They were three days from Road 2 when she laid out all three versions of the Seventeen Borrowed Steps on the floor of her room.

She had been building toward this session for weeks, since the first conversation at the fence when she had described the connection between his injury and the technique documented in the Reedfall archive. What had followed was the process of extracting the relevant material from her memory in written form, working through the notation system, comparing the three distinct versions she had in her archive and documenting the specific ways they diverged.

The floor was the only surface large enough.

She had organized the pages in the order she intended to work through them: the Reedfall version first, her grandfather's original documentation with the margin note about unverified load distribution; the second version, from a text donated to the archive by a practitioner from the eastern schools, which she had read twice and which differed in specific places from the Reedfall version; and the third version, from a fragment in the archive's oldest section, which she had read once and which was incomplete but whose visible portions were consistent with the second version in most respects and differed from the Reedfall version in the same places.

She had additional notes along the margins: her analysis of the specific points of divergence between the versions, the biomechanical implications of each variant, the cross-references to other techniques in the archive that shared structural elements with the seventeenth step.

Ilyun came in, looked at the floor, and sat down across from the pages.

"Three versions," he said.

"Yes. Plus your account, which is a fourth."

He looked at the pages without touching them. She let him.

---

"Walk me through the first one," he said.

She walked him through the Reedfall version, using the notation she had developed, moving between the written description and the physical demonstration she provided where the notation was insufficient. She had been developing her own execution of the sequence in the context of the training drills and her execution was still imperfect, but it was accurate enough for demonstration purposes: she could show where the technique went and what it required without claiming to have mastered it.

The seventeenth step of the Reedfall version involved a rotation through the wrist joint under extension at the end of the movement arc, placing the load at the farthest point from the center of mass. She demonstrated this slowly.

"Show me," she said.

He demonstrated the version he had been taught, the version that had injured him. Slowly, as she had, stopping at the seventeenth step.

She watched.

She was cross-referencing immediately, the archive running parallel to the observation, looking for the alignment between what she was seeing and what she was holding. The comparison resolved in approximately two seconds.

The movement was from the Reedfall archive. She had known this since the early conversations. What she was seeing now, with both versions in front of her, was the specific nature of the difference: it was not a different technique. It was the same technique executed through a different structural path, a path that placed the load at the same final point but arrived there through a different arc, one that committed the wrist to the extension earlier in the movement.

She had not seen this from the written descriptions alone.

She had needed to see both versions demonstrated to see it.

"Show me again," she said. "Just the seventeenth step."

He showed her.

She watched and compared and thought about the specific load placement at the moment of full extension and what the medical text had said about the vulnerability of the wrist joint at that position under that kind of torque.

"The arc is different," she said.

He did not speak.

"In the Reedfall version, the wrist commits to the extension at the end of the rotation. The load accumulates through the movement and arrives at the joint at the peak of the arc." She demonstrated. "In your version, the commitment happens earlier. The joint is in full extension while the rotational load is still building. The load and the extension peak simultaneously."

Silence.

"That is the vulnerability," she said. "Not the technique. The arc."

She looked at him.

He was looking at the page with his version, not at her. The specific quality of stillness she had learned to read in him, the stillness of someone who was receiving something that mattered, was present in a form she had not seen before. Quieter. More complete.

"My instructor told me I committed early," he said. "That I rushed the seventeenth step."

"In the Reedfall version, committing early would be an error," she said. "Your version is built to commit early. You were executing your version correctly. The arc in your version creates the loading condition."

He looked at the floor for a long time.

---

She began the comparison of the second and third versions, moving through the divergences systematically, noting where versions two and three agreed with each other and where they diverged from version one. The pattern was becoming visible in a way that she had begun to see in the written documentation but that was clearer now with the physical demonstrations alongside it.

All three versions produced the same final position. All three put the load on the same joint at the same degree of extension. They arrived there through different arcs, with different timing, with different distributions of the rotational force through the movement.

The question she was building toward was whether one of these arcs was safer than the others, whether the load distribution in the movement could be arranged differently and whether "differently" meant "less likely to produce the injury she was documenting."

She was not there yet.

She knew she was not there yet.

But she was closer than she had been, and the closeness had a specific quality, the quality of a problem whose shape was becoming clear enough to work with.

"There are more versions," she said. "In schools I have not encountered. If I advance through the Assembly, I will be in contact with practitioners from traditions I do not have in the Reedfall archive. I can gather more versions."

"Yes," he said.

"The picture will be more complete with more versions."

"Yes."

She was gathering the pages from the floor, organizing them back into their folders. He was watching her work and the watching had the particular quality that she had been filing and not examining.

"You are going to figure this out," he said.

She stopped.

It was not the kind of thing people said to encourage. She had received encouragement about the investigation from practitioners at the Assembly who expressed it in the way people expressed polite support for ongoing projects: as possibility, as hoping, as aspiration. What Ilyun had said did not have the quality of possibility or hope.

He said it the way her father said the things he was certain about. The way a practitioner described the outcome of a drill they had already run enough times to know what the result would be.

She looked at him.

He was looking at the pages she had been collecting from the floor.

She finished organizing the pages and put them in the folder and set the folder on the desk.

"I know," she said.

She did not mean: I am also confident. She meant: I heard the way you said that.

He met her gaze for a moment.

Neither of them said anything further. The room held the afternoon light and the pages of the investigation and the small significant weight of what they had not said.

She opened the folder again and took out the first version, her grandfather's, and began reading through the section that followed the seventeenth step, looking for whatever her grandfather had understood when he had written the margin note: the note that said approach with caution, the note that she was now certain he had written because he had known, in some way she had not yet reconstructed, that the arc was wrong.

She read until the light failed.

The investigation was not finished. The technique was not solved. The central question, which was not only about the technique and had not been only about the technique for some time now, was not answered.

But the shape of it was clearer.

She could see where it was going.

She had been preparing for this session for three weeks.

Not by drilling additional material or reviewing documentation she had not yet reviewed. She had been preparing in the way she had learned to prepare for the things that mattered most: by letting the relevant knowledge settle into the order it wanted to be in rather than the order she had imposed on it. She had laid out the versions on the floor of her room and sat with them and not tried to resolve them, not yet, waiting for the picture to develop in the way pictures developed when you gave them time.

The picture was almost developed.

She could feel the shape of something, not the final shape, but the shape of the shape: a structural principle that was common to all three versions of the technique and that explained, in a way she had not yet been able to fully articulate, why the differences between the versions produced such different physical outcomes.

She had the picture. She did not have the language.

She needed to see Ilyun's version alongside her own versions to complete the picture.

---

After he demonstrated the seventeenth step and she made the comparison and said "the arc is different," she watched him receive the information.

She had been watching him receive information about his injury since the first conversation at the fence, and she had built a comprehensive record of how it happened: the stillness, the specific quality of the stillness, the way it was not the same stillness as ordinary consideration but a deeper kind, the stillness of someone who is encountering something that has been absent from their understanding for a very long time and is now present.

This was the most complete version of that stillness she had seen.

She had told him things before that had landed with weight: that his injury was not random, that it was connected to the Reedfall documentation, that the version he had been taught was one of multiple versions that diverged in specific places. Each of those had produced a version of this stillness.

What she had said just now was different. She had not told him that the documentation was flawed, or that there was a question about which version was correct. She had told him something more specific and more complete: she had told him the precise mechanism by which the injury had occurred, the specific moment in the movement where the loading condition created the vulnerability, and she had told him that this mechanism was a consequence of the arc of the movement and not a consequence of his execution of the movement.

You executed it correctly. The arc is wrong.

She had not said this last part in those words. She had described the arc and the loading condition and let him arrive at the conclusion. He had arrived at it. She could see him arriving.

Five years.

She thought about what it was to carry a piece of knowledge about yourself for five years and have that knowledge be wrong, and to know it was wrong and have no way to prove it, and then to have the proof arrive from a direction you had not anticipated.

She thought about what it meant that the proof had come from her.

Not from the texts, not from other practitioners, not from the school that had trained him. From her. From the archive she had been carrying without knowing she was carrying it, from the three hundred and forty-two texts she had read and retained, from the specific margin note her grandfather had written in a text that had survived a fire because she had happened to read it twice before the fire began.

---

"There are more versions," she said. "In schools I have not encountered."

"Yes."

"If I advance through the Assembly, I will meet practitioners from traditions I do not have in the Reedfall archive. Some of them will know versions of this technique."

"Yes."

"I will document every version I find."

He looked at the folder she was holding.

"You have been working on this since the fire," he said.

"Since before the fire. I noticed the inconsistencies in the archive before the fire. The fire made the problem visible because it forced me to distinguish between what I remembered reliably and what I was uncertain about."

He was quiet.

"The archive burning was a bad thing," she said. "I have not changed my mind about that. But it forced a precision in the record that I would not have done otherwise. I was treating the archive as reliable. The fire showed me the places where it was not."

He looked at her. "Your grandfather knew."

"I think he suspected. The margin note is too specific for general uncertainty." She paused. "I think he found the inconsistency in the version he had received and noted it and did not have access to other versions to resolve it. The archive only had one version before the donations I read as a teenager. He may not have known other versions existed."

"He had the instinct."

"Yes." She was quiet for a moment. "He also had the instinct that the technique was important enough to preserve anyway, even flagged as uncertain. That is why it survived in the archive."

"And in your memory."

"And in my memory."

She stood with this truth, the complete chain of it: her grandfather's instinct, her father's maintenance of the archive, her own reading, the fire that forced the distinction between reliable and uncertain, the agreement at the fence that had put her in the same room as a practitioner who held the injury as evidence.

The chain had links she had not constructed. It had been forming for a long time.

"You will figure this out," he said.

She looked at him.

"I know," she said.

She meant: I heard what is underneath that. I hear the way you say the things you are certain about. I have been listening to you long enough to know the difference.

He held her gaze for a moment.

She looked back at the folder. She opened it and took out the first page and began reading toward the section that followed the seventeenth step, looking for what her grandfather had known, working toward the understanding that was almost complete.

Outside, the last light of the afternoon moved across the floor of the room.

The investigation was not finished. It would not be finished for a long time. But the central question was no longer: what is wrong. The central question was: how do we build the thing that replaces what was wrong.

She read.

Beside her, he was also reading.

They worked in silence until the light was gone.

After he said you will figure this out and she said I know, they worked in silence for another two hours.

She was reading the Reedfall version of the seventeenth step, the section that followed it, her grandfather's notes on the structural logic of the sequence overall. She had read this section before, twice, but reading it now, with both demonstrations available for cross-reference, she was seeing things she had not seen before.

Her grandfather had written the margin note about unverified load distribution.

Below the margin note, in smaller characters, in what looked like a later addition, he had written four words that she had always read as a general caution: examine the arc first.

She had read this as: when attempting this technique, examine the arc of the movement before proceeding.

She was reading it now as: the arc is where the problem is.

He had known. Not in the complete way she was now able to describe, not with the full analysis of the loading condition and the joint vulnerability at the peak of extension. But he had known that the arc was the thing to look at. He had named it. And then, because he did not have the other versions to compare against, he had not been able to resolve it.

She looked at the four words for a long time.

She thought about her grandfather, who had founded Reedfall Hall and written the methodology text and noticed something in a technique that had been transmitted to him and had not been able to resolve what he had noticed and had still written it down. She thought about what it meant to notice something you cannot resolve and write it down anyway, the faith that the writing would find its reader.

She was the reader.

She had been the reader since she had first opened the methodology text and she was still the reader now, in this room, with the pages on the floor and the other version visible in the memory of a man who had carried the wrong version for five years and was sitting across from her reading her reconstruction.

---

She looked at him.

He was reading the pages carefully, the same quality of attention he gave everything, but she had been watching him long enough to read the layers of the attention, and this layer was different from his assessment attention or his training attention. This was the attention of someone engaging with something that had been personal for a long time.

She thought about what it was like to have your own injury described accurately, the precise mechanism of it named, the arc identified. She thought about what it meant to have the correct naming come from outside the tradition that had formed the injury, from an archive that had been largely destroyed and was now living in someone else's memory.

She thought about her grandfather's four words.

She thought about the chain of events that had put those four words in the same room as the injury they described: the archive built over generations, her own reading, the fire that had made the distinction between reliable and uncertain visible, the agreement at the fence that had put her in proximity to someone who held the injury as evidence.

She thought about the quality of the moment they were in, this room, the pages on the floor, the afternoon light, the silence that held both the investigation and whatever else it held.

"My grandfather wrote four words under the margin note," she said.

He looked up.

"Examine the arc first," she said. "I read it as a general caution. It is not a general caution. He named it. He did not know what he was naming but he named it."

Ilyun was very still.

"He could not resolve it because he did not have the other versions," she said. "If he had had your account, he would have resolved it in an afternoon."

"He did not have my account."

"No. But he named the right place to look." She looked at the pages. "I have his naming and your account and three versions of the documentation. What I need now is the other versions."

"Road 2," he said.

"Road 2," she agreed. "And everything after."

She looked at him.

He looked at her.

The investigation was not a transaction anymore. She was not certain when it had stopped being a transaction. She was not certain it had ever been entirely a transaction, but the last trace of the transactional quality was gone now, in this room, with the four words visible and the chain of events that had made this moment possible arranged in her memory like a sequence she could read in full.

She closed the folder and looked at the road ahead.

Three days to Road 2. The next version of the technique was out there, in a school she had not yet encountered, in a practitioner she had not yet met.

She was going to find it.

They were going to find it.

She gathered the pages from the floor and organized them carefully and did not think too long about the we that had arrived in her thinking, the same way the three inches had arrived, without announcement, already there.