Systemic bias on Wikimedia projects is well documented and well understood. Plenty of internal and external research confirms the unequal distribution of active editors, as well as the skewed representation of all human knowledge in its content. The topic seems to be so well researched that its result is axiomatically accepted, and the thinking seems to be `If they think it is missing, they can join and add it. If they think it is wrong, they can come on board and make it right'.
Geography and demography are not the sole aspect of our projects' systemic bias. Our research compared the ethnology of the Wikimedia editor community with that of the OvaHerero, an indigenous community in South-Western Africa. Even the educated elite of that tribe: academics, businesspeople, politicians, have not joined Wikimedia projects. The results hint at many other incompatibilities that prevent members of this community from participating. Some widespread principles like 'Anyone can edit' or 'Be bold' are so alien to their cultural self that they see the entire Wikimedia movement as unsuitable to preserve and codify knowledge. Moreover, the textual representation of knowledge does not align well with the oral knowledge transfer predominant in many indigenous communities.
My talk showcases a long-term experiment to create a local language Wikipedia that mirrors values and principles of an indigenous community. We used participatory design principles to establish how a group of native Otjiherero speakers would go about writing for Wikipedia. The results are astonishing. We further analysed communication patterns, particularly during knowledge transfer. From the values of the OvaHerero community, their way of organising and transferring knowledge, and the case study in their own environment I demonstrate how a Wikipedia in Otjiherero might look like if it is to be acceptable to that community.
Attendees will learn that there is not only the Western method of knowledge creation, validation, and transfer. Wikipedia mirrors those Western methods. But if all human knowledge is our target, the vessel to store knowledge needs to be designed to also accommodate non-Western repositories, for instance those of indigenous communities.
As a result of this session there should be a dialogue about which aspects of the software, the knowledge organisation, and the community setup favours Western knowledge and disfavours all other systems.