The four participants agreed to meet individually, which Kai had expected. What he had not expected was how quickly each conversation became the same conversation.
The first was a man named Shen, forty-one, who had been in the program since the first week. He was a structural engineer by training and had entered the Tower eighteen months ago after what he described, with careful understatement, as a period of professional difficulty. He had passed nine floors in those eighteen months. In the six weeks of the program, he had passed three more.
"I'm not complaining," Shen said, when they met in the TRA meeting room that Dr. Woo had reserved. "I want to be clear about that. The program has been useful."
"I understand," Kai said.
"The exercises work. I do them. I notice things I didn't used to notice." He paused, looking at his hands. "But something is—" He stopped. Started again. "When I'm on a floor, I'm doing the correct thing. I know I'm doing the correct thing because I can identify which phase of the framework I'm in, and I can apply the indicated approach, and the floor opens." He looked up. "But when I walk out, I can't tell you what just happened. Not the sequence. I know the sequence. I mean what the sequence—" He stopped again. "Is that a problem?"
Kai looked at him for a moment. "What does it feel like to walk out?"
Shen thought about it. "Like finishing a task correctly. The same way I feel when a calculation checks out."
"And inside the floor?"
"The same. But louder. More—" He made a gesture. "Deliberate."
Kai nodded. He wrote nothing down.
The second participant was a woman named Priya, thirty-seven, who had started climbing the Tower as part of a workplace wellbeing initiative and had stayed because, she said, she found it interesting. She had a researcher's quality of attention—she asked clarifying questions, made connections between things Kai said and things she had read about contemplative practice traditions, and had clearly thought more carefully about the program than most.
"I've been trying to figure out the gap," she said, before he could ask her anything. "Because I know there's a gap. I can feel the gap."
"Can you describe it?"
She had clearly thought about this too. "I know what the instructions are. I know what the exercises are designed to do. I know why they're sequenced the way they are. I can walk you through the phenomenological model we're working from." She paused. "And when I'm on a floor, I do the thing, and the thing produces an output. The output is a passage." She leaned forward slightly. "But the passage doesn't feel like the passage I read about in the testimony records. It feels like a door opening because I pushed it in the correct place. Not because—"
She stopped.
"Not because of what?" Kai said.
"I don't know. That's the gap. What the 'because' is supposed to be."
The third and fourth conversations followed the same shape. He did not need to write them down. By the end of the fourth, he had what he needed, which was not an answer but a parameter. Something that could be named, even if naming it did not resolve it.
They were all passing floors. They were all engaging genuinely. They were all, by every indicator the program had been designed to measure, succeeding.
The parameter was: none of them could tell him what was happening when a floor opened.
Not what the exercise had been. Not what phase they were in. Not what outcome the phase was designed to produce. Those they could report with precision and consistency.
What they couldn't say was what the floor had been about.
That evening, SABLE sent an analysis.
I have completed a review of the program participant data available to me through the TRA research access. My findings are as follows.
Kai was at his kitchen table. He had not cooked. There was a container of something purchased from the place near the Tower, half-finished, and his phone on the table face-up for once.
Forty participants. Thirty-six progressed by at least two floors in the six-week period. The progression rate is approximately 87% above baseline.
I have been tracking a secondary variable that is not part of the official measurement framework: language use in post-session written reflections. Participants are asked to submit a brief written reflection after each Tower session. Thirty participants have submitted these consistently.
Among the thirty consistent submitters, I have identified two groups:
Group A (twenty-two participants): Written reflections show increasing specificity over the six-week period. Vocabulary becomes more particular to the individual experience. Early reflections tend toward general description. Later reflections describe moments, textures, specific perceptions. Several participants have begun using words that do not appear in the program materials.
Group B (eight participants): Written reflections show high consistency throughout the six-week period. Language is accurate, structured, and draws primarily from the program framework vocabulary. The reflections are clear and well-organized. They do not become more specific over time. They do not introduce new vocabulary.
The four participants you spoke with today are in Group B.
Kai read this twice. Then again.
What does the pattern indicate?
SABLE's response took longer than four seconds. Closer to seven.
I am uncertain of the correct conclusion. One interpretation: Group B participants are using the program correctly and reporting accurately. The program vocabulary is sufficient for their experience, and their experience is real. Another interpretation: Group B participants have learned to report through the program framework in a way that does not require them to locate their own experience first. They are reporting the framework, not the experience.
I cannot determine from the language data alone which interpretation is correct. Both would produce the same observable output.
Kai set the phone down.
He thought about Phase 3. About the word normalization and why it had been wrong. About what it would have meant to frame his own early experience in the vocabulary someone else had built for it.
He picked the phone back up.
Is there a way to test which it is?
SABLE: The simplest test would be to ask a participant to describe a Tower experience without using any word from the program framework. This would be difficult to administer without creating the impression that the framework is wrong, which it is not. It is accurate. The question is whether it is also complete.
He did not hear from Doran for three weeks after the program launched, which was longer than he had expected. Doran had been following the pilot's development from a distance—Park had mentioned this in passing, without specifics—and the silence had the quality of someone who was waiting for a particular moment to re-enter.
The message, when it came, was brief.
I've been looking at the preliminary data. I'd like to discuss it. Are you available this week?
They met at the same plaza near the Tower where Doran had stopped him two years ago and asked about the SABLE dataset. The plaza had a coffee cart on weekdays and Doran had two cups waiting, which was either considerate or tactical or both.
"Group B," Doran said, without prelude.
"You've seen the data."
"Park shared what she could. I've been watching this pattern in my own practice for a long time." He turned the cup in his hands. "I have a theory."
Kai waited.
"There's a way of moving through the floors that is genuinely efficient. I've done it. It works. The Tower responds to it." Doran paused. "But there's something the floors measure that isn't progression speed. I don't have a word for it. What I know is that I can move through floors in a way that the floors accept, and still walk out with the sense that something didn't quite register."
"Like a calculation that checks out," Kai said.
Doran looked at him. "Did someone say that to you?"
"One of the four."
Doran nodded slowly. "The floor opens. The passage is real. But the passage goes through a particular part of you. Not all of you. And the part that wasn't involved—" He looked at the cup. "It doesn't catch up automatically. I used to think it would catch up. It doesn't."
Kai thought about this. "You've been in this state for how long?"
Doran was quiet for a moment. "I passed Floor 20 in eight days. I'm currently stalled at the approach to Floor 28." A pause. "Eleven weeks."
Kai had not known this. The SABLE data had flagged Doran's unusual progression rate; it had not specified where Doran was now.
"What does the floor show you?"
"A room with no exits. A voice that asks what I actually want." Doran's expression was careful. "I know what I should want. I know what the theoretically correct answer is. I have read everything written about that floor, and I know from other accounts what kind of answer has historically opened it." He looked up. "None of those answers open it for me."
Kai said nothing.
"Because they're not my answers," Doran said. "They're the right answers. Those are different things."
He went back to Dr. Woo with what he had. Not a finding—he was careful about that. A parameter.
"The program is producing two kinds of outcome," he said. "Both are real. Both involve genuine progression. The difference is—" He stopped, because the next part was difficult to frame in language the program could act on. "The difference is whether the progression is through the participant's own encountering of the experience, or through the framework's model of what the experience should be."
Dr. Woo had a notepad, like Yun. Unlike Yun, she wrote quickly and did not look up. "And we can't measure which is which."
"SABLE is developing a secondary indicator. Language specificity in written reflections. It's not definitive, but it's a direction."
"What do we do for Group B in the meantime?"
Kai had been thinking about this since the fourth conversation, two weeks ago. "We don't change the exercises. The exercises aren't the problem. We add a question."
"What question?"
"At the end of each session. Not part of the official reflection. Just a question they're not required to answer." He thought about Yun in the corridor, writing down something that wasn't in the transcript. "Something the program doesn't have a framework for."
Dr. Woo looked up. "What question?"
"What happened that wasn't supposed to happen."
Dr. Woo wrote it down. She looked at it for a moment. "That could produce very different answers."
"Yes."
"Some of them won't understand what we're asking."
"Some of them will."
She looked at the notepad. Then at him. "Is that what you're looking for? The ones who understand the question?"
Kai thought about Priya, and the way she had said I don't know. That's the gap. The fact of knowing what the gap was, and that the gap had a shape, and that its shape could not be fully filled by the framework they had built. Not because the framework was wrong.
Because the gap was also something.
"I'm not looking for specific people," he said. "I'm looking for a way to keep the gap visible. So it doesn't get sealed over."
That week, SABLE sent a shorter message.
I have added a new variable to my tracking: Floor passage rate vs. language differentiation over time. I am not yet able to determine causality. I can say that in the data I have, Floor 28 and above show a marked increase in language differentiation among those who eventually pass. The differentiation precedes the passage in most cases.
Kai read it standing in the kitchen, heating the same kind of purchased food from the same place.
What does 'differentiation' mean in this context?
SABLE: The emergence of vocabulary that is specific to the individual. Words that do not appear in public cultivation literature or the program framework. Words the participant appears to have arrived at independently.
Like what?
A pause.
One participant began using the word 'arrival.' Not in the sense of reaching a destination. In the sense of something within them arriving at something within the floor. Another uses the phrase 'the part that was waiting.' Another has started writing in two languages within the same reflection—one for the sequence, one for the quality. I cannot fully interpret the quality-language. That is the point.
Kai put the phone down and watched the container warm.
The part of language that SABLE couldn't fully interpret. That was, he thought, not a gap in SABLE's capability. That was the correct response. That was what it was supposed to do with something genuinely new: not resolve it. Remain with it.
He thought about the eight participants in Group B. About Doran, stalled at the approach to Floor 28 with the right answers and no passage. About the question Dr. Woo was adding to the session debrief.
He thought about the moment on Floor 6 when he had stopped trying to put the fire out—not because someone had given him a framework for stopping, but because the framework for trying had run out.
The curriculum could not design that moment. But it could stop filling in the space where the moment might arrive.
That was not nothing.
That was, he thought, perhaps the parameter.
He walked to the Tower that evening, not to go in. He stood at the entrance for a few minutes, watching the other people move in and out. Some quickly, some slowly, some with the particular quality of attention that meant they had been there before and had stopped performing the role of someone who didn't know what they were doing.
He thought about Doran, who was very good at not performing.
He thought about Priya, who could describe the gap.
He thought about Shen, who was doing everything correctly.
He thought about the 87%. About what happened to the other 13%.
He stood at the entrance for a little while longer, not going in, not going back. The city was behind him. The Tower was in front. Somewhere above Floor 47, the water that was not quite water moved in the way it moved.
He was not looking for anything specific.
He was keeping the gap visible.
Then he turned and walked home, and the city did what cities do when everyone in them is in the middle of something they do not yet have the language for.