Lab Books

Christoph Engemann, Thomas Nyckel, Isabel Schmiedel, Mary Shnayien, Florian Sprenger

Lab books are more than iterative records of observations and data—they are an integral part of scientific practice. By storing and assembling protocols, series of data, tentative interpretations, messy observations, and speculations pertaining to the research process, they support the reproducibility of experiments and participate in the production of scientific facts. They transform the “protean state of the lab”1 into knowledge, but in doing so create sources of uncertainty for this knowledge themselves. They document places and times of discoveries and have already played decisive roles in invention and patent disputes. In this sense, the lab book has a disciplinary, an epistemic, as well as a poetic function, since it represents, in Bruno Latour’s sense, one of the mediators that translate between different instances of knowledge production, thereby transforming observations into language and a transportable, variable medium.

As the history of science demonstrates, lab books should not be considered as representations of what happened in the laboratory. Rather, they can be understood as media of intersecting modes of knowledge production and knowledge processing: They structure processes between human and nonhuman actors in the lab. Given the division of labor in laboratory practice, they are rarely private. By providing a medium for different acts of writing and cooperating, lab books do not simply document the research process but take an active part in them. The example of Thomas Edison’s laboratory books illustrates how experiments and approaches to solutions are taken up and—sometimes retrospectively—teleologies of the research process are constructed.

Today, analog as well as digital, collectively edited lab books exist in tandem in many laboratories. One advantage of digital lab books, which often resemble databases while still retaining the character of a notebook, lies in their ability to integrate diverse datasets, while also making them easily available to co-researchers or an interested public.

For the Virtual Humanities Lab at Ruhr University Bochum, a media studies lab for analysing contemporary phenomena, technologies, and practices of (and in) virtuality, we have set up our own digital lab book using Obsidian, a Wiki software based on simple text documents in Markdown format.2 In the fashion of open science, our lab book is adapted to humanities-specific workflows of research, writing and teaching as a medium of generating and reflecting knowledge alike. This duality is characteristic for hybrid labs, where the basic skill of academic writing is combined with new modes of knowledge production. In a self-reflexive loop, we are interested in the media of such knowledge production, which is why our lab book is both part of our methodological approach and a research object itself. Furthermore, our lab book serves to make the lab’s work visible to the outside world, while emphasizing research as a process with its possible dead ends, repetitions, and detours rather than presenting final results or products.

Before deciding on Obsidian, we had tested several networked lab books which left us unsatisfied, so we formulated minimal requirements our lab book should fulfill. Amongst these are the integration of as many types of data as possible (text, photos, videos, datasets, diagrams, code, etc.), a topology of entries via links and tags to show (unexpected) similarities and differences, the ability to truly own our data and evade lock-in effects, cross-platform operation of the software to ensure maximum compatibility, as well as a high usability so the entry barriers for new users could be as low as possible. These criteria, we imagined, would subsequently lead to a low administrative effort, while also reducing hierarchies among writers. Last but not least: these features should ensure that the lab book could grow with increasing numbers of users in the future, i.e. scalability. Based on these requirements, we tried different digital lab books with a publishing option. Commercial solutions such as CERF, LabFolder or SciNote were excluded for financial reasons and because their respective publishing options are quite limited.

We decided to use Obsidian, because the software itself is free and fulfills our criteria, using the plugin Obsidian Publish to present selected entries of the lab book online by costs of 170€ per year. The Obsidian interface is intuitive to use and Markdown easy to learn. As is characteristic of markup languages, formatting characters are integrated directly into the content of the files, which can be edited using any software if they are saved in the .md format. Embedding files and diagrams is realized via classic markdown commands as well as additional JavaScript plugins. In this way, different types of data can be displayed and linked with each other while providing minimal entry thresholds and distractions from academic writing. As Obsidian is designed to be used by individuals only, it has no generic multi-user accounts. We decided to use a file hosting service for synchronizing the lab book files amongst its users, as well as assigning individual folders to each user to avoid conflicting copies.

The use of lab books is always an object to unexpected usage practices and effects that are linked to scalability: Although the software’s usability held up to our expectations, a certain form of messiness has taken hold. This messiness resembles the chaotic structure of many written notebooks. It introduces new challenges for an open lab book that does not confine to the categories, forms and masks that are characteristic for commercial digital lab books. Although the participants in our lab-based teaching did not accidentally damage each other’s entries, embedding photos and videos has created a collection of files with unclear origins clogging Obsidian’s root directory. Furthermore, almost all students named their notes “Lab Book,” which led to indistinguishable items of unclear authorship shown via graph and search function.

Reacting on that, we decided to introduce a mandatory header for each contribution to our lab book, containing a brief list of self-chosen metadata and tags. Thereby researchers will now be able to work with the metadata of the lab book. At the same time, this experience of messiness consolidated our view of the lab book as a laboratory itself, as an infrastructure that fosters knowledge production by intervening into its manifestations: a lab book, whether analog or digital, should be the place where knowledge and its materialities convene and messiness is not a problem, but generative.

In this regard, our self-made lab book necessitates two strategies that oscillate between technical and humanities-centered rationalities, and thus also between two ways of dealing with messiness. Following a technical-informatic approach, the main concern is system maintenance, i.e., how to keep project inputs standardized to prevent messiness and keep the lab book tidy. Following a humanities-centered perspective, we must find ways to embrace messiness, which allows for “transformative relationships that might change the project as elements are added”3—allowing for a certain kind of messiness that should not be overruled in favor of rigorous scalability. Future use will show to what extent scalability can meaningfully structure a lab book for a hybrid lab as the VHL.

Christoph Engemann is postdoctoral researcher at the Collaborative Research Center Virtual Lifeworlds at Ruhr-University Bochum. Research and teaching on AR and Architecture, Genealogie of Graphs and Media of Statehood and eGovernment.

Thomas Nyckel is postdoctoral researcher at the chair for Virtual Humanities at the Institute for Media Studies at Ruhr-University Bochum. His research interests are digital media and hypercomputation, Karen Barad’s agential realism, the work of Alan M. Turing and Kurt Gödel and physical computing/programming.

Isabel Schmiedel is research assistant at the Virtual Humanities Lab at the Institute for Media Studies at Ruhr-University Bochum, currently writing a Master’s thesis on the intersection of autonomous driving, LIDAR and dis/ability.

Mary Shnayien is a postdoctoral researcher currently associated at the Institute for Media Studies at Ruhr-University Bochum. Her research interests include the history of science of cryptology and IT security, political affects in digital cultures, as well as the intersections of media, gender, and race.

Florian Sprenger is Professor for Virtual Humanities at the Institute for Media Studies at Ruhr-University Bochum. He has published on the biopolitics of artificial environments, automation and self-driving cars, epistemologies of surroundings, and mobile media.