He arrived in September, the week after the bank holiday, when the office was still carrying the particular looseness of the late summer and people were not yet at their full alertness. His name was Brennan. He had come from a company two sectors over, and someone in management had described him, at the all-team meeting, as a strong hire. I noted the phrasing. Strong implied something specific about the way he had been chosen. I filed it without knowing yet why I was filing it.
The first test came on the Friday of that week.
We had been introduced in the corridor, briefly, and he had said something about my work that was structured as a compliment but functioned as an assessment. He said he’d heard I was good with the client-facing reports — which I was — and that he’d be interested in seeing how I approached them, which I had not offered. It was the kind of statement you cannot object to because it contains nothing objectionable. It was also the kind of statement that tells you something very specific about the person making it.
I told a colleague about it that afternoon. She said he sounded nice. I did not disagree. I went home and wrote down what he had said, and the date, and why it had registered.
October. The tests moved into a new category.
Brennan began asking questions about colleagues that were positioned as professional curiosity. He would ask, with the tone of someone simply trying to understand the landscape, how long someone had been in their role, or whether a particular person was well regarded by management, or whether a team arrangement had always been structured the way it was. These were not unusual questions for a new person. They were the questions a new person asks when they are mapping the territory.
What marked them as something other than ordinary was the precision of what he chose to ask about, and the timing of when he asked. He asked about the colleague I was closest to at work just after she and I had been visibly in conversation. He asked about the team arrangement involving my direct line manager in the week that arrangement had come under review. He asked about a person I had once had a significant professional disagreement with — a disagreement that had not been widely discussed — apparently at random, on a Tuesday, with a casualness that I found, on reflection, very carefully constructed.
I added these to the record. The record was now a notebook, spiral-bound, that I kept in my bag rather than my desk.
November was the month of the professional category.
Brennan began using my work in ways that were not quite sanctioned. He mentioned an approach I had developed — in a client meeting, in front of two senior colleagues — as something he had been “thinking along similar lines to.” He did not attribute it to me. He did not take credit explicitly; he created the conditions in which credit was not assigned to anyone in particular, and then stood in the space where it might have landed.
When I raised it with him afterward, he was genuinely puzzled-looking. He said he hadn’t meant to give the wrong impression, that he’d been trying to frame the approach as a team direction rather than an individual contribution, that if he’d been unclear he was sorry. The apology was complete and contained nothing that could be recorded as an admission. The thing he was apologising for was a misimpression, not an act.
I wrote this down too. I wrote down his exact words, which I had learned to retain with some precision by that point.
In the same month, he began cc’ing my manager on replies to my emails in situations where cc’ing the manager was not customary but was not forbidden. Each individual instance had a rationale — he was keeping the team informed, he wanted to make sure nothing fell through the gaps — and the aggregate effect was a steady low-level signal that my work required monitoring.
December. The physical category.
I want to be precise about this, because precision is the only thing that makes it legible. Brennan was not physically aggressive. He did not touch me in any way that would survive a description offered to a third party. What he did was manage proximity in a way that had a particular quality: a way of standing close enough that I was aware of it, of leaning across me to reach something on a desk rather than asking me to pass it, of positioning himself in doorways at the moment I needed to pass through them.
None of these things, individually, would constitute a complaint. Collectively they communicated something very clear, and the communication was the point.
I noted: date, location, what had been reached for, how long the proximity lasted, whether others were present.
January. The social category.
By January, Brennan had become integrated into the informal social arrangements of the department in a way that was, in itself, unremarkable. People liked him. He was fluent and warm in group settings and he had a quality of attentiveness that people experienced as interest. He was interested in them. This was not, as far as I could determine, false. He was genuinely interested in people in the way that a person who relies on information is genuinely interested in its sources.
What I began to notice in January was the specific pattern of integration. He had become close to people adjacent to me: the colleague I was closest to, the colleague who had been my informal advocate in a reorganisation two years earlier, the colleague who handled scheduling for our shared projects. He had not replaced my relationships with any of these people — I want to be accurate about this — but he had established himself as a node in each of those relationships, so that conversations that had previously been bilateral now had a third point.
I was not isolated. I was slightly less central than I had been, and this had happened through the ordinary social mechanics of a new person joining a team, and I could not point to any moment where it had been done to me rather than simply happening.
In February, I went to HR.
I had prepared carefully. I had the notebook. I had dates, times, locations, exact words where I had them, approximate words where I did not. I had organised the record into the four categories — professional, informational, physical, social — and I had written a summary that I thought was clear and well-structured.
The HR officer was a woman I had spoken to once before, in a different context, and she was competent and professional and she read through what I had given her with the attention it deserved. When she finished, she was quiet for a moment. Then she picked up one item from the notebook — the doorway incident from December — and read it again.
“This is ambiguous,” she said.
I said I knew it was ambiguous. I said that was why I had brought the whole record rather than the single incident. I said the pattern was the point.
She looked at the record again. She said she understood what I was showing her, and that she could see I had been thorough, and that she would need to think about how to proceed given that each individual incident would be difficult to act on.
I asked whether the pattern, in aggregate, was actionable.
She said she would need to take advice.
I left the meeting and walked back to my desk and sat down and looked at my screen for a while without seeing it. I thought about the word ambiguous. I thought about how she had reached for it immediately — not as a conclusion she had arrived at after consideration but as a category she had recognised at once. I thought about why that category had been available to her so readily.
And then I understood something I had been approaching for several months without quite reaching: that the ambiguity was not a property of individual incidents that happened to be unclear. The ambiguity was the method. It had been engineered, with considerable care, specifically so that the word would be available to the person who eventually heard the complaint.
The design was in the gaps. The gaps were the design.
I opened my notebook to a new page and wrote that down.