Six years ago Seel began keeping documentation. She started with a notebook, moved to voice notes, added photographs when she understood that words alone were not reliable. The system was simple: record each event immediately after it occurred. Note the time, the location, the other person present, what was said. The system was meant to answer a single question she could not otherwise answer: did this happen?
She was reviewing a photograph on her phone now. It showed her and Cren at a conference: a corridor, coffee cups in their hands. She could see the cups clearly. Both full. She remembered the conversation: Cren had said the project would extend into winter, and Seel had said she needed to think about whether she could stay with it.
She could not remember who had taken the photograph.
This was a new problem. She knew how to verify what had been said in a conversation. She could find her voice note from three minutes after, could cross-reference with Cren. What she could not verify was the photograph itself, because the photograph required someone to have taken it, and she had no record of who that was, and asking Cren to confirm who had taken a photograph was a conversation she did not know how to begin.
She found the voice note. Her own voice, three minutes after the timestamp on the photograph: had coffee with Cren, eleven-fifteen. She said the project extends into winter. I said I need to think. Follow up next week.
She listened to it twice. The voice note was unambiguous. She had recorded it as memory, not intention. She had been careful, in the early years, about that distinction. She had learned to say what had happened rather than what she hoped had happened or what she meant to do.
The photograph was taken from the front. Cren was looking at the camera. Seel was looking at her cup. The cups were full.
She put the phone face-down on the table. She did not make a note.