# Virtual Data: Research Data Practices in the Postdigital Humanities > [!tip] Reference > Paper for the panel *Data multiple: An inquiry into methodologies and ontologies of data* at STS Graz, May 5. ## About the session Critical Data Studies has now been around for at least a decade, and research on how data is made, processed, and put to work is as relevant as ever. What was once called Big Data has, in a sense, fulfilled its promise in the form of Large Language Models. As you all know, these systems could not exist without ridiculous amounts of data — and an equal amount of data work performed by fellow humans. So, it makes sense to keep following these large quantitative datasets. At the same time, there is also a great deal to be gained from keeping a keen eye on smaller data and on mundane, everyday data practices. Whether large or small, _Data Multiple_ suggests that both deserve our attention, if we want to document the diversity of data. And the session is meant to push this further and to invite even more kinds of differences. For example, by researching research data and its management in different disciplines, we can hope to find all kinds of data practices. I picked this example not only because it is my personal field of research; it is also a case for the political relevance of the _Data Multiple_. At least in the German academic sector, research data management is still in the process of becoming infrastructure. This carries the real risk that a limited understanding of research data will be inscribed in institutions and technologies. The situation might be similar in other fields. This is why this session is interested in small-scale, qualitative, non-digital, non-measuring, sensing, and more-than-human data practices. The concept of the _Data Multiple_ is meant to suggest that every data is multiple in an ontological sense — this is drawing on Annemarie Mol and Klaus Hoeyer, of course. The point is not only that there are multiple _kinds_ of data, datasets, and data practices; it is that any given data might be enacted, used, and held together in multiple different ways, and thereby exist in many versions as _more than one, but less than many_. That is also the overarching question for this panel: how can this multiplicity of data be studied, and what would be the political and infrastructural implications of this? ## Introduction to the talk Research data are essential for planning, conducting, and funding all kinds of research, but disciplines handle their data very differently. Especially in the humanities, there is no consensus on the proper role and value of research data. My talk addresses this situation by building on critical data studies, research on smaller datasets, and everyday data practices. The talk draws on an ongoing, multi-year praxiography of a research institution from the humanities. The notion of _virtual data_ from the title refers to the multiplicity in which researchers in my field create and use data. This includes mundane practices such as note-taking, annotation, and sorting PDFs in folders. I understand these as data practices that aim to _catch traces of meaning_, and I qualify these as _virtual_ because they are necessarily preliminary and incomplete. But let's start at the beginning. ## Field Let me introduce you to my field. I work in what is called a _Sonderforschungsbereich_ (SFB), a Collaborative Research Centre. It is one of the larger funding formats of the DFG, the German Research Foundation. My Research Centre has been running since 2022, and the maximum is twelve years. The consortium consists of around 60 people, distributed across 15 sub-projects. The sub-projects come from different disciplines and work on their own, distinct topics, while the consortium shares a common programme that justifies the collaboration. My project within this consortium consists in studying the _research practices_ at the consortium itself. This stands in the tradition of laboratory and infrastructure studies, which also means I focus on practices, techniques, data, and infrastructures. The method of my research is a praxiography — that is, an ethnography of practices drawing on Annemarie Mol. This means participant observation, individual and group interviews, but also participatory workshops, an experimental publishing project, and classical surveys. ![[IMG_1699_compressed.jpg]] ![[IMG_3967_compressed.jpg]] ![[IMG_8566_compressed.jpg]] ![[IMG_7532_compressed.jpg]] Let's look at the Research Centre more closely. In contrast to the classics of science studies, this case is not a natural-science consortium but one for humanities, cultural-studies, and social-science. Overall, the consortium has a strong _media-studies_ vibe. Topics include, for example, contemporary literature under digital conditions, the exhibition of art with VR headsets, digital objects in medieval studies, and the sensors of autonomous vehicles. There is a pronounced interest in digital technology. The everyday ways of working are likewise digitalised, similar to other office jobs: email, Microsoft Office, files in folders. The research methods, however, are not "digital" in the sense of the Digital Humanities. This is what I call the _postdigital humanities_: digital and non-digital topics are equally relevant; everyday work is digitalised; but without the methods necessarily being digital methods. ## Geist Let's take a look at the goals and values of the work. We will get to the practices soon; for now, it is about goals, values, justifications. So, what do the researchers say they want to do, or are supposed to do? In order to describe this, the German term _Geist_ has proved productive. In Germany, _Humanities_ are called _Geisteswissenschaften_, right? So, I started to take the term seriously as a kind of ethnographic strategy of estrangement. It is not unusual that the term _Geisteswissenschaften_ is used affirmatively — on the website, through DFG classification, in project proposals and descriptions. Great, I thought — so it's all about _Geist_; just as the _natural_ sciences study _nature_, the humanities are responsible for _Geist_. There isn't a straightforward translation of _Geist_: it's more than _mind_ or _thoughts_, and it also includes a social or cultural component. Anyway, I asked my interlocutors whether they understood themselves as humanities scholars (_Geisteswissenschaftlerinnen_) and what the goals of their work are. Here are a few answers: ``` I actually want to present problems. I want to make things complex, complicated. ``` ``` [It is about] foregrounding the speculative dimension, that is, what could be, not what is. ``` ``` I like to read. […] Maybe it really is just about text. […] The humanities is about conceptual work and exploring meaning. ``` So it is about the _possible_, the _speculative_, about _contingency_; it is also about _critique_ and about _conditions_ (genealogical, material, technical). All of this can be captured well in an updated, de-idealised concept of _Geist_. I follow a proposal by Niklas Luhmann; he translated _Geist_ with _Sinn_, which is German for _meaning_. For Luhmann, _Sinn_ or meaning is a kind of super-medium that underlies all mental and communicative processes. It's a semantic referential structure that indicates alternative possibilities, and at the same time regulates access to those possibilities. _Geist as Sinn is the medium of all possibilities for thinking and speaking about the world._ In this way, the term points out the priorities of humanities work. It is work with, on, and for _Geist_ – a work on _cultivated contingency_, a systematic and elaborate exploration of what is possible. ## Data Alright, so Geist or meaning or _cultivated contingency_ are what humanities work strives for. But in order to produce _high-quality Geist_, practices have to be carried out; one must _do_ something. And this is where it gets interesting, because it is not at all clear what one is supposed to do. Here are two recent German publications on practices from within the humanities themselves: first, _Geistesarbeit_ by Martus and Spoerhase (2022); and second, from Sybille Krämer, who has been researching the material and practical dimension of humanities work for some time. I show you these to argue that this is not only an ongoing topic for the humanities, but rather a topic that has recently gained attention. My strategy to address this question is to specifically explore data practices in humanities work. Here is my working definition of research data from my field and based on theory: data only exist in the course of data practices (Mol 2002), and these practices aim to _fix traces_ of phenomena (Rheinberger 2021). In the natural sciences, these are traces of natural phenomena; in the case of the humanities, these phenomena are traces of _Geist_ — that is, traces of meaning. These can be words, of course, but also more diffuse qualities. For example, someone underlines a sentence in a text, or drags a PDF into a folder. These are data practices insofar as they fix and materialize a meaningful quality, like "this sentence is relevant", or "this PDF fits into a folder called _Definitely Read Later_". ![[spuren-00_compressed.jpeg]] ![[spuren-02_compressed.jpeg]] These are most mundane examples; there are also more specialized and elaborate procedures: Viscoll, TEI, MaxQDA, Obsidian, bibliographic databases, self-developed note-taking practices. Here is an overview of data practices I found in my own data — sorry for not reading it out. This overview shows the multiplicity of potential research data practices, and each practice can be addressed with the help of different techniques and applications. Here are two quotes to conclude this part. The first one from the philosopher Sybille Krämer, the second one from one of my interlocutors in the field: > "There are no humanities without \[…] objects such as documents, monuments, and artefacts \[…] being searched for, collected, dated, labelled, classified, edited, compared, commented on, and archived" (Krämer 2025: 12f). ``` I: Would you say that you work with research data? P: No. Well, I … Let's put it this way: I can understand … I mean, I know I could also answer the question with yes, but I don't _feel_ it. […] Data, for me, are still numbers. […] I speak with my texts. I don't process them. ``` Because both of these quotes are valid, I argue for a broader concept of research data: first, to emphasize the sometimes neglected data practices and their essential role in the humanities; and second, so as not to leave this multiplicity of data practices solely to the natural sciences. There is a rich tapestry of data practices in humanities research, which are not reflected on enough, inside and outside the humanities, and they do not readily fit into the frameworks of the natural sciences or traditional research data management. ## Outro Let me conclude. I propose the distinction between _Geist_ and _Data_ to point out a crucial dilemma in humanities research: on the one hand, humanities researchers feel responsible to explore what is possible, for cultivating contingency, and for the production of _Geist_; on the other hand, to do that, they need to employ data practices to capture Geist, to temporarily and materially fix traces of meaning. I chose the term _Virtual Data_ to describe the specific role of data in the humanities that comes from this. Following Peirce, Deleuze, and Derrida, _Virtuality_ refers to a quality of "as if" (Peirce 1902; Derrida 1998; Deleuze 1989; Lévy 1995). A _virtual_ something _works_ like something without _actually_ being it. In the case of data, this means _to treat data as if it were fixed_. I hope this concept can help to promote a more conscious and reflective approach to research data, by emphasizing the provisional nature of research data, all while recognising that temporary fixations are a necessary step in the research process. I close with a little reminder by John Dewey: > Data is \[…] something to be thought about. \[…] Data signify "material to serve"; they are indications, evidence, signs, clues to and of something still to be reached; they are intermediate, not ultimate; means, not finalities. (Dewey 1929: 99)