RESUMO
BACKGROUND: Over the years, the knowledge translation (KT) field has moved from promoting linearized models to embracing the importance of interaction and learning. Likewise, there is now increased attention on the transfer of KT approaches to new environments. Some scholars, however, have warned that ideas about transferability still hinge on linear thinking and doing. In the current study, we therefore sought to use a more reflexive approach to KT and to study how actors align KT approaches with their local environments. METHODS: Our (auto) ethnographic study took place in a wider KT project. This project intended to combine three components: (1) co-organizing demand-driven, locally led and embedded KT cycles in Cameroon, Jordan, and Nigeria, (2) building upon established KT methods and (3) equipping and empowering local teams. We conducted 63 semi-structured interviews with key KT actors, observed 472 h of KT practices, and collected a paper trail of documents. At the same time, we also compiled project exchanges, such as project documents, plans, protocols, field notes, meeting notes and an archive of (email) correspondence between project members. We analysed all data abductively. RESULTS: We show that there were numerous moments where the design of our project indeed enabled us to align with local practices and needs. Yet this often did not suffice, and the project design sometimes conflicted with other logics and values. By analysing these tensions, we want to show that doing KT work which acts upon different values and knowledges and is sensitive towards the different effects that it produces demands both structuring projects in a specific way and requires significant alignment work of KT actors in practice. CONCLUSIONS: We show that practising KT more reflexively relies on two important conditions. First, KT projects have to be structured with sufficient discretionary space. Second, even though the structure of a project is important, there will be continuous need for alignment work. It is important to facilitate such alignment work and to further support it. In the discussion of this paper, we therefore articulate three design principles and three sensitivities. These elements can be used to make future KT projects more reflexive and sensitive to (social) complexity.