In this curated analysis, we aim to build a comparative, collaborative understanding of assumptions about “knowledge” in our different domain areas. These are not meant to sound like a cohesive “voice,” but rather punctuated snapshots from each of us. By anchoring our work in a shared analytic question, we create a common frame that allows differences and resonances to surface without forcing consensus. This approach reflects the value of “explanatory pluralism” embedded not only in the data publishing software we leverage (Poirier 2017), but also in our collaborative method itself.
We each answered the question first individually about our own cases and then we came together for a synchronous discussion to tease out the shared learnings and insights across domains. The summary below includes our collective summary of insights as well as our individual responses.
Across our cases, we see a persistent tension between structure and agency. Knowledge infrastructures (including universities, publishing regimes, data systems, urban governance) shape what kinds of action and expression are possible. Individual agency is necessarily constrained; no one possesses total freedom within such systems. But agency is not static. It can be re-membered: first at the psychosocial level, as individuals recognize their own voice and standing within relations of power, and then collectively, as shared practices like graffiti, maandamano protests, AI tools, journal clubs, alternative archives, generate a wider consciousness of constraint and possibility. We think knowledge production becomes crucial in this process, not as a neutral output but as a site where agency is (re)awakened, rehearsed, affirmed, and redistributed. We’ve noted in our various examples that new knowledge infrastructures don’t just emerge out of nowhere. They are often incubated within dominant ones, stretching their limits through a delicate dance of participation and refusal. Once limits are reached, then collective re-membering can give way to the creation of parallel structures in which different perspectives and knowledge can more fully thrive.
This book itself is an experiment in such infrastructural re-membering. Rather than smoothing our voices into a single authoritative voice, we assemble them in a relationship. We are not asking to be accommodated within existing structures. Where space is limited, we stretch it; where it closes, we build otherwise. Re-membering for us is not a technical adjustment but a conscious, collective practice, an ongoing process of coming together to do what none of us could have done alone.
—
Leonida Mutuku
The artifacts showcase assumed citizen agency in different ways: As developers, citizens can harness knowledge infrastructures as active creators of AI tools for civic engagement. The GPTs created to educate citizens on the Finance Bill are a clear example of agency that AI can provide for advocacy. The artifacts also reference the role of communities within knowledge infrastructures as contributors of local context data. This assumes that knowledge infrastructures can also provide collective agency especially through representative AI systems that can be used for grassroots beneficence. The artifacts also show potential for citizen agency in how they control the use of their data through the underlying infrastructure in the form of federated databases and the alternative licenses.
Wambui Wamunyu
Our knowledge ecosystem has treated the citizen as a peripheral contributor, one who submits to the policies and legislation that are developed around the formal education system and other places of knowledge production and dissemination. But following the June 25, 2024 protests in Kenya, the citizens have been emboldened by their outrage, speaking against the rampant corruption, poor management of resources, and lack of equity and access to what they as citizens are owed in services by government. In so doing, they have asserted themselves as contributors and distributors of knowledge through the arts (animated reels, music and spoken word), open source data, social media discussions, and investigative storytelling. They have used those tools and the openness, versatility and reach of social media to speak to one another and re-member what they have known and have come to know about our nation.
Aurelia Munene
Dominant narratives in postgraduate education often treat students as passive recipients of knowledge. This pervasive assumption, deeply embedded within the knowledge infrastructure, actively silences student voices. The unfortunate consequence is a diminished capacity for generating new knowledge and reimagining our world in more interesting ways. This issue is particularly acute for African knowledges and scholars, who are already marginalized in global discussions about knowledge infrastructure. Systematically silencing the voices of emerging scholars further intensifies this marginalization, hindering collective efforts for advocacy and movement building. However, the assumption that students do not push back is being actively challenged. Through my experience with the Africana Journal Club, I have witnessed how students are actively speaking back to these dominant norms. They are taking a more proactive approach to supervision relationships, advocating for their research topics, and generally making their voices heard.
Syokau Mutonga
Graffiti allows people to express themselves in languages and technologies that resonate deeply with them. The artists become emitters and not just recipients of messages. With the 2021 MaVulture Campaign, the artists painted murals in the Central Business District to mock the infrastructure that implied certain parts of the city belong to the state and that citizens have no agency there. Painting murals calling out MPs as vultures on the blank walls near City Hall was a statement that the city is ours, all of us, regardless of how much we earn, our last name and our education. The Gen Z protests of 2024 and 2025 share a similar inspiration: citizen agency that fights for a city where we are all fully and equally human.
Angela Okune
This transcript assumes that owning data production and management enhances agency and that currently there is not enough technical capacity to shape the PID landscape (hence the need for training). This discourse assumes that engagement will be primarily institutional or mediated through professional roles (e.g. research roles or libraries). There is a lack of clarity on how those who are not affiliated with institutions or citizen /lay scientists can directly shape this initiative. I think this is true more generally, most researchers gain access, legitimacy and credibility based on their institutional affiliation.