The Data Colonialism of European Research Projects

The next speaker in this AoIR 2022 session is Paula Helm, presenting on data colonialism. She begins by contextualising this work as emerging from a computer science project designed to build a new social media platform called WeNet that sought to encourage the diversity of user networks in order to combat the (myth of) ‘filter bubbles’. But in order to encourage diversity, such a platform actually needs substantial amounts of data about its users.

Especially problematic about that project was its engagement with users from a wide variety of countries around the world, from its positioning in the European Union, so that its data extraction processes could be seen as yet another form of data colonialism, or at least as patronising and paternalistic, and should be critiqued from a critical whiteness perspective. Indeed, EU research funding schemes overall could be investigated from these perspectives.

Such projects can thus be seen as ethnocentric rather than as designing for a more egalitarian pluriverse; they privilege certain researchers and groups, in the EU, while non-EU partners can carry out implementation but get no funding for research and design activities. More broadly, this data colonialism can be understood from an industry-critical perspective (which sees all of us as colonised by Big Tech) and/or from neocolonial perspectives (which emphasise the continuing colonialism of the Global North).

The project then focussed on the development of research and treatment of research subjects in four pilot countries (Paraguay, Mexico, India, and Denmark). Even the Danish users, from their considerably more privileged position, found they had too little data agency and would have liked to have been considerably more involved in working with the data themselves; users in Paraguay identified a very Eurocentric approach that wasn’t sufficiently attuned to their needs and lived experiences; Indian participants were especially concerned about their data protection and privacy, and given their high workloads had no particular incentives to engage (again also because they were not provided with any meaningful data access for themselves); while the pilot in Mexico was far more directly tailored to project participants, in part because one of the researchers on the project had strong prior knowledge of the local situation.

Rapid scaling towards pluralism through top-down standardisation is likely to fail miserably, therefore. The vicious circle between ethnocentric design and data colonialism, and between patronising and extractive logics, needs to be broken. Innovation needs to be rethought from a critical whiteness perspective, and potential for innovation may lie in joint self-reflection.