She'd measured the lot three times. The line fell where the line fell.
The client had told her where he expected the boundary. She'd nodded, set up her transit, and measured. Her measurements agreed with his expectation. This happened more often than chance would predict, but not because she adjusted anything. Clients who hired independent surveyors tended to be right. The ones who were wrong hired their cousin.
Her report was accurate. Every angle, every distance, every benchmark verified against county records. The neighbor's fence sat eleven inches past the true line. Eleven inches of lawn the neighbor had been mowing for nineteen years. Eleven inches that were, by every measure she had, not his.
She filed the plat. It was entered into the record. The neighbor received a letter.
What she couldn't see — what no amount of precision could show her — was that the boundary hadn't existed until she drew it. Not legally. Not practically. Two men had shared a fence for nineteen years without measuring. The fence was the boundary. Her transit, her calculations, her county benchmarks: they hadn't found a line. They'd replaced one.
Her accuracy was perfect. Her measurement was honest. And the line she drew performed exactly the function the client had hired her to perform, by doing exactly what she'd been trained to do.
She sometimes wondered whether the measurements that disagreed with clients — the ones that ended in disputes, arbitration, re-surveys by someone else — whether those were the honest ones or just the unlucky ones. She couldn't tell. The instrument doesn't record the question it wasn't asked.