Field notes: the day a draft layer got the room arguing

We printed the draft layer on A1 sheets and taped them to the back wall of the ward office, and for the first hour nobody touched them. Forty-some residents sat in white plastic chairs looking at a wall of rectangles like it was a test they hadn't studied for. Then an elder named Mzee Juma got up, walked to the sheet covering his section, and put his finger on a boundary line between two structure footprints.

"That's wrong," he said. "The fence used to run there, before my brother's house. Now it runs here." He moved his finger about four centimeters to the left, which on the ground was maybe three meters.

That one correction broke the dam. Within twenty minutes we had six people at the wall at once, a red marker going from hand to hand, someone's daughter translating for a woman who wanted to explain why a strip along the north edge belonged to her late husband's family and not to the neighbor whose roofline the drawn plot line currently clipped through.

This is the part of tenure regularization that doesn't show up in the project log as a line item, but it's where the whole thing either works or doesn't: the moment a community looks at a drawn plot fabric and starts arguing about it instead of arguing about whether anyone should be drawing it at all.

I've run adjudication sessions both ways. The old way, you walk in with a blank cadastral sheet, a compass, and a lot of patience, and you spend the first two hours just getting people to agree on where the settlement even begins. People point at hills. People describe boundaries by reference to a mango tree that died eight years ago. You're doing geometry from memory and gesture, and the sketch you end up with by lunchtime is shaky enough that you'll be back out there with a total station three more times before anyone signs anything.

Starting from a drawn layer changes what happens in that first hour. Nobody at the table thought every line on the sheet was exactly right, and nobody needed it to be, because a line that's close gives people something to correct. Correcting a boundary is a much easier civic act than inventing one from memory. Juma didn't have to describe his boundary in the abstract. He had to say "not there, here," and everyone around him could see exactly what he meant and either back him up or push back.

By the end of the session we had forty-one plots marked with consensus lines, eight flagged as contested pending a second visit, and one argument about a shared well that nobody expected to resolve that day and didn't. That's a good day's output for adjudication. The eight contested ones will need a field visit with the parties present, probably with a tape measure and a lot of standing around while people reconstruct a boundary history that only exists in two or three heads. That part of the job doesn't go away, and it shouldn't. Adjudication is supposed to be a human process. The community decides what the parcel fabric is, not the imagery.

We'd flown a VHR pass over the settlement about six weeks before the session and run it through a tool that pulls structure footprints and candidate plot lines out of the imagery, so the field team had something printable before anyone set foot in the ward office. The output is a starting sketch, drawn from what's actually built on the ground, not a claim about ownership. It lines up with the rooflines and fence lines closely enough that people recognize their own plots on first look, and that recognition is what gets a room talking.

I still think the real work happened at that back wall, with the red marker and Mzee Juma's finger moving three meters to the left. The draft layer just meant we spent the morning arguing about boundaries instead of arguing about whether we could draw a map at all.

If your next adjudication session needs something better than a blank sheet to start from, that's what Cadastre Update is built to hand you.