I prepared the way I prepare for everything that matters.
Not by rehearsing. Rehearsal produces rigidity — you build a script and then you spend the interaction managing the distance between the script and what's actually happening, which means you're always running half a second behind the real conversation. What I do instead is prepare the *space* around the interaction. The variables I control in advance, so that during the conversation itself I'm not managing logistics, I'm only managing the human element.
The variables I controlled: I reviewed the Kasahara structural site documentation until I knew it well enough to discuss it without notes, because notes in a preliminary conversation signal that you're not yet serious about the engagement, and seriousness was the first thing I needed to establish. I reviewed three publicly available Yaoyorozu Structural Consulting project summaries — the kind posted on reconstruction authority websites as examples of approved methodology — to understand the vocabulary she used when describing her own work. I made a list of twelve questions, ranked by priority, and then crossed out the bottom eight because you don't need twelve questions to accomplish three objectives in thirty minutes.
I kept four.
The four questions were not the most important information I wanted. They were the questions most likely to produce responses that revealed the information I actually wanted, which is a different selection criterion and produces a different list.
I woke at five-thirty and made coffee and sat at my desk in the apartment and read back through the preparation notes without adding to them, because adding to preparation in the final hours before an interaction is a form of anxiety management rather than actual preparation, and I have limited patience for managing my own anxiety through productive-looking displacement activity.
At six-fifty I picked up the phone.
At seven-oh-two it rang.
---
"This is Yaoyorozu."
Not *Yaoyorozu Structural Consulting, how can I help you.* Not *good morning, this is Momo.* Just the name, stated as identification, which is the phrasing of someone who is already in the conversation before it starts — who has allocated attention to this call and is ready to use it efficiently.
Two minutes late. I noted it and discarded it immediately because two minutes is within any reasonable margin and treating it as significant would be a misread that contaminated subsequent analysis.
"Thank you for making the time," I said. "I'll keep to thirty minutes."
"Appreciated." A brief pause, not hesitant — the pause of someone settling into a position. "You mentioned a land-assessment dispute with structural specifications in question. Can you give me the scope?"
Direct. Not *tell me about yourself* or *how did you hear about us.* Straight to the work, which meant she'd either read my email carefully enough to have the preliminary questions already answered, or she operated by default in the mode of someone who treated social preliminaries as costs rather than investments.
I gave her the Kasahara scope in four sentences — the development site, the assessment dispute, the valuation methodology question, the structural specifications that were in contention. Precise, no filler.
When I finished there was a silence of approximately three seconds.
"The load-bearing specifications are the contention point, or the material specifications?"
"Both. The original assessor's methodology treated them as separable. My client believes they're interdependent in a way that invalidates the valuation."
"They're right," she said, without any particular emphasis. A statement of fact delivered the way people deliver facts they've known for long enough that they've stopped being interesting. "Load-bearing and material specifications on a post-war reconstruction site are always interdependent. Any methodology that treats them as separable is either outdated or was designed to produce a particular valuation."
"Which of those applies in your experience with assessment disputes?"
A beat. Not a hesitation — she was deciding whether the question was genuine or rhetorical. "Both, in roughly equal proportion. The outdated methodology cases are usually low-stakes and resolve quickly. The designed-to-produce-a-particular-valuation cases take longer."
"Why longer?"
"Because the people who commissioned the assessment have an interest in the outcome. When you challenge a valuation, you're not challenging methodology — you're challenging whoever is served by that methodology. That's a different problem."
I was quiet for exactly one second.
This was the first reveal, and it had arrived in the first four minutes. She was not describing a technical problem. She was describing a structural one — the difference between a mistake and a design, the question of who benefits. That was the cognitive framework I needed to know about, and she'd produced it unprompted from a straight technical question.
"The site is on Kawabe Street," I said. "The replacement assessor — the one who produced the disputed valuation — has a prior relationship with a land-development consortium operating in the same ward."
Silence. Three seconds, four.
"Do you have that documented?"
"I have it at the preliminary stage. The documentation is one of the objectives of the current engagement."
"Who's your client?"
"I can't share that at this stage."
"Understood." No push, no negotiation — she accepted the constraint cleanly and moved. "Send me the site specifications and the disputed assessment. I'll review them and give you a preliminary read on whether the methodology is outdated or designed. That should take two days."
"That's faster than I expected."
"Structural specifications are specific. Either the methodology accounts for the site conditions correctly or it doesn't. It's not interpretive work." Another pause, this one with a different quality — something she was deciding whether to say. "Mr—" She checked herself. "You gave a firm name in the email, not a personal name."
"I prefer to operate that way at the preliminary stage."
"Also understood." The quality of her voice had shifted slightly. Not suspicion — something more like the register she'd use recalibrating a structural estimate when new load data came in. "Your firm focuses on consulting contract integrity specifically?"
"We focus on situations where the stated methodology and the actual methodology diverge. Structural assessments. Procurement contracts. Materials supply chains." I paused. "Post-war reconstruction has produced a significant volume of that kind of divergence."
The silence this time was longer. Six seconds. Long enough that I was counting and also doing something else — listening to the quality of the silence, the texture of it, whether it was processing or reacting.
Processing.
"Yes," she said finally. "It has."
And there was something in those two words that I had not expected to hear in the first thirty-minute preliminary call with a structural consultant I was approaching for a land-assessment review. Something that sounded less like professional agreement and more like someone confirming a thing they'd been sitting with for long enough that confirming it out loud felt like relief.
I did not react to it.
"I'll send the specifications this morning," I said. "If your two-day preliminary is promising, I'd like to discuss next steps."
"That's a reasonable sequence." Back to professional register. Clean, efficient. "I'll be in touch by end of business Friday."
"Thank you for your time."
"Thank you for the early hour. I get more thinking done before seven than most people get done by noon."
She ended the call.
---
I sat with the phone in my hand for approximately thirty seconds.
Then I put it on the desk and opened the preparation notes and added a new section underneath the four questions.
I wrote: *Seven minutes and forty seconds. Four primary observations.*
*One: technical precision is automatic. She doesn't perform expertise, she operates from it. The load-bearing / material specification answer was immediate and unqualified. No hedging, no professional liability padding. She said what was true.*
*Two: structural thinking extends beyond structural work. The outdated-versus-designed framework applied to assessment methodology is not standard professional vocabulary. She's mapped the political layer of her field independently, not from training.*
*Three: the six-second silence after reconstruction divergence. Processing, not defensive. She has something she's been thinking about in that space. What.*
*Four: "It has." Two words carrying weight they weren't asked to carry. Something she knows. Something she hasn't said to anyone, or hasn't said in this register.*
I underlined observation four.
Then I added a fifth: *She asked for a name before catching herself. Not paranoid — professionally cautious. And she accepted the constraint without friction, which means she's worked in contexts where information is controlled as a matter of course.*
Hero work. She'd been a hero before the war, at the student level. The habits of operating in environments with information discipline don't fully leave.
I closed the notes and sent the site specifications and disputed assessment to the email address she'd written from, with a brief professional cover note and no additional information.
Then I sat back in my chair and looked at the ceiling.
Option seven.
The argument for it had gotten stronger in seven minutes and forty seconds. The argument against it — that I couldn't predict her decision-making under disclosure pressure — hadn't weakened, but I had more signal now than I'd had before the call. The six-second silence. The *it has.* The structural thinking applied beyond the technical domain.
She was already proximate to something she hadn't named.
Whether she'd already reached it, or was approaching it from a different angle, or had stopped short of it because she didn't have the final piece — that was what I needed the second conversation to tell me.
---
Toga's read, when I debriefed her at nine, was characteristically succinct.
"She already suspects something."
"Possibly."
"The two-word response. The silence. That's someone confirming a thing they've been carrying."
"It's consistent with that interpretation. It's also consistent with a professional who's observed the general pattern of post-war contract fraud without having a specific case in front of her."
Toga looked at me with the expression she uses when she thinks I'm being precise in a way that's covering for something. "Is there a difference, for what you're trying to do?"
I thought about it.
"Yes," I said. "If she's carrying a specific case, she's already in a reactive posture. She'll be harder to approach cleanly because she'll be cross-referencing everything I say against whatever she already has. If she's observed the general pattern, she has the framework but not the application, and I can place the application in front of her in a way she can follow from first principles."
"Which is easier to work with."
"Which produces a more reliable outcome. If she reaches the conclusion herself from evidence I place correctly, the conclusion is hers. She owns it. She acts on it with full conviction instead of borrowed certainty."
Toga was quiet for a moment. "You want her to figure it out."
"I want her to be in a position where she can figure it out. With the right information, in the right sequence."
"And if she already has part of it?"
"Then the sequence is shorter." I picked up my coffee. "Either way, the Friday call is the one that matters."
Toga made a note on her small pad. "The watcher. If they're monitoring the Yaoyorozu building—"
"They'll know we've made contact."
"Yes."
"That's acceptable." I looked at her. "If the watcher is above Moriwaki and is monitoring the architecture, them knowing we've contacted Yaoyorozu through a legitimate consulting engagement tells them we're approaching her as a professional resource. That's the correct reading of what we're doing. We're not hiding the approach — we're making the approach look exactly like what it partly is."
"The visible part."
"Exactly."
She wrote something else. "Hatsue wants to know if the Shinkawa analysis should expand to include the specific contracts where Yaoyorozu's name appears in the specifications."
"Yes. Flag any contract where her structural authority sign-off is present in a project that also involves a Beppuken-network company. Cross-reference the timing — I want to know whether her involvement precedes or follows the Beppuken incorporation on each project."
"That's the sequence question."
"That's the sequence question. If she always comes first, she's being selected after the fact. If the Beppuken companies come first and she's brought in later, someone is specifically recruiting her involvement."
Toga wrote it down. "Two very different problems."
"Two very different problems with two very different response structures. Yes."
She stood and gathered her notes and paused at the door with the particular pause that meant she was going to say something she'd decided to say despite the fact that she wasn't sure it was welcome.
"Kuro."
"Yes."
"When you were on the call. Was she what you expected?"
I considered the question. It was a good one, which is why she'd been deciding whether to ask it.
"No," I said.
Toga nodded once, like that was the answer she'd expected, which was its own kind of read, and went back to the front office.
---
The training session that afternoon was a deliberate deviation from the protocol.
Not a subject-and-command session. Those were the metric-building work, the data collection. What I did instead was spend two hours in a public space — a mid-sized café near the Mizuno Street ward, deliberately within the potential observation zone because I wanted to use the time usefully and I wasn't going to let the observation pattern control my behavior — and I practiced the baseline skills the hostage negotiation book had described as the precondition for everything else.
Labeling. Strategic silence. Mirroring.
Not on anyone specific. On everyone in the café in rotation, passively — the table dynamics, the way conversations shifted when one person mirrored another's posture unconsciously, the way silence in a two-person conversation created different pressure depending on which person had been speaking last and what they'd said.
It was the kind of practice that doesn't produce data. It produces calibration. A finer grain on the instrument.
I'd been measuring Puppeteer's development in numbers because numbers were honest and resisted self-delusion. But numbers only captured what was measurable, and what was measurable was only part of what was developing.
The thing that was also developing, that I didn't have a metric for, was harder to name. A kind of resolution, maybe. The way a lens achieves focus not by adding something but by aligning what's already there. I was becoming more accurate in real-time reads without working harder at them, which meant the underlying processing had improved rather than the conscious effort.
I thought about Yaoyorozu's six-second silence and what I'd heard in it.
Then I thought about what she'd heard in mine.
That was the part I hadn't been tracking. The bidirectional nature of the read. She was precise, and precision in a technical domain tends to transfer — people who've trained themselves to hear the difference between a calculated structural load and a real one often hear the difference in conversations too without being able to articulate what they're hearing.
She'd asked for a name before catching herself.
She'd caught herself because she'd recognized that the question was premature — that she didn't yet have the basis to expect the information, and that asking for it before the basis was established would change the dynamic in a direction she hadn't decided on yet.
That was a fast read in a short call. Faster than most people do it.
I needed to bring the same quality to Friday that she was going to bring to Friday.
I paid for the coffee and left without looking toward the observation point and walked back through the afternoon Musutafu crowds with the calibration running quietly underneath everything else and the October light doing its low-angle thing off the east scaffolding.
The city still hadn't decided what it was going to be.
Neither had I, entirely. But the shape of the decision was becoming clearer, the way structural load distribution becomes clearer when you know where the weight is actually sitting.
---
The migraine came at eight that evening from the morning's session, delayed as usual now — the time gap between use and onset had become reliable enough to function as a planning window. I took it lying down with the compress and the timer and the particular discipline of not fighting it.
Five-point-one.
Forty-one minutes into the timing it broke the way they break now, not gradually but in a drop, as though something had been held at tension and then released.