Sylvia had been working at Meridian Consulting for four years before Leon joined the team. She was competent, organised, and well-regarded � the sort of person who was given more complex work not because anyone announced it but because the work found its way to her desk and stayed there. She had cultivated, over four years, a reputation that was quiet and solid, which she preferred to the alternative.
Leon arrived in October. He was personable, well-dressed, and attentive in the way new colleagues are attentive when they are still learning the topology of a workplace. He sat two rows from Sylvia. Within the first week he had asked her a question about the client filing system, and she had explained it, and he had thanked her in a way that was specific rather than general � thanking her for the particular information she had provided rather than for being helpful in the abstract. She had noticed this and thought he was probably good at his job.
He was also, she learned gradually, very good at being helpful.
The first incident she could reconstruct in retrospect was in November. She had been preparing a client brief � a standard deliverable she had prepared dozens of times � and had sent it to the internal review folder. The following morning her manager had flagged a formatting error: the headers in section three did not match the client template. Sylvia was certain she had used the correct template. She opened her saved file. The headers were wrong.
"You might have grabbed an old template by accident," Leon said. He said it while passing her desk, without stopping, as if it were a negligible observation. "The old ones are still in the shared folder if you go through the archive subfolder instead of the main one. Easy mistake."
She had thanked him. She had also spent twenty minutes trying to remember if she had opened the archive subfolder, finding, as she did with many things that week, that her memory of the precise sequence of her own actions was less reliable than she expected it to be.
This was, she would understand later, by design.
The pattern that emerged over the following months was one that Sylvia could not perceive while she was inside it, which was the nature of the pattern. What she perceived was a cumulative sense of her own unreliability: small errors that she was certain she had not made but could not prove she hadn't; details she had communicated that colleagues seemed not to have received; a general impression, surfacing gradually in team meetings, that her contributions required more verification than other people's. She began double-checking things she had previously done automatically. She began arriving early to review her own work. She began, in the vocabulary she used to herself, to be careful.
Leon was helpful consistently. When something went wrong � and something went wrong on a regular schedule that she could not at that point identify � he was nearby, with a gentle explanation: an alternative account of what had happened that was always plausible and always positioned Sylvia's error as accidental and understandable. He said things like these systems are confusing and it could happen to anyone and I do the same thing sometimes, and what these phrases accomplished, she understood later, was to make it impossible to dispute the premise that an error had occurred in the first place.
She told her partner, in February, that she was struggling. Her partner said she seemed distracted. She agreed. She went to her GP about sleep. The GP said stress was a common cause of concentration difficulty and asked about her workload.
She did not consider, at that point, that the difficulty was not originating with her.
The shift began in March, with a colleague named Priya, who sat in the corner by the window and who had been at the firm longer than anyone except the managing director. Priya said, one afternoon in the kitchen: "I've been watching Leon correct you in meetings."
Sylvia said that she had been making mistakes.
"Have you," Priya said. It was not phrased as a question.
What followed was two weeks of Priya paying the same quality of attention to Leon that Leon had been paying to Sylvia. What Priya found, and documented in the straightforward way of a person who had been in workplaces for thirty years and knew how they worked, was a pattern of intervention: Leon was correcting Sylvia's work at a rate that did not match the error rate of anyone else on the team, including recent hires. Some of his corrections were accurate. A significant proportion of them were corrections of work that had been, before his intervention, correct.
This was difficult to prove in the way that difficult things are always difficult to prove: not because the evidence was absent but because the evidence required someone to decide that the pattern was meaningful. Priya had already decided. She had decided in November, watching Leon's expression in a team meeting while Sylvia apologised for a deliverable that Priya was fairly certain Sylvia had submitted correctly.
Priya took what she had found to HR. HR, to its credit, investigated. The investigation was thorough, uncomfortable for several people, and produced an outcome that Sylvia would later describe to a friend as: not justice, exactly, but legible. Which turns out to be what I needed first.
What took Sylvia longer than the investigation was the internal reconstruction. She had to unlearn, carefully and in sequence, the version of herself that eight months of careful management had installed: the version that was careless with templates, forgetful in meetings, prone to errors that she couldn't quite account for. This took longer than the professional process. It required her to look at specific incidents and ask which account of events was more parsimonious given what she now knew about the conditions.
The answer, almost invariably, was not the account she had accepted at the time.
She was competent. She had always been competent. The competence had not been disrupted by some organic failure of her attention or memory. It had been targeted, methodically, by someone who understood that the most efficient way to diminish a person's professional standing was not to attack it directly but to make the person do it themselves � to create the conditions under which they would correct their own record, apologise for their own accuracy, and mistake the revision for the truth.
She returned to work after Leon left � there had been a mutual agreement with HR, the language that means something other than what it says � and she was, within three weeks, back to being the person the work found its way to. This did not repair the eight months. But it confirmed something she needed confirmed: that the eight months had been an imposition, not a revelation.
She kept Priya's documentation. She kept it not because she expected to need it again but because there was something important about having a record � a careful, external, factual account � of the period when the internal record had been compromised. It was useful to know that the truth had a witness, even when she had not been able to witness it herself.