Naia arrived at the third-floor conference room seventeen minutes before nine o'clock.
This was not unusual for her. She had identified long ago the particular cost of walking into a room where other people were already settled, the way it redistributed the weight of a conversation before a word had been spoken, and she had made a habit of arriving early enough to be the one doing the settling. She set her files on the table in the order she intended to reference them. She placed her pen at the correct angle to her notepad. She moved the two chairs so that they faced each other across the table rather than sitting at adjacent corners, because the adjacent arrangement was warmer, and she did not, at this juncture, want warm.
Then she stood at the window for exactly as long as it took to notice that she was standing at the window, and sat down.
The conference room had been designed by someone who believed that serious work required plain surroundings. The table was long and unadorned. The walls were the color of old paper. The single window looked down on a narrow service courtyard where a boy in a grey apron was dragging a handcart across the flagstones, and beyond the courtyard roof the top third of the Registry Tower was visible, pale against a morning sky that had not yet decided to be blue. She had not chosen the room. She had, however, requested it, which in the Registry Wing amounted to the same thing, because the third-floor conference room was the only one without a door that opened directly onto the main corridor, and a meeting that could not be overheard was a meeting that could be conducted honestly.
She did not allow herself to examine why she had wanted that.
The order lay folded in her coat pocket. She had read it four times since yesterday morning, which was three more than its length justified, and she had spent a portion of last night working out what she knew about the man whose name it carried. The answer, assembled with the care she gave to any literature review, was this: Commander Daven Kairu, twenty-seven years old, head of the Imperial Veil Guard, raised in the northern territory of Kelvari, promoted to his present rank younger than anyone in the Guard's recorded history. He had a reputation for exactness. He was said to speak very little and to be wrong even less. Those were the facts available to anyone with access to the palace's public registers.
What she knew beyond that was not information. It was a name written on the inside of her left wrist in her own seventeen-year-old handwriting, and she had made a long and disciplined study of how to treat it as something other than an event.
She had met him at a distance before. Soleth was not a small city, but the Imperial Palace was a managed environment, and managed environments had a way of making the same people visible to one another with a regularity that did not constitute familiarity. She had seen him in corridors. She had seen him from the Wing's upper windows, crossing the central courtyard at the head of a unit drill, his stride so even that the soldiers behind him seemed to be walking to it. Once she had seen him in the formal receiving hall, standing at the edge of a court session where she had delivered testimony on a Misprint case, and she had noticed, with some irritation, that he was the only person in the room who had not looked at her notes while she spoke. He had looked at her. Then at the door. Then at her again.
She had filed that. She filed everything. It was how she remained a person who could be trusted with data.
At nine minutes to the hour, she heard footsteps in the corridor and her pen stopped moving.
They were not his. Too light, too quick, and accompanied by the particular rustle of a scholar's robe that was slightly too long for its wearer. The footsteps passed the door, hesitated, doubled back, and passed again. Naia did not look up. She waited until the third pass and said, in a voice pitched for the corridor and no further, "Lena, if you intend to loiter, do it somewhere with better acoustics."
The footsteps stopped. After a moment, Lena Cae's voice came back, wounded and entirely unrepentant. "I am delivering a message."
"What message?"
"That I am not loitering."
"Understood. Delivered."
The footsteps withdrew, much more slowly than they had arrived. Naia returned to her pen. She was aware of a small, unwelcome warmth in her face and she waited for it to pass with the same patience she would have given an unreliable instrument.
At exactly nine o'clock, the door opened.
Not fifty-nine minutes past eight. Not one minute past nine. The exactness of it was not, she suspected, an accident.
He was taller than he looked from the upper windows. She noted this immediately and with some annoyance, because she had been operating with an inaccurate data point and accuracy mattered. He wore the dark uniform of the Veil Guard without ornament, the high collar closed, the gloves tucked into his belt. His hair was cut short in the military fashion. His eyes were the grey of wet slate and they performed, in the first three seconds of his entry, a sweep of the room that took in the door, the window, the table, and her, in an order and at a speed that suggested the sweep was less a decision than a reflex. Then his attention settled on the materials spread across the table.
He looked at the files. Then he looked at her.
"Your methodology has a problem," he said.
These were the first words Commander Daven Kairu spoke to her.
She set down her pen. She had imagined, in the manner of a person who prepared for conversations the way other people prepared for storms, a number of ways this meeting might open. Most of them involved introductions. Several involved a pause in which each party took the measure of the other. None had involved a verdict.
"Good morning, Commander," she said. "I am Scholar Solven."
"I know. I read your preliminary report." He did not sit. "Good morning."
"You have read it already."
"Yesterday evening. And again at five this morning." He glanced at the top page of her folder as though checking it against memory. "There is a sequencing error in your territory mapping. You are treating Varath and Kelvari as equivalent data sources, but the Kelvari Registry offices operate on a six-week reporting delay relative to Varath's real-time submissions. Your trend line compares data from different time periods as though they were contemporaneous."
She looked at her materials. She looked at him. "That is a legitimate observation."
"It is also a significant one."
"I am aware of what it is, Commander." She drew the relevant file toward her and opened it to the page she wanted. "The delay was accounted for in the baseline calculation on page seven. The trend line uses adjusted figures."
"The adjustment on page seven is incomplete."
He had come around the table by then and was standing at her shoulder, close enough that she could see the page he was indicating but not so close that the distance was a statement. She noted that, too. He had calibrated it.
"You corrected for the submission delay," he said. "You did not correct for the collection method. Kelvari offices take passive reports. A citizen walks in, mentions that their wrist has changed, and a clerk writes it down. Varath conducts active surveys. Registry staff go door to door and ask. You are not merely comparing different weeks. You are comparing different recording intensities. The north looks quieter than it is because fewer people have been asked."
Naia opened page seven. She read it. She closed the file.
She had been correct that the delay was accounted for. He had been correct that the collection method was not. Both of these things were true at the same time, and both of them knew it, and the silence in which they both knew it lasted perhaps two seconds, which was a long time for a silence between strangers.
"The correction would reduce the apparent acceleration rate in the northern territories by approximately twelve percent," she said.
"Based on active-versus-passive comparisons in the military census records, closer to fifteen."
"Twelve to fifteen is a range, not a conflict. The trend holds regardless."
"The trend holds. The magnitude of the trend is less certain than your report implies."
This was, she acknowledged privately, accurate. She disliked how accurate it was. She disliked more that he had said it without any of the small cushioning gestures that most people offered when correcting a senior scholar, the half-smile, the "if I may." He had said it the way a surveyor read a measurement.
"Please sit," she said.
He sat, across from her, in the chair she had positioned there. She had the odd sensation of having arranged the furniture for an opponent who had immediately recognized the arrangement and approved of it.
She pulled her notepad toward her and began to write. "The distinction matters for resource allocation. If I have overstated the acceleration in the north, the court will concentrate its response in the wrong place." She crossed out a figure. Rewrote it with the adjustment. Crossed out a second figure. "What is your source for fifteen percent?"
"Imperial census data from the last three cycles. I can have my lieutenant pull the comparison tables if you wish to verify the method."
"I do. Have him send them to the Wing by this afternoon."
"By noon."
"Noon, then."
He wrote something in a small notebook of his own. She watched his pen move and thought that it moved the way he spoke, in short strokes, few words, nothing wasted. It was the hand of someone who wrote down only what he intended to keep.
"Your account of the causal mechanism is also underdeveloped," he said.
She looked up.
"You document the Fading at length. You attribute it to an unspecified Registry instability. You do not propose a cause. For the court, cause matters as much as effect, because the intervention depends on what you are trying to repair."
"The cause requires further investigation. Publishing a speculative mechanism before the data supports it is not responsible scholarship."
"Publishing no mechanism means the court will supply one. Probably incorrectly. Probably in a direction that serves an existing interest."
She opened her mouth to answer, and found that the answer she had prepared was not the answer that was true.
It was a thing she had learned in eight years of submitting reports to the Imperial Court. A vacuum left by scholarly caution did not remain a vacuum. It was filled, quickly and by whoever was nearest, with political interpretation. She had watched it happen to the Misprint reforms, and again to the survey on regional disparities, and in each case she had told herself that the failure belonged to the court. It had not occurred to her, until a stranger said it across a table at nine in the morning, that her silence had been part of the mechanism.
She had left it out of her own report because including it would have meant admitting that the scholarly method had a structural weakness, and structural weaknesses in one's own method were not comfortable things to document.
"What would you recommend?" she asked.
He regarded her with what she would later learn to recognize as surprise, though at the time it registered only as a change in the stillness of his face. "Present the most defensible mechanism you have. Mark it clearly as preliminary. State what evidence would confirm it and what would disprove it. Give the court something to work with that you control, so they are not tempted to reach for something you do not."
"That is a military answer to an academic problem."
"Yes."
"It is also not wrong."
"I try to avoid that," he said, and there was no smile in it, but something in the line of his shoulders eased by a single degree, and she found that she had noticed it, and that she did not like having noticed.
Outside the window, the boy with the handcart had given up on the flagstones and was sitting on the edge of it, eating something from a paper twist. Beyond him, the Tower held its pale gold against the morning. Naia looked at it for a moment, the way she looked at it every day, as a reading to be taken, and then back at the table.
They went through the report section by section.
It was not a comfortable process. He had read more carefully than she had expected anyone to read, and he read in a way she recognized because she did it herself: he did not concede points he did not believe were conceded, and he did not dispute points merely for the pleasure of disputing them. When she was right, he moved on without comment. When he was right, she adjusted. When the truth lay somewhere between their two positions, which was most of the time, they stayed on the problem until they found where it was.
She argued for the classification scheme she had used for the southern districts. He pointed out that two of her categories overlapped, and that the overlap would inflate the count in precisely the districts she most wanted to emphasize. She argued that the overlap was deliberate. He asked why. She explained, at some length, that people who had lost a Bond often continued to describe themselves as bonded for weeks afterward, and that she had wanted to capture the lag between the event and the acceptance of it. He listened to all of it. Then he said, "That is not an overlap. That is two variables. You should separate them and report both."
She stared at him.
"The lag between the event and the acceptance is itself a finding," he said. "If you fold it into the count, you hide it."
She wrote it down. She did not say thank you. She was aware that her pen was moving slightly faster than before.
At the half hour, they reached the Fading timeline, which she had built with particular care and expected to defend. She had charted the weekly counts, fitted three regression models to them, and shown that the acceleration was robust across all three. It was the strongest page in the report. She turned it toward him with the faint, private satisfaction of a person laying down a good hand.
He read it for some time. Then he said, "How are the dates assigned?"
"By date of report."
"Date of report to a Registry office."
"Yes."
"Not the date the mark faded."
She hesitated. "The date of fading is not recorded in most cases. People do not know the hour it happened. They notice, they come in, and the clerk enters the date they came in."
"Then your weekly counts are not counts of fadings. They are counts of the days people could get to a Registry office." He turned the page toward her again and laid one finger beside the lowest point of her chart, which fell every seventh week as regularly as a tide. "The offices open on weekdays. The Guard assigns men to the Registry queues when they run long. In the last six weeks I have read eleven reports of queues stretching around the block on a Monday morning and none at all on a Sunday. You have fitted a smooth curve to the rhythm of an institution."
She did not move.
Three of her colleagues had reviewed this chart. A senior scholar from the eastern desk had called it exemplary. She had run the three models herself, with great care, on the assumption that the dates were sound, and not one of them had asked what a date was.
It was a very small silence. Half a second, perhaps. A real pause, the kind she did not make in front of other people, in which she looked at her own best work and saw, for the first time, the shape of the thing it was standing on.
Then she reached for the problem, because that was what she did. "If the reports lag the events by a variable interval, then the true acceleration began earlier than the chart shows, and the curve is steeper than I have claimed."
"Probably."
"Which means the court has had less time than I told them."
"Yes."
"I would need the clerk intake logs. They record the hour of each entry. Some of them record what the citizen said about when it happened, in the margin."
"I can have access arranged by tomorrow."
"Tomorrow is acceptable." She was already writing. "That is a significant flaw, Commander."
"It is a correctable one."
"It is a flaw that three other scholars missed."
"They were reading the chart," he said. "I was reading the queues."
It was the longest sentence he had offered without being asked, and she turned it over for a moment, as one turns over a coin that has turned out to be of an unfamiliar mint. A man who read queues. A man who stood in a corridor and counted the people waiting outside a Registry office, not because he had been told to, but because a thing was happening to the empire and he was the sort of person who looked at what was in front of him.
She set the thought aside. It did not go as far as she intended.
Out in the corridor, there was a faint sound, a creak of old floorboards that did not belong to either of them. She glanced toward the door. It was closed. She glanced toward the window, which gave onto the courtyard and not the hall. After a moment, she let it go, because there was a lot of work to do and the sound was not part of it.
They settled the remaining sections more quickly. The framework they agreed on was a compromise, but not the kind that satisfied no one. She would revise the northern figures with his correction. She would separate the lag variable and report it independently. She would add a preliminary mechanism, flagged and caveated and bounded, in the form he had described. And she would rebuild the timeline from the clerk intake logs and issue it as a supplement within the week. It was a better report than the one she had brought into the room. It was also, she suspected, a better one than she would have produced alone, and the suspicion was not entirely welcome.
When the last revision had been recorded, he closed his notebook and stood. She stood as well, since it seemed the thing to do.
"The Varath survey data will be more reliable than the Soleth archives for the active comparison," she said, gathering the pages. "If the southern acceleration matches the revised northern figures, we will have a cleaner picture of the real distribution."
"We may need to go to Varath directly."
"The regional scholars are due to send their preliminary counts. I would want to review those before committing to field work."
"Reasonable." He paused. "Scholar Solven."
"Yes?"
He seemed to consider whether to say the next thing, and for a moment she saw something cross his face, quick and unreadable, the look of a man who had put his hand on a door and decided not to push it.
"Your handwriting is very clear," he said. "It made the report easy to read."
She blinked. "Thank you."
He nodded once, and went out.
The door closed behind him. Naia remained where she was, her hands full of paper, listening to his footsteps recede down the corridor at the same even pace with which he had approached it. They did not hurry. They did not slow. They were the footsteps of a man who had, as far as she could tell, no idea that he had just said the first personal thing she had heard him say.
Her handwriting was very clear.
She looked down at the notepad, at the neat, upright lines of her own script, the cross-outs, the revised figures, and then, involuntarily, at her left wrist, where the bandage lay flat and white against her skin. She had spent a very long time making sure that no one would ever ask about what was written under it. She had spent seven years arranging her life so that the handwriting on her wrist would never be compared to anything.
It was a coincidence. A remark about legibility, offered by a man who read reports for a living. She told herself this once, clearly, and set it in the category where such things belonged.
She sat down again. She began transferring the revision notes into the formal structure of the report. Her pen was steady. Her handwriting was even and clear, and she noticed that she was noticing it, and she disciplined her attention back to the page.
It was a full minute before she realized that she had been listening, in the quiet of the empty room, for the sound of him coming back.
He did not come back. What came instead, after a decent interval, was a soft tap on the door and Lena Cae's face around the edge of it, composed into an expression of such elaborate neutrality that it had circled back around to being a statement.
"Well?" said Lena.
"Well what?"
"How was it?"
"Difficult." Naia gathered the last of her files. "He found a flaw in the timeline."
"A flaw in your timeline." Lena came in and shut the door, her robe trailing. "Someone found a flaw in your timeline. In my three years in this Wing I have never heard those words in that order."
"It was a good catch."
"You said that without being forced. Naia, are you feeling unwell?"
"I am feeling occupied. There is a revised report to produce and a set of clerk logs to retrieve." She stood. "And I would be grateful if you would stop standing outside doors."
"I was not standing outside the door. I was standing near the door. In the corridor. For the purposes of availability." Lena studied her with the frank, bright attention that had made her indispensable and occasionally unbearable. "You know your face does not look like someone who just had a difficult time."
"My face is professionally neutral."
"Your face is doing something I do not have a word for." Lena tilted her head. "It looks like the face you make when you have been handed an extremely interesting problem."
Naia went past her into the corridor. The Wing was waking up around them, scholars drifting between the offices with their arms full of folded paper, the brass lamps along the ceiling not yet needed in the strengthening light. A door opened and closed somewhere. A clerk laughed too loudly at something. Normal sounds, the sounds of a building going about its business, and she walked into them gratefully.
"It is a problem," she said, over her shoulder. "That is all it is."
"Of course it is," said Lena, falling into step beside her with a smile that she was kind enough to keep to herself.
They walked together toward the stairs. Naia did not answer, and Lena did not press, and neither of them mentioned the fact that Naia had said nothing about the one thing the Commander had said as he left.
Her handwriting was very clear.
She kept her left hand at her side as she walked, the bandage pressed lightly against the fabric of her sleeve, and she did not look at it again until she was alone.
You've reached the end of this chapter
Naia arrived at the third-floor conference room seventeen minutes before nine o'clock.
This was not unusual for her. She had identified long ago the particular cost of walking into a room where other people were already settled, the way it redistributed the weight of a conversation before a word had been spoken, and she had made a habit of arriving early enough to be the one doing the settling. She set her files on the table in the order she intended to reference them. She placed her pen at the correct angle to her notepad. She moved the two chairs so that they faced each other across the table rather than sitting at adjacent corners, because the adjacent arrangement was warmer, and she did not, at this juncture, want warm.
Then she stood at the window for exactly as long as it took to notice that she was standing at the window, and sat down.
The conference room had been designed by someone who believed that serious work required plain surroundings. The table was long and unadorned. The walls were the color of old paper. The single window looked down on a narrow service courtyard where a boy in a grey apron was dragging a handcart across the flagstones, and beyond the courtyard roof the top third of the Registry Tower was visible, pale against a morning sky that had not yet decided to be blue. She had not chosen the room. She had, however, requested it, which in the Registry Wing amounted to the same thing, because the third-floor conference room was the only one without a door that opened directly onto the main corridor, and a meeting that could not be overheard was a meeting that could be conducted honestly.
She did not allow herself to examine why she had wanted that.
The order lay folded in her coat pocket. She had read it four times since yesterday morning, which was three more than its length justified, and she had spent a portion of last night working out what she knew about the man whose name it carried. The answer, assembled with the care she gave to any literature review, was this: Commander Daven Kairu, twenty-seven years old, head of the Imperial Veil Guard, raised in the northern territory of Kelvari, promoted to his present rank younger than anyone in the Guard's recorded history. He had a reputation for exactness. He was said to speak very little and to be wrong even less. Those were the facts available to anyone with access to the palace's public registers.
What she knew beyond that was not information. It was a name written on the inside of her left wrist in her own seventeen-year-old handwriting, and she had made a long and disciplined study of how to treat it as something other than an event.
She had met him at a distance before. Soleth was not a small city, but the Imperial Palace was a managed environment, and managed environments had a way of making the same people visible to one another with a regularity that did not constitute familiarity. She had seen him in corridors. She had seen him from the Wing's upper windows, crossing the central courtyard at the head of a unit drill, his stride so even that the soldiers behind him seemed to be walking to it. Once she had seen him in the formal receiving hall, standing at the edge of a court session where she had delivered testimony on a Misprint case, and she had noticed, with some irritation, that he was the only person in the room who had not looked at her notes while she spoke. He had looked at her. Then at the door. Then at her again.
She had filed that. She filed everything. It was how she remained a person who could be trusted with data.
At nine minutes to the hour, she heard footsteps in the corridor and her pen stopped moving.
They were not his. Too light, too quick, and accompanied by the particular rustle of a scholar's robe that was slightly too long for its wearer. The footsteps passed the door, hesitated, doubled back, and passed again. Naia did not look up. She waited until the third pass and said, in a voice pitched for the corridor and no further, "Lena, if you intend to loiter, do it somewhere with better acoustics."
The footsteps stopped. After a moment, Lena Cae's voice came back, wounded and entirely unrepentant. "I am delivering a message."
"What message?"
"That I am not loitering."
"Understood. Delivered."
The footsteps withdrew, much more slowly than they had arrived. Naia returned to her pen. She was aware of a small, unwelcome warmth in her face and she waited for it to pass with the same patience she would have given an unreliable instrument.
At exactly nine o'clock, the door opened.
Not fifty-nine minutes past eight. Not one minute past nine. The exactness of it was not, she suspected, an accident.
He was taller than he looked from the upper windows. She noted this immediately and with some annoyance, because she had been operating with an inaccurate data point and accuracy mattered. He wore the dark uniform of the Veil Guard without ornament, the high collar closed, the gloves tucked into his belt. His hair was cut short in the military fashion. His eyes were the grey of wet slate and they performed, in the first three seconds of his entry, a sweep of the room that took in the door, the window, the table, and her, in an order and at a speed that suggested the sweep was less a decision than a reflex. Then his attention settled on the materials spread across the table.
He looked at the files. Then he looked at her.
"Your methodology has a problem," he said.
These were the first words Commander Daven Kairu spoke to her.
She set down her pen. She had imagined, in the manner of a person who prepared for conversations the way other people prepared for storms, a number of ways this meeting might open. Most of them involved introductions. Several involved a pause in which each party took the measure of the other. None had involved a verdict.
"Good morning, Commander," she said. "I am Scholar Solven."
"I know. I read your preliminary report." He did not sit. "Good morning."
"You have read it already."
"Yesterday evening. And again at five this morning." He glanced at the top page of her folder as though checking it against memory. "There is a sequencing error in your territory mapping. You are treating Varath and Kelvari as equivalent data sources, but the Kelvari Registry offices operate on a six-week reporting delay relative to Varath's real-time submissions. Your trend line compares data from different time periods as though they were contemporaneous."
She looked at her materials. She looked at him. "That is a legitimate observation."
"It is also a significant one."
"I am aware of what it is, Commander." She drew the relevant file toward her and opened it to the page she wanted. "The delay was accounted for in the baseline calculation on page seven. The trend line uses adjusted figures."
"The adjustment on page seven is incomplete."
He had come around the table by then and was standing at her shoulder, close enough that she could see the page he was indicating but not so close that the distance was a statement. She noted that, too. He had calibrated it.
"You corrected for the submission delay," he said. "You did not correct for the collection method. Kelvari offices take passive reports. A citizen walks in, mentions that their wrist has changed, and a clerk writes it down. Varath conducts active surveys. Registry staff go door to door and ask. You are not merely comparing different weeks. You are comparing different recording intensities. The north looks quieter than it is because fewer people have been asked."
Naia opened page seven. She read it. She closed the file.
She had been correct that the delay was accounted for. He had been correct that the collection method was not. Both of these things were true at the same time, and both of them knew it, and the silence in which they both knew it lasted perhaps two seconds, which was a long time for a silence between strangers.
"The correction would reduce the apparent acceleration rate in the northern territories by approximately twelve percent," she said.
"Based on active-versus-passive comparisons in the military census records, closer to fifteen."
"Twelve to fifteen is a range, not a conflict. The trend holds regardless."
"The trend holds. The magnitude of the trend is less certain than your report implies."
This was, she acknowledged privately, accurate. She disliked how accurate it was. She disliked more that he had said it without any of the small cushioning gestures that most people offered when correcting a senior scholar, the half-smile, the "if I may." He had said it the way a surveyor read a measurement.
"Please sit," she said.
He sat, across from her, in the chair she had positioned there. She had the odd sensation of having arranged the furniture for an opponent who had immediately recognized the arrangement and approved of it.
She pulled her notepad toward her and began to write. "The distinction matters for resource allocation. If I have overstated the acceleration in the north, the court will concentrate its response in the wrong place." She crossed out a figure. Rewrote it with the adjustment. Crossed out a second figure. "What is your source for fifteen percent?"
"Imperial census data from the last three cycles. I can have my lieutenant pull the comparison tables if you wish to verify the method."
"I do. Have him send them to the Wing by this afternoon."
"By noon."
"Noon, then."
He wrote something in a small notebook of his own. She watched his pen move and thought that it moved the way he spoke, in short strokes, few words, nothing wasted. It was the hand of someone who wrote down only what he intended to keep.
"Your account of the causal mechanism is also underdeveloped," he said.
She looked up.
"You document the Fading at length. You attribute it to an unspecified Registry instability. You do not propose a cause. For the court, cause matters as much as effect, because the intervention depends on what you are trying to repair."
"The cause requires further investigation. Publishing a speculative mechanism before the data supports it is not responsible scholarship."
"Publishing no mechanism means the court will supply one. Probably incorrectly. Probably in a direction that serves an existing interest."
She opened her mouth to answer, and found that the answer she had prepared was not the answer that was true.
It was a thing she had learned in eight years of submitting reports to the Imperial Court. A vacuum left by scholarly caution did not remain a vacuum. It was filled, quickly and by whoever was nearest, with political interpretation. She had watched it happen to the Misprint reforms, and again to the survey on regional disparities, and in each case she had told herself that the failure belonged to the court. It had not occurred to her, until a stranger said it across a table at nine in the morning, that her silence had been part of the mechanism.
She had left it out of her own report because including it would have meant admitting that the scholarly method had a structural weakness, and structural weaknesses in one's own method were not comfortable things to document.
"What would you recommend?" she asked.
He regarded her with what she would later learn to recognize as surprise, though at the time it registered only as a change in the stillness of his face. "Present the most defensible mechanism you have. Mark it clearly as preliminary. State what evidence would confirm it and what would disprove it. Give the court something to work with that you control, so they are not tempted to reach for something you do not."
"That is a military answer to an academic problem."
"Yes."
"It is also not wrong."
"I try to avoid that," he said, and there was no smile in it, but something in the line of his shoulders eased by a single degree, and she found that she had noticed it, and that she did not like having noticed.
Outside the window, the boy with the handcart had given up on the flagstones and was sitting on the edge of it, eating something from a paper twist. Beyond him, the Tower held its pale gold against the morning. Naia looked at it for a moment, the way she looked at it every day, as a reading to be taken, and then back at the table.
They went through the report section by section.
It was not a comfortable process. He had read more carefully than she had expected anyone to read, and he read in a way she recognized because she did it herself: he did not concede points he did not believe were conceded, and he did not dispute points merely for the pleasure of disputing them. When she was right, he moved on without comment. When he was right, she adjusted. When the truth lay somewhere between their two positions, which was most of the time, they stayed on the problem until they found where it was.
She argued for the classification scheme she had used for the southern districts. He pointed out that two of her categories overlapped, and that the overlap would inflate the count in precisely the districts she most wanted to emphasize. She argued that the overlap was deliberate. He asked why. She explained, at some length, that people who had lost a Bond often continued to describe themselves as bonded for weeks afterward, and that she had wanted to capture the lag between the event and the acceptance of it. He listened to all of it. Then he said, "That is not an overlap. That is two variables. You should separate them and report both."
She stared at him.
"The lag between the event and the acceptance is itself a finding," he said. "If you fold it into the count, you hide it."
She wrote it down. She did not say thank you. She was aware that her pen was moving slightly faster than before.
At the half hour, they reached the Fading timeline, which she had built with particular care and expected to defend. She had charted the weekly counts, fitted three regression models to them, and shown that the acceleration was robust across all three. It was the strongest page in the report. She turned it toward him with the faint, private satisfaction of a person laying down a good hand.
He read it for some time. Then he said, "How are the dates assigned?"
"By date of report."
"Date of report to a Registry office."
"Yes."
"Not the date the mark faded."
She hesitated. "The date of fading is not recorded in most cases. People do not know the hour it happened. They notice, they come in, and the clerk enters the date they came in."
"Then your weekly counts are not counts of fadings. They are counts of the days people could get to a Registry office." He turned the page toward her again and laid one finger beside the lowest point of her chart, which fell every seventh week as regularly as a tide. "The offices open on weekdays. The Guard assigns men to the Registry queues when they run long. In the last six weeks I have read eleven reports of queues stretching around the block on a Monday morning and none at all on a Sunday. You have fitted a smooth curve to the rhythm of an institution."
She did not move.
Three of her colleagues had reviewed this chart. A senior scholar from the eastern desk had called it exemplary. She had run the three models herself, with great care, on the assumption that the dates were sound, and not one of them had asked what a date was.
It was a very small silence. Half a second, perhaps. A real pause, the kind she did not make in front of other people, in which she looked at her own best work and saw, for the first time, the shape of the thing it was standing on.
Then she reached for the problem, because that was what she did. "If the reports lag the events by a variable interval, then the true acceleration began earlier than the chart shows, and the curve is steeper than I have claimed."
"Probably."
"Which means the court has had less time than I told them."
"Yes."
"I would need the clerk intake logs. They record the hour of each entry. Some of them record what the citizen said about when it happened, in the margin."
"I can have access arranged by tomorrow."
"Tomorrow is acceptable." She was already writing. "That is a significant flaw, Commander."
"It is a correctable one."
"It is a flaw that three other scholars missed."
"They were reading the chart," he said. "I was reading the queues."
It was the longest sentence he had offered without being asked, and she turned it over for a moment, as one turns over a coin that has turned out to be of an unfamiliar mint. A man who read queues. A man who stood in a corridor and counted the people waiting outside a Registry office, not because he had been told to, but because a thing was happening to the empire and he was the sort of person who looked at what was in front of him.
She set the thought aside. It did not go as far as she intended.
Out in the corridor, there was a faint sound, a creak of old floorboards that did not belong to either of them. She glanced toward the door. It was closed. She glanced toward the window, which gave onto the courtyard and not the hall. After a moment, she let it go, because there was a lot of work to do and the sound was not part of it.
They settled the remaining sections more quickly. The framework they agreed on was a compromise, but not the kind that satisfied no one. She would revise the northern figures with his correction. She would separate the lag variable and report it independently. She would add a preliminary mechanism, flagged and caveated and bounded, in the form he had described. And she would rebuild the timeline from the clerk intake logs and issue it as a supplement within the week. It was a better report than the one she had brought into the room. It was also, she suspected, a better one than she would have produced alone, and the suspicion was not entirely welcome.
When the last revision had been recorded, he closed his notebook and stood. She stood as well, since it seemed the thing to do.
"The Varath survey data will be more reliable than the Soleth archives for the active comparison," she said, gathering the pages. "If the southern acceleration matches the revised northern figures, we will have a cleaner picture of the real distribution."
"We may need to go to Varath directly."
"The regional scholars are due to send their preliminary counts. I would want to review those before committing to field work."
"Reasonable." He paused. "Scholar Solven."
"Yes?"
He seemed to consider whether to say the next thing, and for a moment she saw something cross his face, quick and unreadable, the look of a man who had put his hand on a door and decided not to push it.
"Your handwriting is very clear," he said. "It made the report easy to read."
She blinked. "Thank you."
He nodded once, and went out.
The door closed behind him. Naia remained where she was, her hands full of paper, listening to his footsteps recede down the corridor at the same even pace with which he had approached it. They did not hurry. They did not slow. They were the footsteps of a man who had, as far as she could tell, no idea that he had just said the first personal thing she had heard him say.
Her handwriting was very clear.
She looked down at the notepad, at the neat, upright lines of her own script, the cross-outs, the revised figures, and then, involuntarily, at her left wrist, where the bandage lay flat and white against her skin. She had spent a very long time making sure that no one would ever ask about what was written under it. She had spent seven years arranging her life so that the handwriting on her wrist would never be compared to anything.
It was a coincidence. A remark about legibility, offered by a man who read reports for a living. She told herself this once, clearly, and set it in the category where such things belonged.
She sat down again. She began transferring the revision notes into the formal structure of the report. Her pen was steady. Her handwriting was even and clear, and she noticed that she was noticing it, and she disciplined her attention back to the page.
It was a full minute before she realized that she had been listening, in the quiet of the empty room, for the sound of him coming back.
He did not come back. What came instead, after a decent interval, was a soft tap on the door and Lena Cae's face around the edge of it, composed into an expression of such elaborate neutrality that it had circled back around to being a statement.
"Well?" said Lena.
"Well what?"
"How was it?"
"Difficult." Naia gathered the last of her files. "He found a flaw in the timeline."
"A flaw in your timeline." Lena came in and shut the door, her robe trailing. "Someone found a flaw in your timeline. In my three years in this Wing I have never heard those words in that order."
"It was a good catch."
"You said that without being forced. Naia, are you feeling unwell?"
"I am feeling occupied. There is a revised report to produce and a set of clerk logs to retrieve." She stood. "And I would be grateful if you would stop standing outside doors."
"I was not standing outside the door. I was standing near the door. In the corridor. For the purposes of availability." Lena studied her with the frank, bright attention that had made her indispensable and occasionally unbearable. "You know your face does not look like someone who just had a difficult time."
"My face is professionally neutral."
"Your face is doing something I do not have a word for." Lena tilted her head. "It looks like the face you make when you have been handed an extremely interesting problem."
Naia went past her into the corridor. The Wing was waking up around them, scholars drifting between the offices with their arms full of folded paper, the brass lamps along the ceiling not yet needed in the strengthening light. A door opened and closed somewhere. A clerk laughed too loudly at something. Normal sounds, the sounds of a building going about its business, and she walked into them gratefully.
"It is a problem," she said, over her shoulder. "That is all it is."
"Of course it is," said Lena, falling into step beside her with a smile that she was kind enough to keep to herself.
They walked together toward the stairs. Naia did not answer, and Lena did not press, and neither of them mentioned the fact that Naia had said nothing about the one thing the Commander had said as he left.
Her handwriting was very clear.
She kept her left hand at her side as she walked, the bandage pressed lightly against the fabric of her sleeve, and she did not look at it again until she was alone.
You've reached the end of this chapter