There is a moment in fieldwork that I have never seen documented in a communications framework, but that I have experienced enough times to know is structurally significant.
A community member describes something they have observed: about the reef, about the seaweed, about the way fishing patterns have shifted over three seasons. The knowledge is granular, located, empirical in the truest sense of the word. It comes from years of daily contact with the system they are describing.
By the time that knowledge enters an institutional report, something has happened to it. The granularity has been smoothed. The location has been generalised. The three seasons of observation have become a data point. The person who held the knowledge has become a beneficiary category tagged by ID.
This is not malice. It is the logic of the document type. Institutional reports require standardisation. They require comparability. They require a register that funders in Geneva or London or Washington can read without a country briefing.
But the translation carries a cost. And that cost has two dimensions, one ethical, one strategic.
When community knowledge is translated into institutional language without a preservation layer, the organisation loses one of its most valuable assets: the granular, place-based understanding that no satellite data can replicate.
The ethical cost: communities whose knowledge is extracted and translated without credit, without return, without meaningful participation in how it is used experience a recognisable form of dispossession. They become data sources rather than knowledge holders. The expertise that is most relevant to the intervention is the expertise that is least visible in the documents that govern the intervention.
The strategic cost: the translated version of community knowledge is a diminished version of it. What makes community knowledge valuable (its specificity, its depth, its embeddedness in a particular place and time) is precisely what institutional translation strips away. The organisation ends up presenting a generalised claim when it could be presenting a specific, evidenced, irreplaceable insight.
The solution is not to stop translating. Institutional audiences are real and their document requirements are real. The solution is to build a two-layer documentation system: one that preserves community knowledge in its original form (recorded, attributed, accessible) and one that translates it for institutional purposes without claiming that the translation is the whole thing.
I have seen this firsthand across organizations as a live challenge. Coral reef restoration involves communities whose knowledge of reef behaviour is sophisticated and location-specific. Women-led seaweed enterprises produce market and ecological observations that are rarely captured in programmatic documentation. Youth marine leadership programs generate insights about coastal futures from the people who will actually live them.
The question I return to constantly: are we building an organisation that holds this knowledge, or one that processes it and moves on?
Does your organisation have a systematic way to preserve community knowledge before it is translated? Or does translation happen once, and what is lost stays lost?