Off the record, though he asked me twice why it needed to be. Blademaster Stonemantle. A training yard in the Chiming Snow, after evening forms. My own notes, made at the time, are in brackets.

(Stonemantle doesn't sit down for anything. We walked the perimeter of the yard the whole time, and he answered every question on the first ask, at length, with several digressions, most of which I've cut.)

Scholar: Do you place the same three in every child?

Stonemantle: God, no. We test them and then we choose.

(He scraped frost off the top rail with his thumb as he went past it. He did that on every circuit, apparently without noticing.)

Stonemantle: Vaulin watches and we push. Mira gets results I can't argue with, mind. I just want to know what a child is before I put something permanent in them, and you find that out faster by making life difficult.

Scholar: Difficult how?

Stonemantle: Cold. Weight. A bit of fear, though there are rules about that now, and they're good rules. We needed them.

(He caught my expression and laughed.)

Stonemantle: Nobody's frightening children in my yard. You put a nine-year-old in front of a problem that's genuinely too big and you watch what they do with it. Some of them go through. Some go round. Some go and get three friends and come back with a plan, and those are my favourites.

Scholar: And that tells you enough?

Stonemantle: It tells you more in an afternoon than a year of watching them be comfortable.

Scholar: What's your success rate?

(He laughed properly this time, loud enough that two students at the far end looked over.)

Stonemantle: Terrible. Worst of the five by a long way. Seln has the figures if you want them. She'll have brought them to a meeting at some point to make a point about us.

Scholar: You don't seem bothered.

Stonemantle: I'm delighted.

Scholar: You'll have to explain that.

Stonemantle: We aim high.

(He stopped at the rail and leaned on it.)

Stonemantle: Sablemire places what's safe. Same three in every child, and it works well enough for everyone. That's what they need. We look at a nine-year-old and we place for the practitioner we think that child could be at their best, forty years on. So when we're right about them we're very right, and you've seen what comes out of this yard.

Scholar: And when you're wrong.

Stonemantle: They mutate. Years of it, and rough years.

(He was still cheerful, which I found hard to reconcile.)

Stonemantle: I've had students in here convinced for a decade that they were a disappointment. Two tried to leave over it. One of them actually went.

Scholar: And knowing that, you'd still aim high.

Stonemantle: Look at what comes out the other end.

(He nodded at the yard.)

Stonemantle: The ones whose forms had to rebuild themselves, they're the best practitioners I've got. People will tell you hardship improves people, and that's rubbish, hardship mostly just damages people. What happened is they spent ten years living next to something that was actively changing shape, and nobody else in the Heartlands gets that education. Everyone else's three sat perfectly still their entire life. Mine had to watch theirs change and work out what it was turning into.

(He started walking again, and I had to catch up.)

Stonemantle: Sablemire produces reliable people. We produce people who've had to work out from the inside what a technique actually is, because theirs stopped making sense when they were fourteen. I know what it costs, and I'll take that trade.

The one digression I kept

(Partway round the third circuit he dropped his voice, which he hadn't done all evening.)

Stonemantle: I got one badly wrong. Girl I'd have staked my name on. Warhand, obvious, everything about her said she'd hold a line.

Scholar: And she didn't.

Stonemantle: Came out a Channeler, and it wasn't even a near miss. Nothing about her final shape had anything to do with what I put in her at nine.

(He walked a few paces.)

Stonemantle: She's doing well. She's teaching now, down south.

Scholar: How long did it take?

Stonemantle: Seventeen years. And she doesn't speak to me. That's her right.

(He shrugged with one shoulder.)

Stonemantle: That's what it costs. Seventeen years, and sometimes you lose the person.

Scholar: Is there a word for it? For a miss.

(He looked at me for long enough that I stopped writing.)

Stonemantle: Yes.

Scholar: Will you give it to me?

Stonemantle: No. And you'll have noticed by now that none of us will.

(He set off again at his usual pace.)

Stonemantle: It's the same word, Scholar. All five Houses, same word. None of us says it out loud, and we never agreed that between us at any point I'm aware of.

Scholar: Then where did it come from?

Stonemantle: Same place the placing did.

(He said this as though it were the least interesting thing he'd told me all evening.)

Stonemantle: You've been going round the five of us asking who worked out how to set a template against a child. Nobody worked it out. We were taught it, whoever taught us was taught it, and if you follow that back far enough you're out the other side of the Houses entirely and into the Ascent.

Scholar: That's a very long way back.

Stonemantle: It is. I'm not claiming to know it, mind. I'm telling you what a Blademaster gets told at forty when they hand him the responsibility, which is that this is a training method, and it's older than the Four, and we didn't invent one hair of it.

(He shrugged.)

Stonemantle: Makes sense to me. It's a method, it's got the shape of a method. Someone drilled this, in a yard, when the ground was still coming apart. They had a word for when it didn't take. We've all still got the word, and none of us knows who first said it.

Scholar: Does that not trouble you?

Stonemantle: Course it troubles me. I use it anyway, four times a year, on children. We're using something somebody else made, and nobody wrote down how it works.

(That's closer than anyone has come. It's also the first confirmation that the word is shared, where I'd assumed five separate reticences, and that changes what every silence I've logged was actually telling me.

It puts placement, and the silence around its failures, in the Age of Ascent. I record that as a Blademaster's inherited belief rather than as a finding. Hollowglass keeps records that predate the Houses, and Seln has already declined to date them for me, though she hasn't contradicted him either.)

Scholar: One more. Does it not strike you as strange that what you place has nothing to do with a child's Source?

Stonemantle: Should it?

Scholar: Most people I ask don't know that it's true.

Stonemantle: Course it's true. Fire child, water child, same first form, works the same both times.

(He was entirely unbothered.)

Stonemantle: That's the craft, that is. Any fool can teach a fire child to set things alight. Putting something in a child that doesn't care what they're joined to, that took a thousand years, and I had nothing to do with any of it. I just use it.

Scholar: Do you know how it works?

Stonemantle: Not the faintest idea. Nor does anyone. We know how to do it. Hollowglass will tell you they know why, and they're guessing too.

Scholar's note

An hour with Stonemantle produced more than four days in the Lowlands, and he attached no conditions to any of it. He isn't careless. He told me at the start that Ironvale publishes what it can, on the grounds that a frontier which keeps secrets from its neighbours eventually loses one.

There are two findings. The first is the shared word, which reframes every silence in this file. The second I didn't expect from Ironvale of all Houses. He's the only person I've interviewed who treats the Source-independence of placement as inherited competence that nobody living understands. Everyone else has treated it as too obvious to discuss. He thinks it's a mystery and has decided it isn't his problem.