The data arrived on the third day, as projected.

  He found the notification in his student record at 6:47 in the morning — a single line entry confirming that the assessment platform log had been processed and the full output file was available for download under subsection seven of the institutional data access policy. The entry had a timestamp of 3:12 AM. The system processed overnight.

  He downloaded the file.

  It was larger than he'd expected.

  ?

  The standard assessment record for any given student occupied approximately two kilobytes — ID numbers, tier classification, merit score, the summary output the platform generated after processing. Clean, minimal, designed for administrative use.

  His full log was 340 kilobytes.

  He looked at the number for a moment.

  Then he opened it.

  ?

  The outer ring data was not formatted for human reading.

  He'd expected a structured table — rows, columns, labeled fields. What he found was a continuous string of instrument readings logged at intervals of 0.4 seconds across the full duration of his assessment, each entry a seventeen-digit value with no attached label, no defined scale, no reference baseline. The calibration study Aldea had shown him had described the outer ring output as "secondary field signature characteristics," which was precise enough to be accurate and vague enough to be useless as an interpretive framework.

  He needed a reference.

  He needed to understand what the instrument had been built to measure before he could understand what his 340 kilobytes of values meant.

  He went to the library.

  ?

  The calibration study itself was not in the main archive — he'd confirmed that the first time. But the study had cited its own methodology, and methodology citations were recorded in the Index even when the primary document was classified. He spent forty minutes with the Index, working backward from the call numbers he already knew, cross-referencing against the citation list he'd reconstructed from adjacent documents over the previous weeks.

  He found four reference texts the study had drawn on. Two were available on open shelves. One required a faculty access code. The fourth was listed as "restricted — physical copy only, second-floor reading room by appointment."

  He filed the appointment request before he left the library.

  He checked out the two available texts, signed the log, and carried them back to Block Four under his arm. The morning was cool and clear, the kind of autumn light that made the campus look like something deliberate — like someone had arranged the stone buildings and the old trees specifically for this quality of illumination. He noted it and moved past it. He had work to do.

  ?

  The first text was a 1970s instrumentation manual for resonance-class assessment platforms — pre-SR framework, written before the standardization that had made the outer ring secondary. He read it at his desk with a pencil and a separate sheet of paper for notes, the way he read anything he expected to need precisely later.

  The chapter on outer ring design function took him forty minutes to work through.

  What he found there changed the shape of everything he'd been building.

  The outer ring had not been designed as a secondary system.

  In the original SR-platform architecture, the outer ring and the active core had been equal components of a dual-channel assessment — active core reading output quantity, outer ring reading what the manual called "field signature coherence." The two channels were intended to be analyzed together. The quantity score would be interpreted against the coherence reading to produce a complete picture of a practitioner's field. Neither channel was primary. Neither was supplementary.

  The 1987 calibration study had not been the first time someone had noticed the direction signal. It had been the most recent formal attempt to understand something the original engineers had built the platform to detect, and that two generations of administrative framework had progressively deprioritized until the outer ring's data was being logged automatically but never read. The logging infrastructure remained because it was built into the hardware. The analytical framework for using it had been quietly retired.

  He looked at what he'd written.

  He thought about forty years of practitioners walking through enrollment assessments, producing data in two channels, going home with a number from one of them.

  He looked at his desk, at the window, at the morning light on the stone wall of the building across the courtyard.

  Then he opened his outer ring data again.

  ?

  He needed a baseline.

  His data was 340 kilobytes of unlabeled values. To interpret any of it, he needed to know what a normal outer ring reading looked like — what the instrument produced for a practitioner with standard field signature coherence, so he could identify what his readings differed from and by how much.

  The 1987 study had included comparative baseline data for its three observed cases. The study was classified. The methodology citations he'd found in the Index gave him the original framework the study had used, but not the study's own numbers.

  He wrote down what he had and what he needed.

  *Have: raw outer ring output, 340KB, unformatted, 0.4s intervals*

  *Have: original instrument design documentation — outer ring = field signature coherence channel, originally equal status to active core*

  *Have: pre-standardization methodology framework*

  *Need: baseline coherence readings for standard practitioners*

  *Need: the 1987 study's raw data or its calibrated reference values*

  He looked at the second list for a while.

  The restricted physical copy in the second-floor reading room was scheduled for tomorrow morning. He didn't know yet what document it was — the Index listed it only by call number. It might be the study itself. It might be something adjacent that cited the study's data. It might be neither.

  He filed this under *variables currently resolving* and moved to what he could work with now.

  ?

  Without baseline values, he couldn't interpret the outer ring data in absolute terms. But he could look for internal patterns — how his readings changed across the duration of the assessment, whether there was structure in the variation, whether the seventeen-digit values had a recognizable shape when plotted against time.

  Taken from Royal Road, this narrative should be reported if found on Amazon.

  He had the assessment sequence memorized — he'd requested that documentation separately, months ago, as part of his standard review of the evaluation methodology. Timestamp: entry to the assessment chamber. Timestamp: platform initialization. Timestamp: active core measurement begin. Timestamp: active core measurement end. Timestamp: exit. Five events. Precise times.

  He borrowed Sirin's calculation tablet for two hours. She handed it over without asking why, which he noted — she had asked once, early in the term, when he'd borrowed it for something unrelated. She didn't ask anymore. He wasn't certain what to make of that pattern, so he filed it alongside several other observations about Sirin Vael that he was accumulating without yet having sufficient context to draw conclusions from.

  He transferred his outer ring data to the tablet and spent ninety minutes building the plot. Manually, value by value, because the tablet's automatic graphing function required labeled axes and he didn't have labels yet. He wasn't certain what scale the seventeen-digit values used. He could only plot relative magnitude — value against time, with the sequence of assessment events marked.

  The result was not what he'd expected, which meant it was useful.

  He had expected — if there was anything to find — a gradual curve. A buildup, a peak during the core measurement, a return toward baseline. The standard shape for most resonance-class phenomena: accumulation, peak, dispersal. The shape that would correspond to the active core's reading profile.

  What the plot showed was a step function.

  Flat for the first forty seconds. Then an abrupt jump to a new level — not a curve, a vertical step, the value changing within a single 0.4-second interval and then holding. Then flat again at the new level. Then another step. Then flat. Then — at the moment the active core measurement had begun — a large step to a sustained plateau approximately 2.4 times the initial level, where it remained for the rest of the recording.

  He looked at the shape for a long time.

  Three steps. Each corresponding exactly to one of the five timestamped assessment events.

  First step: when he'd entered the assessment chamber.

  Second step: when the platform had initialized and the field reading had begun.

  Third step — the large one — at the moment the active core measurement had started, and his field had engaged fully with the platform's reading apparatus.

  Not output. Not the measurement of what he was producing. The moment of contact between his field signature and the instrument.

  He set the tablet down and looked at the window.

  The word had come three days ago.

  *Direction.*

  He'd written it in the back of his notebook because it was true before he could prove it. Now he had the first shape of the proof.

  Standard resonance output was a reservoir model — the field had a capacity, and the assessment measured how much it could produce during the measurement window. The curve would be gradual because it was drawing down a stored quantity, the way water drained from a tank at a rate proportional to pressure.

  His outer ring reading was not a draining tank.

  His outer ring reading was a contact response.

  It changed when it touched something.

  ?

  Sirin found him at the window at 4:30, which was when the afternoon light moved off the desk and made the calculation tablet difficult to read without adjusting the lamp. He'd been at the window for fifteen minutes by then, not thinking about anything in particular — or thinking about many things simultaneously in the way that sometimes looked, from the outside, like thinking about nothing.

  "What did you find," she said. Not a question. She'd learned, somewhere in the past month, that questions opened space he didn't always fill.

  "A step function," he said. "Three steps. Each one corresponds with a moment of contact or initialization during the assessment sequence."

  She crossed the room and looked at the tablet. He watched her read the plot without explaining it. She was quiet for almost a minute.

  "It's not a buildup," she said.

  "No."

  "It responds to specific events, not duration."

  "Yes."

  She looked at the three steps. Her finger moved along the time axis, stopping at each jump.

  "The direction shows up when there's something to interact with," she said.

  He'd reached the same formulation earlier. Hearing her say it clarified something about its implications that he hadn't finished working through — not the interpretation itself, which he'd already established, but its extension. If the direction signal was relational, not intrinsic, then it wasn't a property of his field in isolation. It was a property of how his field related to what it encountered.

  Which meant the framework's mistake wasn't only about what it measured.

  It was about what it assumed the thing being measured fundamentally was.

  "The outer ring reads coherence," he said. "Field signature coherence. But coherence with what — coherence as an internal property of the field, how organized it is internally, or coherence as a relational property, how the field relates to what it's adjacent to — that question the original framework left open."

  "And the step function suggests relational."

  "It's consistent with relational," he said. "I don't have a baseline. I can't confirm the interpretation without the 1987 study's reference values, or equivalent data from another practitioner."

  She looked at the plot again. He could track the direction of her reasoning by which part of the graph her eyes rested on. She was on the sustained plateau at the end — the 2.4x level — and she was measuring its flatness. Its stability.

  It hadn't dipped. It hadn't varied during the measurement window. It had stepped up and held.

  "I could request my own data," she said. "Subsection seven applies to all students."

  He had considered this the previous evening and set it aside because a single data point was insufficient to establish a reference range. But he'd also noted that he'd set it aside because it would require asking Sirin to do something with her own record, and he had not yet established whether that fell within the scope of what they were doing together.

  Apparently it did.

  "Your resonance classification is standard," he said. "Your outer ring readings would give me a baseline value, but a single comparison point is insufficient to establish a reference range."

  "Still better than nothing."

  "Yes. It would eliminate some interpretations and narrow others."

  She filed the request from her student record while he watched. The process took forty seconds. She didn't ask him to confirm that he wanted it, or describe what she was doing, or frame it as a favor. She looked at the filing confirmation for a moment.

  "Three days for processing," she said.

  "Probably two, this time of term. The system runs faster mid-week."

  She handed him back the calculation tablet.

  He noted the action — the return of the tablet, the matter-of-fact completion of the filing, the absence of any framing around either — and added it to the set of observations about Sirin Vael that he was still accumulating.

  The set was getting larger.

  He wasn't certain what to do with that yet.

  ?

  He returned to the library at closing time and confirmed tomorrow's appointment with the desk attendant.

  Second floor, reading room, 9:00 AM.

  The attendant turned the register around.

  *Calibration Study: Assessment Platform SR-Series, Outer Ring Output — Field Signature Secondary Characteristics. Authors: [institutional committee]. Year: 1987. Classification: Restricted internal use. Access: reading room only, request required.*

  He looked at it for a moment.

  "The study itself," he said.

  "That's what's listed," the attendant said.

  He thanked him and went back to Block Four.

  ?

  He wrote in the notebook until midnight.

  The new section was denser than anything he'd put in the back pages before — not the tentative shape-before-precision of earlier entries, but something tighter, more structured. He had enough pieces now to move from *direction exists* toward *direction behaves in a specific way*, and he could feel the difference between those two positions the way you felt solid ground after uncertain footing.

  *The outer ring data shows a step function, not a curve. Three steps, each corresponding to a discrete event during assessment — entry to chamber, platform initialization, active core engagement. Not buildup. Response. The field signature coherence doesn't increase gradually with effort or duration. It increases in discrete jumps at moments of contact.*

  *This is inconsistent with the scalar output model, which treats practitioner field as a reservoir with a fixed capacity being measured. If the outer ring reading were tracking output quantity, it would correlate with the active core reading — both should show a gradual increase as the assessment progresses and the conversion process continues.*

  *They don't. The outer ring responds to events. The active core responds to sustained output. Two different things being measured. The original engineers understood this. The current framework doesn't account for it because the current framework was built after the outer ring channel had already been reclassified as supplementary.*

  *The step function is consistent with relational coherence — the field signature's response to contact with other field structures. Each contact event produces a discrete elevation that holds. Not a quantity that accumulates. A relationship that establishes.*

  *Tomorrow: the 1987 study. Reference values, methodology, what the original researchers concluded before the recommendation was reclassified. The reading room has no copies. I'll need to work from what I can hold.*

  He looked at what he'd written.

  Three days since the word.

  Tomorrow he would find out what the people who'd seen it before had understood about it — and whether what they'd understood was sufficient for what he was beginning to see.

  He closed the notebook.

  Outside, the campus ran its evening routines. Somewhere in the main building, a light was still on in the academic wing's fourth floor. He looked at it for a moment — the lit window against the dark stone, a familiar feature of this hour — and then pulled the curtain.

  He had an appointment at 9:00 AM.

  He slept.