The documentation cleared in forty-one hours.

  Kael had expected forty-eight.

  That was the stated processing window.

  Forty-one meant one of three things: someone had moved the request early, the file had been expedited, or the Academy's published timeline was a ceiling rather than a prediction. He had assumed the timeline was accurate on average, with variance in both directions depending on the current load on the administrative office and the priority of the specific file.

  The documentation for his enrollment variance wasn't a high-priority file.

  Forty-one hours meant someone had given it attention it wasn't supposed to receive.

  He updated the number in his notebook.

  On the morning of the tenth day, the second combat rotation was posted outside the administrative wing.

  His name was on it.

  Time. Location. Assessment format.

  Confirmed.

  He stood at the board for ten seconds and looked at the names around his. Aldis was on the list. Third rotation. The roster had been compiled through the standard randomization process, which he had observed on two separate occasions to confirm it was genuinely random and not subject to influence.

  It appeared to be genuinely random.

  The composition of the rotation was still information.

  Then he went to find the letter.

  ?

  The training hall manager's name was Corven.

  His office sat on the north side of the practice grounds, beside the calibration room where the assessment instruments were stored and maintained.

  Kael had spoken to him once before, the day after the training session. Corven had confirmed a letter had been sent. He had not confirmed who it was addressed to, which was his right—the content of staff correspondence was not a student matter.

  Today Kael did not return to Corven.

  He went to the records annex.

  The annex kept a log of all formal correspondence sent from Academy staff to the examination board, its committees, and its standing members. Students were allowed access under subsection eleven—a regulation added sixteen years ago, after the board had been accused of processing disciplinary actions without documentation. The regulation had been designed to ensure transparency in the formal disciplinary process.

  The current board, apparently, had never bothered closing the gap.

  Kael requested the log.

  The clerk had to look up the subsection.

  He found it. Processed the request.

  Kael read the relevant period in eight minutes.

  Corven's letter had been addressed to the Committee for Academic Standards and Integrity.

  Three members.

  Two names meant nothing to him.

  The third did.

  Instructor Maren Solt.

  Advanced Theory Division.

  Fourth floor of the academic wing.

  ?

  He had heard the name before.

  Not directly. The way first-year students heard certain names—attached to upper curriculum, advanced access, quiet reputation. The kind of reputation that arrived not through active promotion but through the way other instructors' voices shifted slightly when the name came up, the way administrative staff answered direct questions with slightly different words than they used for other names.

  A name worth treating carefully.

  Now he looked properly.

  The faculty research index was public.

  He spent forty minutes with it.

  Twelve papers across nineteen years.

  The first four were standard. Careful. Foundational. Safe. The kind of work that built a reputation without committing to any position that could become a liability. Thorough citations, conservative conclusions, contributions that other researchers could build on without the original author being associated with anything controversial.

  The fifth had been published in the same year as Aldea's paper on the upper limits of Stability measurement.

  The sixth cited Aldea.

  Not in agreement.

  The language was academic. Careful. Restrained.

  The argument was not.

  Aldea's questions about Stability ceilings, Solt wrote, came from a misreading of calibration data. The dissipation assumption remained the simplest explanation for practitioner limits. And theories that challenged that assumption risked damaging the Academy's ranking system—by implying that what students were being measured against was not the actual boundary of human capability, but an administrative convenience.

  The conclusion read:

  *Such frameworks, however intellectually interesting, serve no productive function in an institutional curriculum.*

  Aldea stopped publishing theory the following year.

  Kael sat with that sequence for a moment.

  Not: *you are wrong.* But: *being right here is not useful, and pursuing it will be treated as institutional disruption.*

  You could dispute *you are wrong.* You could argue evidence, methodology, the interpretation of data. You could build a case and present it.

  You could not dispute *this line of inquiry is not permitted in this institution* with argument alone.

  You could only dispute it with evidence that couldn't be dismissed.

  He looked at the research index for another few seconds.

  Solt had been on the Committee for Academic Standards and Integrity for eleven years. The Committee reviewed flagged student files. It processed formal requests from administrative staff. It made recommendations—formal and informal—about student standing.

  Informal recommendations left no record.

  If you discover this narrative on Amazon, be aware that it has been stolen. Please report the violation.

  No record meant no appeal.

  He understood the architecture.

  ?

  Kael closed the index and sat still for ninety seconds.

  He was building a map.

  Not a conclusion.

  Not yet.

  What he had:

  Nineteen years ago, Solt publicly rejected Aldea's theoretical line.

  Now Solt sat on a committee that had received a letter about a first-year student raising the same questions in class.

  That student's file had been flagged.

  The result had been exclusion from an assessed practical session.

  Whether those things were connected in the way they appeared to be connected—or whether the appearance was a coincidence he was weighting too heavily because it was the most coherent pattern available—he could not determine yet.

  He could determine one thing.

  If he intended to keep working on the dissipation assumption, he could not do it in public.

  One notation was already on his file. A second would compound in ways that were not recoverable within the current term.

  He had the intermediate texts. He had the annex. He had his notebook.

  He did not need to make an argument in a public room to do the work.

  He returned the research index.

  Then he went to Zone Three.

  ?

  The second combat rotation was in two days.

  He had one full session before then.

  Zone Three was empty at fourth bell. Always.

  He had used it four times since finding the differential points. Twice for the efficiency drill. Once to map saturation timing. Once to test whether the three anomalies interacted under simultaneous load.

  They did. Slightly.

  The overlap between the first and second points produced a fourth zone at the midpoint—where field bleed compounded and raised saturation another four percent above corridor average.

  Today he would use it.

  He positioned himself at the midpoint and loaded to seventy percent.

  The differential appeared immediately.

  The slight increase in ambient density.

  Stable.

  He shifted laterally until he found the center of the overlap zone—the narrow half-meter window where the two anomalies combined.

  The saturation rose.

  Not by much.

  Enough.

  He loaded to eighty-five percent and began the drill.

  ?

  He completed four cycles before he noticed the variance.

  Not in the field.

  The field was clean.

  The variance was in the reload.

  On the fifth cycle, the transition carried a fractional hesitation that had not been there before. Not instability. Not the threshold effect at ninety-six percent.

  Something smaller.

  Finer.

  He stopped after cycle four and sat on the corridor floor.

  Slow internal check.

  Capacity intact. Affinity stable. Second presence inert at this load.

  Expected.

  The fracture, however, was clearer here than it had been yesterday.

  Not worse.

  Just clearer.

  Zone Three did not amplify it.

  Zone Three removed noise.

  Lower ambient saturation meant less interference.

  The fracture was easier to hear.

  He opened the notebook.

  *Zone Three. 85%. Cycle 4. Fracture perceptibility increased.*

  A pause.

  *Not a function of load. Function of environment.*

  *Lower ambient saturation removes masking effect.*

  *Fracture presents as hairline inconsistency in reload texture.*

  He looked at the page.

  *Cannot confirm non-progression from two data points. Suspend judgment. Measure over fourteen days.*

  He finished cycles five and six.

  The variance remained the same.

  No worse. No better.

  Stable.

  Information—even if not the information he wanted.

  ?

  When the drill ended, he stayed seated for a few more minutes.

  At rest, the fracture was still there.

  Stable. Precise. His.

  He would learn it the same way he learned everything else.

  By measurement.

  Not interpretation.

  He would not test maximum load before the assessment. He would not stress the fracture without a longer baseline.

  For the second combat rotation, he would use what he already had.

  Zone Three's efficiency gains. Cleaner technique. Ten days of mapped parameters.

  He closed the notebook and stood.

  Then left the corridor and walked back toward main campus.

  ?

  Sirin was at the courtyard study tables.

  Three texts open. Papers arranged. Settled in.

  She looked up when he entered the courtyard, tracked his path, and decided before he arrived.

  Kael sat across from her.

  She did not begin with the problem in front of her.

  "Solt is on the oversight committee," she said.

  Kael met her eyes.

  "I know."

  A small shift in her expression.

  A recalibration.

  "I found it this morning," she said. "After I saw the notation on your file."

  He was quiet.

  She had looked into it on her own. Not casually. Deliberately.

  "The oversight committee reviews flagged student files," she said. "They can recommend holds on assessed activity."

  "I know that too."

  She inclined her head once, accepting the correction without resistance.

  "The flag on your file was informal," she said. "It doesn't appear in the committee's formal record."

  No formal record meant no appeal.

  "How do you know?" Kael asked.

  "I asked the clerk."

  Then:

  "She told me the formal record didn't include it. I asked why. She said informal requests were processed at member discretion. I asked what discretion meant. She stopped answering."

  Kael looked at her.

  For ten days he had been working alone. Running drills in empty corridors. Collecting data. Building maps.

  He had treated Sirin as resolved—the contact she'd given him had been used, the research proposal was in process, the exchange had been useful and contained and categorized.

  Useful. Capable. Parallel.

  An ally at one degree of separation, working on adjacent problems that occasionally intersected with his.

  He had not expected her to be running her own line of inquiry at the same speed.

  He adjusted the model.

  Not dramatically. Just precisely—the way you adjusted a calculation when a variable you'd held constant turned out to be dynamic. Sirin Vael was not working adjacent to his investigation. She was running the same investigation from a different starting point, using different access pathways, arriving at the same coordinates.

  The difference between an ally and a partner was whether they would have arrived at the same place without being directed there.

  She had arrived.

  "The dissipation assumption," she said. "You were thinking about it before you came here."

  "Since before I arrived."

  "How far back?"

  "Three years."

  A pause.

  She looked at him with the expression she had at the study tables, ten days ago—the specific look of someone revising a prior estimate upward.

  "The notation on your file was placed two days after the training hall session," she said. "I checked the timestamps."

  He had checked them too.

  "The notation says pending review," she said. "The committee hasn't acted formally. But the flag is on the file, which means anyone on the committee can see it and respond informally." A pause. "Informally means without documentation. Without documentation means without an appeal pathway."

  "Solt's paper," he said. "The rebuttal."

  Sirin nodded.

  "The argument was technically competent," Kael said. "But it didn't answer the question."

  "No," Sirin said. "It answered a different one."

  He thought of the conclusion line.

  *No productive function in an institutional curriculum.*

  Institutional utility.

  Not correctness.

  "It argued that questioning the assumption was disruptive," he said. "Not that the assumption was true."

  "The two are different."

  "Yes."

  A moment of quiet.

  Students moved through the courtyard behind them.

  Late light crossed the table.

  Then Sirin said: "The dissipation assumption supports the whole ranking structure."

  Kael said nothing.

  "If it's wrong," she continued, "the ceiling the rankings use may not be real. Block assignments. Access tiers. Advancement. All of it depends on a limit that might be administrative, not physical."

  "That was the real threat," Kael said.

  "Not error," Sirin said. "Instability."

  He looked at her.

  She looked back.

  "I'm going to keep working on it," he said.

  Not a declaration.

  A fact.

  "Not in lecture halls. In the annex."

  Sirin opened her notebook.

  Wrote something on a clean page.

  Tore it free.

  Slid it across the table.

  A reference number. A shelf code.

  "Restricted section," she said. "The calibration study Solt cites. The original version."

  Kael looked at her.

  "There's a faculty liaison who processes provisional access for documented research," she said.

  "Aldea won't sign it."

  "No."

  Sirin had already written the next part.

  "The rule only requires a current instructor," she said. "It doesn't say the research has to be in that instructor's field."

  He held the paper for a moment.

  He had eleven days left before the consequences of the notation fully unfolded.

  He had Zone Three. Intermediate texts. Two footnotes leading to an older study that had been replaced without its reservations being settled.

  And now a path to the original document.

  "You'd co-sign," he said.

  Not a question.

  "My proposal is already written," Sirin said. "It only needs your name."

  Kael looked at her for two seconds.

  Then he took the page, wrote his name at the bottom, tore it along the seam, and handed it back.

  Sirin slid it into the back of her notebook and returned to her work.

  No ceremony.

  Outside, the campus was shifting toward evening.

  The rule field had begun its slow descent into lower nighttime saturation. Students crossed between the academic wing and the dormitory blocks. Formal hours were over.

  Informal work was not.

  In the annex on the north side of the maintenance building, the morning had left no visible trace.

  Kael took the western route back to Block Four.

  His notebook held two open threads.

  The fracture, which would require fourteen days—a minimum baseline before he could determine whether it was stable, progressing, or something else. Fourteen days of the efficiency drill at sub-threshold load, observing the internal texture of his system, looking for any change in the hairline quality he'd first logged after the boundary crossing.

  The calibration study, which might require less.

  Sirin's proposal was in process. The research application was documented, the faculty liaison was named, the co-signatures were in place. Provisional access to a restricted document through a formal process with a formal record—which meant whoever reviewed it would see a documented academic inquiry, not a student poking at a gap they'd found by accident.

  The difference was important.

  He had twelve days before the next formal assessment.

  He intended to move on both threads before then—carefully, without triggering another informal flag, without making the work visible before he had enough of it to be worth defending.

  He walked through the western path and did not look at the fourth-floor window of the academic wing where Solt's office light was still on.

  He had what he needed.

  The rest would arrive at its own pace.