architectures of Early Modern Hospitals in European Cities‹ (PI: Chiara Franceschini). The study focuses on the visual culture of European hospitals from the outbreak of the Black Death (1347) to the Great Plague of Marseille (1720). The contribution recommends the use of WikiFAIR as a FAIR-oriented strategy for ensuring long-term accessibility and collaborative data curation. It demonstrates how Wikidata functions as a reference ontology for navigating, modeling, and linking project-specific data, including the structured documentation of buildings, artworks, and objects. Finally, it presents visualization approaches such as geographic mapping, network analysis, and quantitative evaluation. These methods provide clear benefits for ARCHIATER and offer transferable insights for future art and architectural history research projects.
The aim of this study is to develop a reliable approach to the theoretical formalization of the fairy tale and to demonstrate the possibility of implementing this approach computationally. The most important finding is the identification of the ›appearance of an action-bearing subject‹ as an objective criterion – previously lacking – for determining the smallest structural element of the fairy tale, the motif. In light of these insights, the first part of this study categorizes the actions and the corresponding actors involved in their realization. This is followed by an attempt to schematize the universal structure of the fairy tale. The next section is dedicated to providing markup as a means for the standardized encoding of content. Subsequently, the digital assistant developed for the semi-automatic annotation of the folk tale is presented. In addition, the tool for the visualized representation of the annotated data is introduced.
Museums shape our perception of history and culture. In order to understand object accessions and the composition of museum collections, it is necessary to analyse the collectors from whom the objects were acquired. What were their interests and motivations? What understanding of art and collectable objects still shapes our understanding of culture today? Two influential museums, the Metropolitan Museum of Art in New York (Met) and the Victoria and Albert Museum in London (V&A), were selected as examples from the large number of museums and museum objects. We were able to determine donation trends and structures as well as donors central to the network. We also identified gaps for further research, especially in the field of biographical and provenance research.
How can digital editions be published online in a stable and sustainable manner? One answer to this question is provided by dse-static-cookiecutter, a static site generator developed at the Austrian Centre for Digital Humanities (ACDH). Several digital edition projects are already based on the tool, which generates websites from TEI-XML-encoded text files and publishes them free of charge via the GitHub Pages service. This project presentation discusses the three concepts at the core of dse-static-cookiecutter: firstly, the paradigm of static websites; secondly, the principle of defined workflows (in the sense of automatically running processes that are configured in a central text file readable by both humans and machines); and thirdly, the use of code templates that help to structure different software projects in a similar way. In addition, the advantages and disadvantages of such a static digital edition are discussed, and the dse-static-cookiecutter is contextualised in the current landscape of digital editions.
With the proliferation of accessible machine learning tools, there is a pressing need for ethical frameworks within Digital Humanities. Although traditional source criticism is well established, Digital Humanities require a digital source criticism that considers both the historical sources themselves and the data creation process. Often misunderstood as solely gender-focused, Data Feminism provides such a toolkit for addressing bias and ethics. This working paper discusses how these principles originally focused on data science can be adapted to everyday Digital Humanities practice. It provides both theoretical grounding and practical examples, including a case study from our own work, demonstrating the relevance and application of Data Feminist principles for Digital Humanities.
The project Correspondence of Early Romanticism is dedicated to the potential of digitally supported analyses of letters from the years 1790 to 1802. It uses letter editions of key players in early Romanticism, closes gaps and provides full texts and homogeneous metadata. The gradually emerging corpus of around 7,500 letters will be evaluated using network analysis. In this contribution, initial graphical representations, using the example of greeting instructions in letters, shed light on the structure of the Romantic network of communication. The article discusses techniques of semantic annotation as a form of knowledge representation with a focus on the connexions we have introduced, which reproduce letter statements in the form of triples or quadruples.
Superstition has long played a vital role in both traditional and modern societies. In the Middle Ages, people used sortes (books of fortune) to seek answers about the future, both in serious and in playful contexts. Sortes texts, like the Prenostica Socratis Basilei , provide insights into the mediaeval system of values, as well as in the worries and hopes of the people. However, the majority of these texts have not been edited yet, with none available as a digital scholarly edition. Our novel methodology, which combines graph-based modelling and computational editorial methods, allows for a dynamic, precise representation, enabling deeper analysis and making these historical documents more accessible and meaningful to modern audiences.
Based on the exemplary research project for the investigation of a horror film corpus, the system for the categorical classification and annotation of suspense content as well as the results of the algorithmic recognition of shot lengths, camera sizes, brightness and volume values and faces are presented. If the initiation of a fear reaction can be regarded as the functional core of the suspense experience of a horror film, the work uses explorative-statistical methods to determine regularities in the structure of relevant content in this regard. These largely invariant presentation features are compared with the stimulus materials of experimental work on film perception for so-called functional interpretation. The reference works from the fields of cognitive science and neuroscience make it possible to hypothesise how horror films coordinate certain modalities in order to support receptive activities such as anticipation or reaction to danger. In the end, the quantitatively comprehensible interplay between content and form allows the conclusion that slasher films simulate characteristics of moments of danger to which the audience is naturally sensitised.
In an experimental setting, the interplay of topic modeling and grounded theory annotation in the exploitation of large-scale collections of qualitative research data is approached via a combination of quantitative pre-structuring and qualitative in-depth analysis. This raises issues of epistemology, sense-making, characteristics, compatibility, and different dynamics of both methods. A mixed-methods approach suggests that topic modeling provides a proposition that can be interpreted by researchers for a first structuring of a qualitative analysis. Nevertheless, researchers give away agency through the use of topic modeling. The combination of perspectives and their ethnographic reflection can also meet the demand for evaluation in machine learning.
The project ›MMMMO – Mechthild’s Medieval Mystical Manuscripts Online‹ attempts to digitally provide an overview of the manuscript transmission of a widely transmitted work in the Late Middle Ages, the Liber specialis gratiae by the mystic Mechthild von Hackeborn (ca. 1241–1299). By using web tools and hosting the data in a repository, the project on the one hand ensures a safe and sustainable deposit of the acquired data, yet allows on the other hand a continuous updating with new research findings, for example taking into account new textual witnesses or more precise identification.
Taking the insights from diagrammatology – rooting in German media and culture studies (Medienkulturwissenschaft) and semiology – as background, this essay seeks to answer the question of how the history of the visual representation of temporal processes can be captured, and how the emergence of the statistical time series graph has influenced our ideas about time, history, and stories. To this end, I will provide a cultural-historical overview that, beyond shedding light on the interrelationships between different discourses (e. g., economy, statistics, science, literature), can contribute to a critical examination of perhaps the most widespread type of diagram used today. This overview can be understood more as a proposal for discussion, aimed at encouraging methodological and historical reflections on digital humanities practices, rather than as a presentation of a closed topic.
We document the results of a retro-archiving and literary analysis of Dana Buchzik’s blog Ze zurrealism itzelf . The starting point of our study is to consider how the blog has developed since 2010, what poetic processes characterise it, and how its genesis and poetics can be reconstructed using what is available of the blog in the Internet Archive. We describe the structure of the archival object and use the data available to reconstruct older versions of the blog. We cover the specific types of changes in the course of the blog’s textual genesis, describe them as elements of its poetics, and compare thematic and poetological aspects of three selected reconstructed versions. The methods of retro-archiving and reconstruction as well as the epistemology of the archival object serve as a basis for assessing types of changes and literary versions and constitute a workflow that can be used to analyse other blogs. This workflow is also relevant for the preparation of a text-genetic digital edition of a blog.
Part of a sustainable development of the Digital Humanities (DH) are findings whose relevance extends beyond DH communities because they are providing answers to existing research questions. A possible way of entering into this conversation seems to be a reflection on questions of how digital analyses critically question basic concepts of linguistics and literary studies and thus sharpen their definition. In this paper, we are going to address such questions based on the example of text segmentation which forms the basis for many analyses, and which can be used, albeit in different ways, as a foundation for both reading and computational processing of text. We are taking exemplary issues from linguistic and literary segmentation practice as a vantage point for discussion in this paper.
