While reflecting on the first phase of working on this database, we realized how much we could learn from our own process, and that this process was also revealing of the data itself. Creating an interactive dashboard of what data had been entered initiated a conversation around these learnings.
Initially, we hoped the data could help us answer questions including:
- Are there any locational “hubs” of activity that have emerged from our research?
- Are there key artists and curators in this database? Which people have crossed paths with each other through their work?
- Are there any milestone moments in the timespan we are interested in? How have these milestone moments influenced subsequent activity? and
- What trends exist in the topics of the exhibitions we entered?
Instead of being able to answer such questions after working on data entry for over a year, unintended barriers in how we approached the data entry came to light.
Because we were all given latitude to research and determine which exhibitions and people to enter in order to create a “worlded” data set, each team entered exhibitions that aligned with their existing research interests and expertise. It is important to note that not every team had the capacity to enter data - most data, as demonstrated in the dashboard, was entered by the Heidelberg and Ottawa teams. This led to the formation of areas of independent focus, and subsequently, patches of bias in our database. For example, Emily Putnam from the Ottawa team focused on artist-run centres (ARCs) in Canada, beginning with ARCs in Toronto as the largest, most populated city with the most concentrated number of ARCs in Canada. In the dashboard, most of the Toronto-based exhibitions were held at artist-run centres and entered by Emily. While in Canada, Toronto is one of the major hubs of cultural activity, especially given its globally diverse population, it only appears as a central site in the WPC database because of researcher decision-making. In other words, Toronto only appears in exhibitions related to the worlding of public cultures worldwide because the database was driven by a selective entry process. Likewise, exhibitions themed around the concepts of worlding and transnationalism appear to be concentrated in the Global North. This does not reflect the locational reality of such exhibitions, only the limitations of our method thus far.
Our methodological fallacies include sample bias (analyzing a limited and misrepresentative set of information) and scarcity (the allocation of resources to what people judge to be most important) (Greenwald 2021). Furthermore, the emergence of bias in our data entry could have been somewhat mitigated if the funding structure of the Trans-Atlantic Platform had not limited our research team to the Global North, enabling us to follow intersectional feminist approaches grounded in community proposed by Roopika Risam (2023). One of Risam’s proposed principles, "No data visualization without representation," calls for the inclusion of many voices, particularly those affected by the work, to be part of the entire process.
Overall, this dashboard offers a focused snapshot of where our data entry efforts were directed throughout the first phase of working on this database. Beyond the data entry efforts, significant labour (from Thanasis Velios and Maribel Hidalgo-Urbaneja) went into the development of the database itself. WPC project members also participated in a series of critiques around the intellectual framework of the CIDOC-CRM, on which the database is structured. These are components to our overall efforts as the decolonizing data cluster that are not addressed in this dashboard.
While this type of reflection, too often overlooked in digital humanities projects, is challenging, it provides insights into the limits of our methodologies, and is especially rewarding for revealing learnings that will guide our future directions.
Works Cited
Greenwald, Diana Seave. Painting by Numbers: Data-Driven Histories of Nineteenth-Century Art. Princeton University Press, 2021. https://www.jstor.org/stable/j.ctv15r588t.3.
Risam, Roopika. “No Data Without Representation: Principles and Practices for Intersectional Data.” Presented at the Distinguished Lecture Series, University of Tennessee Humanities Center, February 6, 2023.
JV