‘You doubt what I wrote? Let me show you.’ The very rare and obstinate dissenter who has not been convinced by the scientific text, and who has not found other ways to get rid of the author, is led from the text into the place where the text is said to come from. (B. Latour, Science in Action, 1987).
In the 1970s, sociologist Bruno Latour stepped inside a neuroendocrinology research laboratory at the Salk Institute in California to seek answers to the questions: What happens inside laboratory walls? How are scientific facts produced in a laboratory? Questions such as these stimulated ethnographic studies of laboratories aimed at understanding scientific work. Laboratory ethnography was a seminal movement, which conceived of science as the social practice of constructing scientific knowledge rather than an area that merely seeks to reveal reality. The study of science and technology through direct observation at the root of where knowledge is produced – in the laboratory – contributed to enhancing the public understanding of the substance of science that is, in fact, a network of complex actors and factors. Science is no longer seen as a mystical work in a sterile laboratory but as a planned process of manual and conceptual activities entangled with social, economic, and political aspects.
To comprehend how scientific knowledge is generated, Latour suggested going to “the place where the text is said to come from”. Is it, therefore, possible to understand how humanities knowledge is produced by stepping inside a humanities laboratory and observing research, technical, and administrative teams at work?
Before “entering a lab”, I want to reflect for a moment on why the study of humanities knowledge creation has been largely unexplored and why it is important to undertake it. As Rens Bod asked: “How can it be that humanists care about the history of everything except about their own?” (2017). The reasons for the deficiencies in the studies of the history of humanities and scholarly knowledge production lie in a sustained belief that the humanities have no systematic way of reasoning that leads to discovery. If the field is not based on a systematic approach to knowledge it is not possible to identify and follow steps in the structure of argumentation. However, as Bod shows in A New History of the Humanities (2013), humanities knowledge is based on patterns – seeking regularities in the system – and principles – determining rules underpinning a chain of reasoning. If scholarly knowledge is, in fact, a system of patterns and principles, then not only is there no opposition between the humanities and the sciences but it should also be feasible to indicate and follow the components in the chain of scholarly knowledge production. The effort to open a Pandora’s box of humanities enquiry and discovery means to recognize humanities research as the process of systematic actions and seeking to reveal the components, operations, and rules of that process. It also entails looking at that process as a series of actions intertwined with economic-political and socio-cultural concerns to understand how these aspects determine the construction of the chain of knowledge.
The study of humanities knowledge creation means to detect the process of reasoning that leads to making a claim translated into the text – the final product of humanities works – and also to go behind the text and reveal the mechanism and factors that leads to its formation. This, in turn, takes us to the messy and uncomfortable world of knowledge production and its power dynamics that influence the way we access resources, make our arguments, and articulate them. Going behind the text means to look at that text as the product of reasoning processes that are entangled with issues of the affordances of infrastructure, technology capacity, labour and social power, politics and nationality, and are also determined by many factors, including money, skills, connections, privilege, time, and risk. More methodological, epistemological, and ethical discussions are needed to find an approach which would enable us to dig into these layers and map the complexities of knowledge production.
A lack of understanding of how humanities knowledge is created may lead to a situation when claims and discoveries made by the humanities are falsely credited to science. When Nature Human Behaviorpublished the findings of William Thompson, a researcher in computer science at Princeton University who used machine learning to study how culture affects the meanings of words, the public was divided. The media described the work as “the first data-driven evidence that the way we interpret the world through words is part of our culture inheritance” (Nuwer, 2020). Humanists’ reactions were quite fair and predictable: the lists of philosophers and linguists were shared across social media platforms pointing out that the humanities have been saying that language is shaped by culture for decades. For humanists, the claim stated in the scientific paper sounds trivial, for the public, it was quantitative evidence that the word “beautiful” means different things in different languages. Note that the quantitative part – the process of reaching the results – was translated into colourful images that visualized the finding, the connections and differences between languages. The tension spawned around that paper showed how the public is not convinced by the humanities claims.
Lack of understanding about how the humanities generate knowledge has also been reflected in long-lasting beliefs and debates, such as “the humanities have no infrastructure”, “technology as tools” or “theory and practice distinction”. These statements are a striking exemplification of the fact that we do not know how humanities knowledge is really produced – knowledge that is heavily determined by infrastructure, co-created with technology, and articulated in different forms of materialized objects resulting from theoretical and making approaches. The study of humanities knowledge production can thus help the public to understand the substance of the humanities and help humanists clarify the position of the field in society.
My research interest in this area of inquiry is to study how digital humanities knowledge is produced in a laboratory environment. To approach it, it is necessary to reflect first on the methodological challenges and difficulties of the study. The first thing that comes to mind is the question of “what is Digital Humanities?”. As Stuart Dunn has rightly noted, there is a whole branch of research on this subject (Dunn, 2020). The short definition of the field is that it applies digital tools and methods to the humanities. However, Digital Humanities (DH) is not only about application, it is also about “critical inquiry with and about the digital”, as expressed in the motto of the Department of Digital Humanitiesat King’s College London. The clarification of “what is Digital Humanities” is important as it determines the objects and methods of investigating how DH knowledge is created. It is thus a puzzling and intriguing question to ask: “How to investigate something that is not clearly specified?”. However, I think the ongoing conversations on the definition of DH comes in fact from the lack of studying how DH generates knowledge. If we could understand the processes happening behind the text published in the journals for digital humanities, we would be able to not only specify the substance of the field but also identify what is needed to its better understanding and further development.
The second challenge to conduct this study is the fact that DH knowledge is generated as a result of the collaborative work taking place in the physical and digital realm, and in a particular space and time. The distribution of DH practices requires us to first identify the network of actors taking part in the formation of knowledge and then follow their intersections and actions. This entails building multi-sited fieldwork – introduced by George E. Marcus for the purpose of ethnographic practices – to interrogate more complex objects that cross-cut dichotomies such as the “local” and the “global” (Marcus, 1995). Therefore, although we study the practices of knowledge production in a single-site location, it is imperative to look at the place and products coming from that place as objects that are a part of the construction of a larger system of knowledge. The identification of actors involved in the manufacture of a particular product is important for defining a contour of the study, for its contextualization and specification of the timeframe of fieldwork. The study of DH practices occurring at the physical and digital environment – even more in the digital space in the times of Covid-19 – is a methodological challenge but it should not stop us from seeking to follow digital humanists at work even if that would mean to constantly cross the material and digital worlds.
The last challenge comes from the belief that digital humanities are a heterogeneous and large system comprising unique locally-situated settings and practices. This perspective has been developed within global digital humanities, the branch of DH focused on the global representation of the digital humanities community and the influence of the global academic system on local DH practices. If DH is a body of diverse practices varied due to locality, this would imply that there is no one set of “principles and practices” of DH knowledge creation but many different models that vary across places. While we should be aware that there are many histories and practices of DH, and therefore the chain of reasoning might differ, “the search for patterns and principles is less context-dependent” (Bod, 2020). Any knowledge production occurring in a particular part of the world is built upon the system of “patterns and principles” embedded in a local context. Therefore, while studying a particular local setting and seeking to reveal the systematic way of creating DH knowledge, it is necessary to keep in mind that in fact, we research localities – “the affordances of the study of the local, which invite us to think about local knowledge and the relationship between the local and global” (Pink et al., 2016: 123). The effort is therefore to understand how the situatedness determines the way how the systematic way of reasoning is performed and how the results are achieved. Local studies of DH knowledge production can thus provide a picture of the global configuration of the substance of the field and introduce heterogeneous epistemologies – beyond only Western approaches – of building, collaboration, and experimentation.
Keeping these challenges in mind, I have joined King’s Digital Lab (KDL) as a Marie Curie Research Fellow to conduct an ethnographic study of digital humanists at work, combined with a critical analysis of local infrastructure. KDL is a unique lab that is made up of Research Software Engineers (RSEs) who work on technical research solutions for conducting digital research in the humanities and social sciences. At this point, you’re probably wondering what a RSE-based DH lab can tell us about scholarly knowledge production?
As this is my research problem, I don’t know the answer yet. What I suspect is that in order to understand how DH knowledge is created, one must get into the substrate of DH work – the technical infrastructure layer of producing and providing devices, software, and tools. By starting ethnographic work from the underlying substance of DH work we might be able to comprehend how the production layer determines the process of reasoning and also how it embodies critical insights into the socio-technical world. As Susan Leigh Star showed in her seminal work on the ethnography of infrastructure (1996), technical infrastructure constitutes an integral part of social life that structures systems, conditions social practices, and brings things into being by connecting and materializing them to other things. The practices running on technical infrastructures are not fixed things but they are afforded and constituted by that socio-technical assemblage. Tools, software, and technologies reveal themselves as the relational and dynamic aspects of DH knowledge production. As Nicolas Gold rightly said while discussing the importance of software-based services for humanities research: “Similar approaches have been used to great effect in scientific discovery where the process of reaching a result can be as important as the result itself” (2009). This is why I am interested in interrogating underlying layers to find out answers to the following questions: How does a local infrastructure affect research practice? How are technologies intertwined within scholarly work? What are the roles of material and digital objects in DH knowledge formation?
RSE-based DH labs can also offer an insight into the nature of collaboration between people representing different epistemological backgrounds. How do RSE and DH people work together, and with non-DH researchers? How do they negotiate and reach consensus? How do they establish a common language to understand and inspire each other? The relationship between RSE and DH is getting more and more attention as it is mutually agreed that these two communities need each other for “engineering” a better world, based on values of social justice, equality, and inclusivity. The Alan Turing Institute has recently contributed to this process by releasing a set of recommendations for how DH and Data Science communities can more easily and better work together to realise the full potential of interdisciplinary work: “Without such collaborations, there is a substantial risk that data-driven research does not say anything new or meaningful, repeats well-known distortions, or introduces new forms of bias at an even larger scale” (McGillivray et al., 2020: 11). It is interesting to observe how the humanities have been called to be included in the design, production, and evaluation process of computational tools. In this sense, ethnographic research can contribute to a better understanding of these two communities and identifying ways how digital/humanists can be more included in RSE decision-making strategies.
Over the next two years, I will be exploring KDL as a case study locally grounded but situated in a global dynamic of knowledge systems. I will look at it from the historical perspective of the development of DH in the UK and in the context of the current state of the field across the world. My research project is conducted in close relationship with the King’s Department of Digital Humanities (DDH) on which a 40-year legacy KDL was built. It is thus an epistemologically and historically rich case study of DH supported by two unique institutional units: KDL and DDH. Drawing on theories of STS and Critical Infrastructure Studies, I will thus seek to understand how the Lab work, practices, materials, products, and culture contribute to building something called “DH knowledge”.
Bod, R. 2013, A New History of the Humanities: The Search for Principles and Patterns from Antiquity to the Present, Oxford: Oxford University Press.
Bod, R. 2017, How a New Field Could Help Save the Humanities, The Chronicle Review, February 19, 2017.https://www.chronicle.com/article/how-a-new-field-could-help-save-the-humanities/.
Bod, R. 2020, How to Open Pandora’s Box: A Tractable Notion of the History of Knowledge, Journal for the History of Knowledge, 1.1 (2020): 5. DOI: http://doi.org/10.5334/jhk.28.
Dunn, S. 2020, Digital Humanities: a Department, a Field and an Idea, Stuart Dunn blog, January 31, 2020. https://stuartdunn.blog/2020/01/31/digital-humanities-department-field-idea/.
Gold, N. 2009, Service-Oriented Software in the Humanities: A Software Engineering Perspective, Digital Humanities Quarterly, 3.4 (2009). http://digitalhumanities.org:8081/dhq/vol/3/4/000072/000072.html.
Marcus, G. E. 1995, Ethnography in/of the World System: The Emergence of Multi-Sited Ethnography, Annual Review of Anthropology, 24 (1995): 95-117.
McGillivray, B. et al. 2020, The challenges and prospects of the intersection of humanities and data science: A White Paper from The Alan Turing Institute. Figshare. dx.doi.org/10.6084/m9.figshare.12732164.
Nuwer, R. 2020, Machine learning reveals role of culture in shaping meanings of words, Princeton Engineering, August 14, 2020. https://engineering.princeton.edu/news/2020/08/14/machine-learning-reveals-role-culture-shaping-meanings-words.
Pink, S. et al. 2016, Digital Ethnography: Principles and Practice, London: SAGE Publications.
Star, S. L. 1999, The ethnography of infrastructure, American Behavioral Scientist, 43 (1999): 377-91.
This article was originally published on King’s Digital Lab website.