The department meeting started at 2:04 PM, four minutes late, which was normal. Magda had been in the room since 1:58, her notes on the table, the seminar outline printed twice — one copy for herself, one in case anyone wanted to follow along, which no one would. The room's ambient-index was reading at 14.2, which meant three colleagues had already checked their personal monitors since the last sync. She had stopped watching the ambient-index display months ago. It told you what people were doing with their attention; it did not tell you whether the attention was going anywhere useful.
She had put the Adamski notepad in her bag that morning out of habit, then taken it out and left it on the kitchen table. There was no reason to bring it. Adamski did not work in her department building. The readout was Monday at the earliest. The notepad belonged in the kitchen.
At 2:07 PM, the CAS-monitor sent its calibration packet. She felt nothing; she was not wearing the sensor pad on her body. The patch was on the inside of her left forearm, under her sleeve, cycling its readings without requiring her attention. The UPMC relay node in the building's exterior wall received the packet 0.4 seconds later and routed it through the A2A-UCP pipeline. She was reviewing her notes about seminar structure reform.
The third-year seminar was item three. Items one and two — facilities maintenance and the graduate stipend adjustment — took eleven minutes combined. Tomasz made the case for the stipend increase with the same patient precision he brought to everything: current indexing formula, Q3 cost-of-living data pulled from the CityLens aggregate, three comparison cohorts. The department chair had clearly already read the brief. The vote was brief. Magda agreed, noted her agreement, moved on.
When Tomasz sat down, she stood up.
"The third-year seminar," she said, "has traditionally been structured around the oral examination component." She turned to the whiteboard — a physical one, still, because the room's overlay-projection had been tagged for maintenance since August and no one had filed the repair request, including her. "What I'm proposing is to move the oral component to week nine instead of week twelve, and use the last three weeks for synthesis work rather than preparation work."
She wrote two lines on the whiteboard: ORAL EXAM: WEEK 9 SYNTHESIS UNIT: WEEKS 10–12
The room was quiet in the way department rooms went quiet when people were reading something they had not expected to read.
"The logic," she continued, "is that students currently spend weeks ten through twelve in a kind of holding pattern. The oral examination functions as a terminus. Everything after it is administrative." She turned back to face the table. "If we move the examination earlier, weeks ten through twelve become actual intellectual work."
Irena, who taught the second-year seminar, asked the first question: "What does the synthesis unit produce?"
"A fifteen-page reflective paper. Not a new argument. A map of the student's argument over the year."
"That is a new assignment," Irena said.
"Yes," Magda agreed.
This was the nature of department meetings. Someone pointed out that a new thing was new.
At 2:31 PM, the CAS-monitor sent its half-hour calibration. The relay node in the building exterior confirmed receipt. THRESHOLD-STABLE. The data entered the A2A-UCP pipeline and moved toward the central monitoring system, which logged it with a timestamp and filed it in the running record. The attending physician's pre-appointment review file — auto-populated since Tuesday night, when the pre-imaging disclosure protocol v4.2 forwarded Monday's variance reading without requiring her to take action — received the update. The record grew by one line.
Magda was explaining the rubric. Four criteria: coherence, development, self-awareness, precision. She had brought printed copies. In the room's learning-commons version, this would have been handled through the curriculum-sync portal; in the old building with its broken overlay, she had printed eight copies on the third-floor department printer, the one that still worked.
Benedikt asked about the budget for external speakers. He was not hostile; he needed to understand financial implications before he could agree to anything. Eleven years. She knew this about him.
"No additional cost," she confirmed. "The synthesis unit uses internal resources."
He nodded and wrote something in his notepad. Not the Adamski notepad — his department notepad, green-covered, analog, which he had used since she had known him. Magda noticed the notepad and thought briefly of her own, on the kitchen table, and then Dorota was asking about the self-awareness criterion.
"It is difficult to assess," Irena had said, and this was true, which was why it was on the rubric. Irena smiled in the way she smiled when she was deciding whether to push back. She let it go.
The thing about the MRI data was that it was already in the system. This was not a reassurance; it was a fact. The relay node had confirmed Building C transmission complete at 11:22 AM the previous morning. The attending had the CAS variance readings. Adamski would call when the results were ready for interpretation. None of that changed what was written on the whiteboard. None of it changed whether weeks ten through twelve should be a synthesis unit.
She talked through the rubric for twenty minutes. Someone asked whether the CityLens cohort comparison could be adapted for the synthesis paper's self-awareness criterion — tracking student conceptual drift against a baseline established in the first seminar year. It was a good question. She wrote it down. It would require more than she could commit to in this meeting, but it was worth following up on.
The meeting ended at 3:31 PM. Ninety-one minutes. The two action items on the whiteboard: 1. DISTRIBUTE RUBRIC DRAFT — MAGDA 2. CONFIRM WEEK 9 EXAM SLOT WITH REGISTRAR — DEPT CHAIR
She photographed the whiteboard with her phone, erased it, put the cap back on the marker. People gathering, closing tabs, filing out. Tomasz said he liked the synthesis unit idea. She thanked him. Dorota asked whether the rubric would come through the curriculum-sync portal or the department listserv. Listserv, Magda said — the curriculum portal was still pending review from last term's data-governance audit, which was a different maintenance log than the overlay-projection, and she was tired of waiting on both.
The room emptied. The overhead lights, motion-triggered, stayed on for ninety seconds after the last person left. Magda was the last person to leave.
Her office was on the same floor, four doors down. She hung her coat on the hook, sat at her desk, and took two student essays from the stack on the left side. Third-year students working the same theoretical framework from different angles. She had read the opening pages last week. Now she was reading the middles.
At 3:00 PM — she noted this afterward, looking at the CAS-monitor log she checked once at 3:45 — the device had sent its hourly calibration. THRESHOLD-STABLE. She had not felt it. She had been at her desk with a red pen, writing in the margin: the claim is here but the evidence arrives too late — does paragraph 4 belong before paragraph 2?
The second student had a stronger argument in the first half and a weaker argument in the second half, which was the more interesting problem. She wrote: where did the confidence go? the opening three paragraphs know something the rest of the essay has forgotten.
She checked the CAS-monitor log once at 3:45 PM, out of habit rather than anxiety. THRESHOLD-STABLE: 9 AM, 10 AM, 11 AM, noon, 1 PM, 2 PM, 3 PM. Seven consecutive calibrations since she had come to the office. The A2A-UCP pipeline was running. The data was continuous. The attending's file was current. She closed the log and went back to the essays.
At 4:07 PM she set the second essay down. Two pages of notes; she would write the formal feedback tomorrow. She looked at the window. The light had shifted in the way it shifted in late September when the afternoon sun dropped below the angle of the building across the courtyard. Eleven years in this office. This exact angle of light at this hour. The ambient-index display in the corner of her screen read 3.1: only three active monitors in the building wing at this hour, all presumably office-bound colleagues doing their own quiet work.
She put on her coat, picked up her bag. The Adamski notepad was on the kitchen table. Three questions, seven words. Third question formed and written in Building C's changing room, nine minutes after the technician said we are finished. The question was in the notepad. The scan was in the system. Two separate things pointing at the same data from different angles.
She turned off the office light. Closed the door. Walked past the meeting room, which was dark now except for the thin line of late-afternoon light under the blinds.
Outside the building, the air was cooler than it had been at noon. She walked toward home. The relay node in her building's exterior housing unit would receive the 5 PM calibration in forty-three minutes. She would probably be home by then, sitting in the kitchen, reading something.
The action items were photographed on her phone. The essays were in the stack. The seminar structure was on the registrar's desk, waiting for a slot confirmation that would arrive in the next three to five business days, which was the same timeline as Adamski.
This was Wednesday. She walked home in the light while it lasted.
At 4:59 PM, the CAS-monitor sent its calibration. THRESHOLD-STABLE. The relay node in the housing unit exterior received the packet and routed it through the A2A-UCP pipeline. The record updated. The attending's file updated. The system logged: continuous monitoring, no flags, transmission complete, seventh consecutive stable reading since the start of the post-imaging observation window.
Magda was at the kitchen table, reading. The Adamski notepad was three inches to her left. She had not opened it. The meeting had produced two action items. She had already sent the rubric draft to the listserv. One done.