Section 1
The inquiry sat for eleven months and was joint, which had never been done, and the lag made it slow and merciless in equal parts.
It did not find against her personally, which enraged about nine thousand people and was, the record notes, the only survivable outcome; the finding observed that a hundred and ninety-one thousand people had eaten for nineteen quarters on the strength of the thing she had done, and that no punishment available was proportionate to either side of that, and declined to perform one and call it justice. It quoted, because it could not help itself, four lines from year eighty-one, and everybody knew the lines and everybody read them again anyway.
What it did instead was produce forty-one recommendations, and Turi Lund-Amse counted them and laughed until she had to sit down, and had to explain the joke to eleven people on the ground who had never heard it.
Two of them stuck.
The first was that no quantity appearing in any of the four wall positions could have its definition altered except by ballot of both worlds, eleven days, published question, minority statement in the same size letters. That one was obvious and unopposed and is the sort of thing that gets written into a text and read by children for four hundred years without anybody thinking about it.
The second was the one that mattered, and it came out of the sentence Nea Achebe-Merrick had asked to have written down as the whole of her contribution.
It established the reconciliation as a standing obligation: four times a year, forever, every load-bearing reported figure on either world checked against an entirely independent physical proxy by a body with no interest in the answer, with the method published, the residual published whether interesting or not, and the whole thing performed — and this was the clause that was argued about for two of the four years, and that Ag-3 supported and the Coast Board's technical committee opposed and the year-nine classes of both worlds carried by a landslide in the indicative — by fifteen-year-olds as the autumn exercise.
Not because they were good at it. They were not especially good at it; the published figure for the first century was that a third of the reconciliations contained an arithmetic error serious enough to change the sign of the residual, and that figure was itself published every autumn in the fourth position on the wall, underneath the thing it was checking.
They were used because they were bored, and unimportant, and had nothing invested, and because the entire eight-hundred-year history of both worlds could be read as a single unbroken argument for the proposition that the only reliable auditor is a person with a sore back at three in the morning who has been told to check it again and bring their working.
They called it the reconciliation and then, within about forty years, nobody called it anything; it was just the autumn, the way the wall was just the wall.
Section 2
Nea Achebe-Merrick resigned the chair, which nobody had required, and went back down, and spent the last eleven years of her working life on the coast in the ordinary rota, above gate, in a suit, landing mat.
She published her own hours. That was not required either. Every quarter, in the ordinary weekly reduction, on the ordinary Fourthday, under her own name: hours above gate, this quarter, cumulative, with the panels' current fit applied and the resulting shift in her own median onset stated in years, to one decimal place, with the error bars.
She did it for forty-four quarters. The last one, filed from the eleventh camp in the autumn of year eight hundred and ninety-six, when the cumulative figure had gone past the point where the fit stopped being kind, reads in its entirety:
Q4. 61.9 hrs. Cumulative 2,411. Median onset now 63.1 (58.4–71.9). This is four years and two months earlier than when I signed.
I am not doing this as penance and I would be grateful if nobody wrote it up that way. I am doing it because it is the number and it should be on the wall, and I spent nineteen months arranging for it not to be, and the only correction available to me is to put my own on it every quarter until I stop.
Do not tidy me.
N. Achebe-Merrick, 11th camp, coast
She died in year nine hundred and four at sixty-six, at the eleventh camp, with the fold well advanced and the last four months of her quarterlies filed by dictation and published unedited, including the two that repeat.
They wrote her name on the wall on Deck 5, four hundred thousand kilometres above where her body was, in the ordinary way, in whatever hand was holding the stylus, and there was an argument about it in the corridor beforehand — there is always an argument, it is practically a rite — and the argument was settled in the way it had been settled since year eighty-two, which is that the wall is for everybody, in the order they went, and that whoever they were, they get to be in the same book as somebody's mother.
Somebody put an apple on the deck plate underneath, and somebody else cut it into eight pieces with a knife and gave one to each of the seven people standing nearest and kept the last and put it in her pocket, and nobody asked why, because everyone in the corridor understood that some things are kept.
Section 3
Turi Lund-Amse was fifty-one in year nine hundred and one and had not left the ship.
She had meant to. She had come up on a nine-year rotation at twenty-six and had signed for nine more at thirty-five and then had stopped signing anything and had simply gone on being awake at 0217, which is how the third shift has always retained people and always will. She ran the imager. She ran the reconciliation four times a year with whatever fifteen-year-olds were up on the autumn lift, and she was, by all accounts, an indifferent teacher who was extremely good at not helping.
On the last night of the autumn of that year she went up the spine ladder at 0409 with a girl from the fourth camp who had spent the fortnight being sick from the spin and photographing pages nobody had photographed yet, of which there were, at the current rate, another two hundred years' worth.
Ninety-one rungs. A hook for a lamp at the bottom, and a shelf that somebody had put up in year six hundred and forty for pencils, which was not empty, and which nobody knew who filled.
"Feet against the stringer," Turi said. "Not the pipe — that."
"I know," said the girl.
"Everybody knows."
The crawl was the same. Hot dust, old solder, the coolant run warm against the shoulder and ticking as it warmed, and behind them along the curve of the hull the paper: taped and re-taped, layered, soft as cloth in the deep places, nine hundred and one years of other people's arithmetic running away past the reach of the lamp further than anybody had ever followed.
The girl pulled the array feed, because it was the first thing on her home screen the way it was on the home screen of everyone in two worlds who had ever been fifteen.
TEV-9 flickered in its little grey box. Long, long, short, short, long. The stutter. Then the gap.
She counted it under her breath, which everybody did, and got thirty-eight, and marked it on the slate four times and got 38.11, 38.09, 38.12, 38.10, and wrote that down without any particular ceremony, because the period had been falling for a hundred and ninety years and a girl of fifteen had first caught it at forty point six with a sore back and had thought she had miscounted.
"Still going," she said.
"Still going." Turi had her head back against the skin. "It'll bottom out in a few thousand years, probably. Somebody'll be up here."
The girl was quiet for a bit, and then said: "Is it true nobody's ever been to the bottom of the paper?"
"It's true nobody admits to it."
"Somebody must have."
"Somebody must have, and whoever it was had the sense not to say so, which I've always thought was the most impressive thing anyone's ever done in this crawlspace." Turi shifted her shoulder off the pipe. "There's a paper by a laundry technician in year six-fifty-two arguing the nineteen pages aren't down there at all — that they came out during the disclosure century and the story grew over the gap. It's a good paper. There's a better one from year eight-eleven that says it doesn't matter, and I think that one's right, and I'd rather you read it than went looking."
"Why doesn't it matter?"
"Because the thing she left wasn't the pages." Turi nodded at the dark. "It was that everybody after her also left something, and none of it's indexed, and you had to come up ninety-one rungs and be warm and uncomfortable and bored to read any of it. That's the artifact. The pencil's the point. She'd have hated a relic; she voted her own name off a door four times."
The girl thought about that, and then reached back along the hull, behind the coolant run, into the dark, and found the paper, and took a pencil off her ear.
She wrote, on the back of a fault sheet, badly, because you cannot write well wedged into a crawlspace:
Yr 901 d19, 0409. TEV-9 gap 38.10 s. Still falling. Everyone still hums 41.
Reconciliation this autumn: coast hours above gate, reported 62.1, inferred from waste heat 61.7, residual within error. I got the plant library wrong the first time and had to do it again.
Anyone can check this. The array logs are open. They have always been open.
Bring a pencil. There's a shelf.
S. Vaan-Ouma, yr-nine, 4th camp, Ouma
She taped it up with the rest, where a hand would find it and where nobody would ever index it, and Turi Lund-Amse took the slate and logged the visit — inspection, no fault found — which was true in every word.
Then they went down the ninety-one rungs and out along the amber corridor toward Ag-3, where the false dawn was coming up in the grow-lamps and the night crop crew would be arguing about something, because there was a filter on the list, because there was always a filter on the list.
On the way the girl hummed it, and got the gap wrong at forty-one, and did not correct herself, because by then nobody did.
By year nine hundred and seventy-seven there were two hundred and forty-one thousand people on Ouma and a hundred and ninety-one aboard the Ambit, and the standing complement had been below three hundred for so long that the ballot of year six hundred and eight — not fewer than four hundred, drawn by voluntary rotation from ship and settlement, funded as a first charge on the fabrication of both — was taught in the autumn exercise as an example of a decision that had gone on being obeyed in its letter for three centuries after it had stopped being obeyed at all.
That was published. The shortfall was in the fourth position on the wall in the Long Gallery under the hours above gate, and it had been published every quarter for a hundred and eleven years, and the ground had voted four separate times to make it good and had four separate times funded something else instead, and every one of those ballots was in the archive with its minority statement in letters of the same size.
Section 4
The ship was old. That was also published, in a way, and had been for six hundred years: the spine fatigue index, a single number, plotted quarterly against a model, with its residual underneath. The residual was flat. It had been flat since year three hundred and ninety-one. It was the flattest line in either world and it was on page nine of the year-nine text as the worked example of a well-behaved system, and a hundred and forty generations of fifteen-year-olds had glanced at it on their way to the interesting part of the exercise.
Kesia Roque-Anselm was seventy-one and had been awake at 0217 for forty-four years.
She had come up at twenty-seven on a nine-year rotation and had signed for nine more at thirty-six and had then stopped signing anything, which is how the third shift has always retained people and always will. She had been born at forty degrees south. She had been named for a woman who chaired a Council in year sixty-eight and gave an order at 0340 on the eleventh of the ninth of year sixty-nine, and the name had become common in the settlements in a way nobody planned and nobody prevented, and there was a small unpleasant literature about why, and Kesia had read all of it at nineteen and had decided to go on being called Kesia.
She had gone up the ladder to Frame 44 something over eleven thousand times.
What she had that nobody else had was not skill. She was a competent hull tech and had never been anything better, and she said so, at length, to people who tried to make her into something. What she had was that she had heard the same crawlspace come up to temperature four thousand times in the dark.
The coolant run ticked as it warmed. Everybody knew that; it was the first thing anybody wrote about the crawl, it was in the facsimile pages by the hundred — hot dust, old solder, the coolant run ticking as it warms. It ticked because the pipe grew four millimetres along its length between the cold of the shadow pass and the heat of the sun side, and because it grew against four clips that gripped and slipped and gripped, and every slip was a tick, and the ticks came in a run of eleven as the pipe took up its length, spaced out at the end the way a dropped coin settles.
On the ninth night of the autumn quarter of year nine hundred and seventy-seven, wedged in with her shoulder against the pipe and nothing much on her list, Kesia Roque-Anselm counted nine.
She did not think anything at all for about a minute. Then she thought: I've miscounted.
She waited seventy minutes for the next shadow pass and counted it cold and counted it warm and got nine. She went down and came back on the following night and got nine, and on the night after that she brought a contact microphone she had signed out for a different job and got nine, and the intervals were wrong too — the first four came faster and the last one did not come at all, so that the run stopped short, like a phrase with the end bitten off.
She sat in the dark with her hand flat on the inner skin, which was the same skin, and which hummed, and thought about the fact that a pipe slips against a clip when the pipe grows faster than the frame it is clipped to.
It should have ticked eleven times.
That was all, at first. A technician's small objection.
Section 5
She went to Frame 31 because it was eleven metres aft and because she had shimmed the clip there herself in year nine hundred and fifty-one and had thought nothing of it at the time.
The clip was singing. Eleven hertz, on sustained load, the way it had been singing intermittently for six hundred and sixty-four years.
Kesia pulled the fault history. It was open; it had always been open; nobody had ever needed to petition anything to read a stringer clip's life, and nobody ever had, because a stringer clip's life is the single least interesting document in the possession of two worlds.
It ran to forty-one entries.
Yr 313 d1. Clip singing 11 Hz under new load. Shimmed. Somebody check it in a hundred years. — D. Roque-Amse, third shift.
Yr 411. Singing 11 Hz. Replaced. — Y. Amse.
Yr 524 d1. Singing 11 Hz again. There is no load on it. Shimmed anyway. Somebody check it in a hundred years. — I. Achebe-Ouma.
And on, and on, at intervals of thirty and forty and nineteen and sixty years, forty-one times, in forty-one hands, every one of them a competent person doing a competent job at four in the morning: shimmed, replaced, packed, shimmed, replaced. Four of them had written somebody check it in a hundred years. One, in year seven hundred and two, had written this is the fourth shim in my own service and I do not understand it, and had shimmed it.
Nobody had ever asked it why.
Kesia Roque-Anselm sat on a stringer in the amber dark with a slate in her hand and understood, with a completeness that made her feel physically cold in a compartment at twenty-one degrees, that for six hundred and sixty-four years a piece of this ship had been trying to say something and had been silenced forty-one times by people who were good at their jobs.
That was the finding. Everything that came after was arithmetic.
Section 6
She took nine days, because nine days was the interval and had been since before anybody, and because she was seventy-one and had been told by two separate directors of the Bureau of Verification over the course of her life that the only thing that ever went wrong with a finding was hurry.
The first thing she checked was the strain gauge archive on the spine, frames 31 to 61, six hundred and eighty-six years of it, quarterly, open, indexed, and beautifully behaved.
The residual against the model was flat. She sat with the plot for an hour and could not make it be anything but flat, and had the brief unearned happiness that everybody who has ever done this gets and that everybody writes down afterwards, and she wrote it down.
Then she went and read the model.
The spine fatigue model was a nine-parameter fit. It had been written in year three hundred and eighty-one for a ship under a twentieth of a gravity, revised in year four hundred and forty for a ship in orbit, and it was refitted — this was the part Kesia read four times and then went and stood in a corridor — refitted annually, automatically, against the incoming gauge data, and the residual published each quarter was the residual of the data against the fit that the data had just produced.
For five hundred and thirty-seven years the Ambit had been checking whether its own measurements agreed with a curve drawn through its own measurements.
There was no dishonesty in it anywhere. Kesia went looking, and looked for four days, and found the whole provenance in the open: a technical committee in year four hundred and forty that had adopted annual refit because the ship's thermal environment was changing under deceleration and a static model would have been useless; a note in the minute observing that the residual would thereafter be a measure of scatter and not of model error, and recommending an independent five-yearly validation against a physical proxy; the validation performed in year four hundred and forty-five, and four hundred and fifty, and four hundred and fifty-five, and four hundred and sixty; and then a line in the year four hundred and sixty-four workplan, four words long, in the same flat administrative register that had once written no operational requirement against a cometary survey: validation to biennial cycle.
And then to five-yearly again, and then to eleven-yearly, and then, in year five hundred and one, in a schedule consolidation that ran to nine hundred items and was published in the ordinary weekly reduction on an ordinary Fourthday and read by nobody, off the list.
Four hundred and seventy-six years. The instrument had never lied. The gauges were good; four independent laboratories confirmed them inside a fortnight. The number had never been falsified and no one had ever concealed anything, and there was not one human being anywhere in it who had made a decision they would have been ashamed to publish, and that was the whole trouble.
What the physical proxy said, when Kesia Roque-Anselm finally built one — a laser interferometric span across the spine's inner skin between frames 39 and 51, cobbled up in two nights out of survey spares, a bench-grade instrument that any competent hull tech could have built in any century — was that the frames in that section had lost eleven percent of their bending stiffness.
That does not sound like much. It is the difference between a coolant pipe that ticks eleven times and one that ticks nine.
She ran the crack-growth arithmetic on paper, with a pencil, three times over two nights, because she had stopped trusting anything with a processor in it, and she got a through-thickness date in the inner skin of frames 41 to 47 of somewhere between year one thousand and forty-one and year one thousand and ninety, with a wide interval and an ugly tail.
Sixty-four years. Not centuries. And the worst of it, the deepest of the growth, was at Frame 44, because Frame 44 had a warm coolant run laid four centimetres off the cold skin, and had had for nine hundred and seventy-seven years, and every shadow pass of every orbit for five hundred and forty of those years had put a thermal gradient across that skin and taken it off again, two point six million times.
The crawl was warm. That was the first thing anybody said about it. It smelled of hot dust and old solder and it was warm, and everyone from the ground remarked on it, and the warmth was the thing that had been eating the hull for as long as there had been anybody to be warmed by it.
Kesia had the whole reduction repeated independently by the Ag-3 night crop crew of the third rota, four of them, using duplicate instruments under a beans rota, because Turi Achebe's nine pages had been law since year eighty-four and applied to a hull tech of seventy-one exactly as they applied to everything, and the crew got twelve percent and an argument about the span calibration that took two more days and improved the estimate.
She published on Fourthday, in the ordinary weekly reduction, at 0400 ship-time, with a four-line covering note.
Then she went and cleaned a filter in Ag-3, because it was on her list, and the filter was a filter, and the false dawn was coming up in the grow-lamps, and two of the night crop crew were arguing about whether the mat counted as a vegetable.
Section 7
The argument took four years and two months.
Nobody remarked on that either. There is a paper about it, from a laundry technician in year eight hundred and eleven, and it is one of the good ones, and its conclusion is that four years and two months is simply the length of time it takes a quarter of a million people to say everything they have to say twice.
There were three positions and everybody knew all three by the end of the first month.
Sleeve her. Cut and fit a structural sleeve over frames 39 to 51, nineteen segments, in the same way and to the same tolerances as the interferometric truss had been built five centuries earlier by people whose names were on the wall on Deck 5. Cost: eleven years of both worlds' fabrication as a first charge, which meant eleven years of no new cold tunnel at the four far settlements, eleven years of the coastal landing gear running on the old plant, and — this was the number that did the work, and the Coast Board put it on the wall in the fourth position under the hours, because that was the law now — eleven years at an estimated additional four point one hours above gate per capita per year, because a fabrication charge is paid in labour and labour on Ouma is paid in exposure. Buys two hundred years.
Seal her. Pump down frames 39 to 51, weld the bulkheads at 38 and 52, and let that section of spine be vacuum forever. Cheap. Four months of work. Costs the crawl above Frame 44, and everything taped inside it.
Bring her down. Reverse the ballot of year six hundred and eight. Recover eleven thousand tonnes of hull stock over forty years, which was more structural material than Ouma had fabricated in its entire history, and which the far settlements had been quietly and not so quietly wanting for two hundred years.
The case for coming down was made in the fourth session by a woman of thirty-eight from the eleventh camp, and it was not sentimental and it was not cruel, and it is on the syllabus.
"I want to answer the orchardist," she said. "The one from Ag-1, year six hundred and eight, with the shavings in her hair. Everybody quotes her and nobody has answered her and I have had three hundred and sixty-nine years to think about it, which is more than she had.
"She said: in four hundred years there will be a generation that has never seen this ship, and somebody in a room will have a very good reason to make something simpler than the truth, and when that happens I want there to be a thing in the sky they can point an instrument at.
"I am that generation. I have never seen the ship. My mother never saw it and my grandmother never saw it. And I would like the record to show that what has kept us honest for three hundred and sixty-nine years is not the object in the sky."
Eight seconds.
"It's the autumn. It's forty-one settlements doing the reconciliation four times a year with a third of them getting the sign wrong. It's the wall in every assembly hall and the minority statement in the same size letters and a woman at the eleventh camp who published her own median onset every quarter until she couldn't hold a stylus. Those are ours. We built them down here, in the cold, and we would go on doing them if that ship burned up tomorrow, and if we wouldn't then the ship was never what was holding us up. "And here is the part nobody up there will say to your faces, so I will. There are four settlements that have been waiting eleven years for a cold tunnel. Not for comfort. For four degrees. There are children in the far south sleeping at six because the plant they were promised in year nine hundred and sixty-one was deferred, and six is not four, and the panels' fit is not a metaphor. That is exposure, and it is accruing tonight, and it is accruing in children.
"So when you cost me eleven years of fabrication to sleeve a hull, I am going to make you write down what the eleven years buys and what it costs in the same units, because that is the whole discipline and this ship invented it. Two hundred years of an archive in orbit. Against four point one hours above gate per capita per year for eleven years, and four cold tunnels that do not get built.
"I have done that arithmetic on paper with a pencil and I did it four times and I want somebody in this hall to tell me which term in it is wrong."
Eight seconds, and then, from a hall of a hundred and ninety-one people four hundred thousand kilometres up, nothing at all came back, and the chair, after some time, moved to the next speaker.
Section 8
Sealing was the position of the technical committee and it was the position that made everybody unhappiest, because it was cheap and sound and correct and it took the paper.
There were four thousand three hundred and eleven pages in facsimile by then, or rather there were eleven thousand and six, because the plate-camera work had gone on for three hundred and sixty-two years at the pace of a fortnight's autumn exercise and had reached a depth of one thousand nine hundred and forty on the marker, and the published estimate said another hundred and ninety years to the end at current rates, and that estimate had been revised nine times and never once met.
Everything above depth nineteen-forty existed in facsimile in every settlement school and every ship-side lab. Everything below it existed once, taped to the inner skin of a hull that was going to open in sixty-four years.
The proposal to go in and get it — to strip the crawl, sheet by sheet, before the weld — was tabled in the first month, and it was reasonable, and it was defeated in the ninth by seventy-nine percent, and the sentence that beat it was said by a conservator of forty-one who had spent her working life arguing for exactly the opposite.
"I have wanted that paper indexed for nineteen years," she said. "I have filed four proposals. I have lost all four and I was right all four times about the risk and wrong about everything else.
"Here is what I have finally understood, and it took a stringer clip to teach me. That crawlspace has been read continuously for nine hundred and seventy-seven years by people who had to be uncomfortable to do it. Forty-one hands shimmed a clip and none of them asked why, and the reason we caught it is that a woman of seventy-one had heard the pipe tick four thousand times.
"You cannot get that out of a facsimile. You cannot get it out of an index. It is not information. It is a person having been in a place enough times to notice that it had changed, and there is no archive in the universe that stores that, and the only way to keep it is to keep sending people up ninety-one rungs with a pencil.
"So: if we seal, we do not strip it first. We do not go in and save the relic. We photograph what the autumn reaches in the time it has, at the ordinary pace, by fifteen-year-olds, and whatever we do not reach we do not reach, and the weld goes over it.
"I hate that. I want it minuted that I hate it. But the alternative is that we spend four months teaching two hundred and forty thousand people that when a thing is finally about to be lost, we stop doing it the slow way and send in professionals — and that lesson would outlive the paper by a thousand years and it would be the worst thing we ever taught them."
Section 9
The ballot ran eleven days in the autumn of year nine hundred and eighty-one. Turnout was ninety-six percent, joint, and the question was three sentences long because the drafting committee had lost an argument about brevity.
Sleeve carried with fifty-four point one percent.
Not sixty-one. Nobody had expected sixty-one for eight hundred years; the superstition had died in year one hundred and forty-two along with the woman who had been wrong about it, and it was on page nine of the text as the worked example of an anchor destroyed by publication, and there was always somebody in the Long Gallery on ballot night who said it's always sixty-one as a joke, and it never was.
Fifty-four point one. Four percent of a quarter of a million people had voted to bring the ship down and their reasons were in the record under their own names, and the minority statement went onto the plate in the same size letters, and the load-bearing part of it was the woman from the eleventh camp's four point one hours, printed exactly as she had said it, in the units she had demanded.
The four cold tunnels were funded in the same ballot, as a separate question, by ninety-one percent, out of the settlements' own charge and eleven years late.
Section 10
They cut the first sleeve segment in year nine hundred and eighty-four and fitted the nineteenth in year nine hundred and ninety-five, and four people died doing it, and their names are on the wall on Deck 5 in the order they went, in whatever hand was holding the stylus.
The crawl above Frame 44 was open the whole time. That had been a condition, argued for two years, carried by a majority so large that the drafting committee had trouble filling the space beneath the minority statement: the sleeve went on the outside, the working faces were rerouted, and at no point in eleven years was the ladder closed to the autumn exercise.
They moved the coolant run. That was the actual engineering answer and it took nine months and it was the single most consequential thing anybody did to the Ambit in its second millennium: the warm pipe came off the inner skin at frames 39 to 51 and went up onto the stringers, eleven centimetres out, with a radiation shield between, and the thermal gradient across the skin at Frame 44 fell by a factor of nine, and the crack-growth date went from year one thousand and forty-one to something in the four thousands with an interval so wide that the panels declined to quote a median.
The crawl was cold after that.
Nobody had thought about it. It was in nobody's assessment and nobody's minority statement, and the first autumn class to go up after the reroute — nineteen of them, in year nine hundred and ninety-six, from four settlements — came down at the end of the fortnight and reported it as a fault.
Kesia Roque-Anselm was ninety and did not go up any more and heard about it in the Bureau room, where somebody read her the girl's log entry out loud: Crawl above Frame 44 is at four degrees. Feels like home. Recommend investigation.
She laughed until she had to sit down and could not explain it to anybody in the room, all of whom had been born on the ground.
"You don't understand," she said, when she had got it back. "The whole of it. Nine hundred and ninety-six years of children coming up ninety-one rungs to be warm and uncomfortable, and every single one of them wrote it down, hot dust and old solder and the pipe ticking — and we fixed the hull and now it's four degrees up there and a girl from the fourth camp says it feels like home."
"Is that good?"
"It's not good or bad. It's just the last thing that was still the same." She wiped her face with the back of her wrist, like a woman clearing a work surface. "Somebody write it down."
Section 11
Somebody wrote it down. Everybody wrote it down; that was the point and had been for nine hundred and ninety-six years.
The pipe stopped ticking. That went into the fault log as a closure — coolant run 39–51 relocated, thermal cycling on inner skin reduced by 9x, tick signature no longer present, this entry closes the anomaly first reported yr 977 by K. Roque-Anselm, third shift, hull & thermal — and the anomaly's whole life ran to four lines and forty-one prior shims, and a hull tech in year one thousand and eleven, reading the file for an unrelated reason, wrote in the margin of her own working: forty-one people fixed this and none of them asked. Do not be the forty-second.
That sentence went into the year-nine text in year one thousand and forty, on page nine, in the margin, where it is now.
Frame 31's clip sang for another two hundred years and was shimmed nineteen more times, by people who all knew the story and all wrote investigated, no underlying cause found, shimmed, because there genuinely was no underlying cause; it was a clip with a resonance, and it had been singing since a woman shimmed it on the first night of a deceleration burn six hundred and eighty-three years earlier, and it is singing now.
Section 12
Kesia Roque-Anselm died in year one thousand and two at ninety-five, aboard, in a chair by a window at eye height, which had been the arrangement for the old and the dying since before there was anything else to arrange.
There were nineteen windows in the Deck 2 annex and there had been since year eighty-six. She was at the fourth one, which looked out along the spine at nothing in particular: a great deal of dark, and stars, and off to one side, if you had the geometry and a good decade, a dim thing that nobody bothered to point out any more because there was no vector for it to be nineteen degrees off.
She had asked, in the last month, for the sports and got them, and for the array feed and got that too, and in the last fortnight, when she could not hold a slate, a hull tech of nineteen sat with her at 0217 and counted the gap out loud.
Thirty-six point nine.
"It's still going," Kesia said.
"It's still going."
"Everybody still hums forty-one."
"Everybody still hums forty-one." The girl had the ship's own habit of not filling silences and was good at it. "There's a paper this year saying we should stop and teach the current value and revise it annually."
"There's a paper like that every eighty years."
"It got sixty-one percent in the indicative."
Kesia Roque-Anselm laughed, which cost her something, and said: "Then it'll fail in the real one," and it did, four years and two months later, and the little star went on being hummed at wrong by a quarter of a million people, and page four kept 41.00 with a date attached, and the marginal line beneath it kept saying the song is not the measurement.
They wrote her name on the wall on Deck 5 in the ordinary way, in whatever hand was holding the stylus, which turned out to be a survey technician of twenty-four who got the K slightly wrong and was sick with fright. Nine hundred people came, which was five times the standing complement, most of them up from the ground on the autumn lift, which had been rescheduled by four days for the purpose and which was the only concession anybody made.
Somebody put an apple on the deck plate underneath, and somebody else cut it into eight pieces with a knife and gave one to each of the seven people standing nearest and kept the last and put it in her pocket, and nobody asked why.
Section 13
The last thing to say about Frame 44 is that nothing happened there.
That is the whole of it and it is the reason the place survived a thousand years: it was a crawlspace on a spine with a bracket in it, and the bracket had been reseated in year eighty by a woman of thirty-nine who had come up to do a job and had brought her slate out of habit, and the bracket was fine, and had been fine for nine hundred and twenty years, and was checked every autumn by somebody who put a hand on it and wrote no fault found.
In year one thousand and eleven the autumn exercise reached depth two thousand and forty on the marker.
The girl who took the photographs was seventeen, which was two years late, because she had failed the derivation twice at fifteen and had come back for it, which was permitted and unremarkable and happened to about one in nine.
She was from the far south, from one of the four settlements that had waited eleven years for a cold tunnel and had got it in year nine hundred and ninety-two, and she had been born in the year it opened, and she had never in her life slept above four degrees.
The crawl was cold. She wedged herself in — feet against the stringer, not the pipe, that, although the pipe was eleven centimetres away now and cold and nobody could tell you why the instruction had survived — and she photographed pages at eleven an hour with the lamp moved four times a page, and it was tedious in the way she had been promised and had not believed.
At depth two thousand and thirty-eight she found four sheets in a hand she recognised, because everybody recognised it, because two of the four were on page nineteen of the text: a hull tech turned biologist who had gone down at thirty-six and had not come up, three of them wrong.
She photographed them, including the wrong ones, because you do not tidy people.
At depth two thousand and forty she stopped, because the fortnight was over, and she reached back into the dark past the marker with her arm to the shoulder, and felt paper going away from her further than she could follow, and did not follow it.
She had brought a pencil. There was a shelf at the bottom of the ladder that somebody had put up in year six hundred and forty and it was never empty and nobody knew who filled it, and it is one of the four genuinely unsolved questions in the archive of two worlds, and there are papers.
She wrote, on the back of a fault sheet, badly, because you cannot write well wedged into a crawlspace:
Yr 1011 d19, 0409. Autumn plates to depth 2040. Reached I. Ouma's four sheets, photographed all four incl. the wrong ones.
Bracket at Frame 44: no fault found. It's been fine for 931 years, which everyone knows, which is why I put my hand on it.
Pipe is cold. Crawl is at four degrees. I'm told that's new. It feels like home and I'm told that's funny.
TEV-9 gap 36.4 s. Everyone still hums 41. I hum 41.
Anyone can check this. The array logs are open. They have always been open.
Bring a pencil. There's a shelf.
A. Achebe-Lund, yr-nine (repeat), far south, Ouma
She taped it up with the rest, where a hand would find it and where nobody would ever index it, and took the slate and logged the visit — inspection, no fault found — which was true in every word.
Then she went down the ninety-one rungs, past the hook and the lamp and the shelf with its pencils, and out along the amber corridor toward Ag-3, where the false dawn was coming up in the grow-lamps and the night crop crew would be arguing about something, and where there was a filter on her list, because there was always a filter on the list.
On the way she hummed it.
Long, long, short, short, long. A stutter. Then a gap of forty-one seconds — which was wrong, and had been wrong for three hundred years, and which she knew was wrong because she had counted it herself four hours earlier and got thirty-six point four, and which she hummed at forty-one anyway because that was the song, and because a woman on third shift had memorised it in nine hours in a crawlspace in year seventy-six the way you memorise a song you hear during a bad week, and had hummed it, and had written in her personal log 41s, call it 41 exactly, cute.
A star, keeping its own time, a thousand years and eleven light-years behind, indifferent, unhurried, changing at its own pace for its own reasons and telling nobody, having been made a clock and then a metronome and then a deadline and then a song — and remaining, through all of it and for as long as anybody kept climbing ninety-one rungs at four in the morning to look, simply something that anyone could see.
By year one thousand and thirty-nine there were three hundred and eleven thousand people on Ouma and a hundred and forty aboard the Ambit, and the four wall positions had not changed in two hundred years: the world, live, with its weather; the phase box, with a star whose gap was now thirty-five point one seconds and falling and which everybody still hummed at forty-one; the four air chemistries from four laboratories forbidden to share instruments; and, last, where the eye rested, hours above gate.
It was 64.9 that autumn, and rising, and it had been written by a girl of fifteen with a stylus while eighteen others watched her hand, and the reconciliation underneath it — inferred from waste heat by the autumn class of the ninth rota, four of whom had got the sign wrong and said so in the note — read 64.6, residual within error.
Section 14
The ship was sound. The sleeve had held for forty-four years and the model that watched it was validated five-yearly against a physical proxy by a body that had no interest in the answer, which was recommendation eleven of the inquiry of year nine hundred and eighty-one and was the least glamorous thing that inquiry ever did and the only one anybody would have missed.
The crawl above Frame 44 was at four degrees and had been since the reroute, and every autumn class that came up wrote that it felt like home, and by then nobody found it funny except historians.
Bakri Ouma-Vaan was sixty-one and had been awake at 0217 for thirty-four years.
He was radio and thermal, which meant the ground link, which meant the two dishes on the dorsal truss that carried nine hundred channels of session traffic and school and weather and the weekly reduction down to a world four hundred thousand kilometres below with a lag of four seconds each way, and it meant the radiators that took the heat off the transmitter chain, and it meant, four times a year, a housekeeping document that nobody had ever read for pleasure.
The system noise temperature log.
Every receiver in the universe has a noise floor and it is measured in kelvin, because noise is heat and heat is the one thing that always arrives. The Ambit's ground-link receivers logged theirs every eleven seconds and had done for nine hundred and forty years, and the log existed because a warm receiver is a sick receiver and because a hull tech in year one hundred and something had wanted to know when to clean a feed horn. It was open. It had always been open. Nobody had ever petitioned anything to read it because nobody had ever wanted to.
On the ninth night of the autumn quarter, with the transmitter down for a feed swap and nothing much on his list, Bakri Ouma-Vaan pulled a year of it to see whether the west dish was ageing, and it was, by about a millikelvin a decade, tediously, as expected.
There was also a bump.
It was four millikelvin high and it lasted about nineteen minutes and it happened once per orbit, and it had happened once per orbit for as long as the log went back, and it was so small and so regular that every automated cleaner in nine hundred and forty years had classified it as a spacecraft artefact, which is what you call a thing that recurs with your own period.
Bakri looked at it for a while.
Then he thought: it should have moved.
That was all, at first. A technician's small objection, and the whole of what followed came out of it. The Ambit's orbit was not fixed in the sky. It precessed. Slowly, four degrees a year, the way any orbit does around an oblate world, so that the plane of it turned like a slow hand — and a thing that happened once per orbit because of the spacecraft would keep happening at the same place in the orbit forever, and a thing that happened once per orbit because the sky was in a particular direction would creep.
He pulled forty years and fitted it.
It crept. It crept at four degrees a year, exactly against the precession, which meant it was not in the ship. It was outside, and it was fixed in the sky, and it had been sitting in the same place in the sky for nine hundred and forty years while a receiver's fourth sidelobe swept across it once every ninety-one minutes and warmed by four millikelvin and nobody in two worlds had ever asked why.
Bakri Ouma-Vaan sat in a compartment on Deck 3 with his feet on a crate and worked out the direction with a pencil, because he had stopped trusting anything with a processor in it about an hour earlier, and the direction was fifteen light-years away and had a name that every child in two worlds learned at nine and no adult in either had said out loud in earnest for eight hundred years.
Section 15
He took nine days, because nine days was the interval and had been since before anybody.
The first thing he checked was his own instrument, and he spent one shift very happy, because the west feed had been swapped four hundred times in nine centuries and a periodic thermal artefact in a receiver is the most ordinary thing in the world. He put a source in front of the chain. He ran the east dish, which had different feeds and a different amplifier and had been built two hundred years later, and the east dish saw it too, at the same place in the sky, four millikelvin high.
He wrote the happiness down and wrote down that it had been unearned, because everybody did, because it was on page nine.
The second thing was the direction, properly, with the interferometric baseline, which took two nights of somebody else's telescope time and an apology. The source was point-like to the resolution of a four-kilometre array. It was not the star. It was within nineteen arcseconds of the star and it was not the star, and the star at that distance was a fourth-magnitude nothing that the survey people had been using as a photometric standard for six hundred years without ever once mentioning what it was, because it was on the standards list under a catalogue number.
The third thing was the spectrum, and this was where he stopped being a man with a tedious log and became a man who could not eat.
It was narrow. It was a carrier about one and a half gigahertz and it was narrower than any natural thing has ever been or ever will be, narrower than the receiver could measure, and the entire four millikelvin was sitting inside a band that a hydrogen cloud could not have produced in the lifetime of the universe.
The fourth thing was the drift, and the drift was the proof, and Bakri Ouma-Vaan understood while he was doing it that he was performing the oldest exercise in the culture that had raised him.
A signal in your instrument does not move. A signal fifteen light-years away moves in frequency exactly as much as your own motion makes it move: the ship around Ouma, ninety-one minutes; Ouma around its dim orange sun, four hundred and eleven days. Two periods, two amplitudes, both computable from published ephemerides by anybody with a pencil, neither of them adjustable by any wish of his.
He measured the ninety-one-minute term first and it was right to four parts in a thousand.
Then he pulled nine hundred and forty years of noise floor and folded it on four hundred and eleven days, because the bump's height changed as the sidelobe pattern swung, and there in the fold was the annual term, in the amplitude rather than the frequency, riding along on an orbit the Ambit had been making around a star it had chosen out of a junk catalogue.
The source was outside. It was narrowband. It had been there for nine hundred and forty years at least, which was as far back as the log went, and it was in the direction of Sol.
He had the whole reduction repeated independently by the Ag-3 night crop crew of the eleventh rota, four of them, using the duplicate radiometer that had been kept under a beans rota since year eighty-four because a woman named Turi Achebe had written nine pages and lost an argument publicly and won it in the end, and the crew got four point one millikelvin and an argument about the sidelobe model that took two more days and improved the estimate.
He published on Fourthday, in the ordinary weekly reduction, at 0400 ship-time, with a four-line covering note, and because it was 0400 the first people in two worlds to read it were hull techs and recyclers and night crop crews standing in amber corridors with a slate in one hand.
Then he went and cleaned a filter in Ag-3, because it was on his list, and the filter was a filter, and the false dawn was coming up in the grow-lamps, and two of the night crop crew were arguing about whether the mat counted as a vegetable, and one of them said, without looking up, "Is it them?"
"It's narrowband and it's fifteen light-years away in the right direction," said Bakri Ouma-Vaan. "That's all four measurements say and I'm not going to say a fifth."
"But is it them."
"Check it again," he said. "Bring me your working," and threw her a tomato, and she laughed at him, and he went to bed and did not sleep.
Section 16
Register of Consequential Omissions, entry four thousand one hundred and eleven, filed by B. Ouma-Vaan, third shift, radio and thermal, in the fourth month of year one thousand and thirty-nine:
No receiver on this vessel or this world has ever been pointed at Sol.
Not once, in one thousand and thirty-nine years. Not for an hour. There is no line in any workplan declining to do it, because there is no line in any workplan proposing it. I have looked for four days with an archivist and we cannot find the decision, and the reason we cannot find it is that there wasn't one.
This is the largest entry in this Register and it is nobody's. I want that stated plainly before anybody goes looking for a villain, because there isn't one, and the ease of it is the horror and not the malice, which is a sentence a woman said in a hall in year sixty-eight and which we have been quoting for a thousand years without noticing that it applies to omissions as well as to acts.
Here is the best reconstruction the archive supports. For eighty years the vessel believed it was in transit and there was nothing to hear from and no one who could have proposed listening without being told why. In year sixty-nine we lost the ability to return and in year eighty-two we published that we had lost it, and somewhere between those two dates Sol stopped being a place and became a direction. Nobody wrote the second date down. After that there was the disclosure century, and the tugs, and the departure, and the burn, and Kesh, and four hundred years of a world that killed at seventy-one, and at every point a proposal to build a receiver and point it back the way we came would have been answered — correctly, by good people, in the open — with the observation that we could not go there, that any news would be fifteen years stale, and that the fabrication charge would be paid in hours above gate by people who had not asked to pay it.
No operational requirement. That is the phrase. It is item four hundred and eleven of this Register, filed by I. Vaan in year eighty-three about a cometary survey that turned out to be the difference between a house and a ship, and I have read that entry perhaps four hundred times in my life as a piece of school and I did not recognise it when I was standing inside it.
I have no recommendation. I measured a noise floor.
B. Ouma-Vaan
Section 17
The receiver took eleven years.
The ballot ran eleven days in the autumn of year one thousand and forty, joint, turnout ninety-eight percent, the highest ever recorded in two worlds, and the question was one sentence long because for once the drafting committee won its argument about brevity.
It carried with eighty-eight point one percent, and the minority statement went onto the plate in the same size letters, and it was not stupid and nobody pretended it was. It said that a fabrication charge of nine years was four point four hours above gate per capita per year for nine years and that this was three hundred and eleven thousand people paying in the far end of their lives for the privilege of overhearing something they could not answer, verify, or use; that the last time this ship had built a great instrument out of longing it had been the interferometer and it had been worth it, and that this was not an argument, it was a precedent, and a precedent is what you have instead of an argument.
It was signed by four thousand and six people, which the archive notes was a coincidence and which nobody has ever entirely believed.
They built a phased array nine hundred metres across on the dorsal truss out of eleven segments over nine years, and two of the segments failed and were remade, and nobody died, which was remarked upon at the time as the first great fabrication in the history of either world to be completed without a name going onto the wall on Deck 5.
What they got, over the following two decades, was this.
The carrier was there and it was steady to eleven parts in a billion, which is a stability no natural object achieves and no drifting wreck maintains, and which meant somebody was keeping a clock.
It was modulated, extremely slowly. The bit rate, when they finally established it in year one thousand and fifty-one, was four bits a second, which is the rate you choose when you have decided that reaching a very long way matters more than saying very much, and it meant that the thing took nine days to transmit what a settlement school sent to the ship in a heartbeat.
It repeated on a cycle of four hundred and forty-one days.
And inside the cycle, when the array had been running for a decade and the error correction had been worked out by three separate groups who were not allowed to share instruments, it was a catalogue.
That is the flat word and it is the right one. It was a list. It was a very long list of vessels, with dates, with mass and drive class and declared destination, and with a field at the end of each entry that took the survey people four years to interpret and which turned out to be, when they were sure, the date on which Sol had last had any information about the vessel at all.
The Ambit was on it.
She was the ninth entry in the fourth block, launched year zero by the reckoning of a calendar that had a different zero, mass and drive class correct to within the tolerance of the published specification, declared destination Kesh, and her last-information field held a date that corresponded to about eleven months after departure, which was when the Ambit had passed out of usable range of the deep net and had begun, as every vessel of her generation had begun, to be a rumour.
There were nine hundred and forty-one entries.
There were nine hundred and forty-one, and Ouma read the list for a year, and the thing that took the ground apart was not that the Ambit was on it. It was the last-information dates. Four hundred and eleven of the entries had a date within a decade of launch, like theirs. Sixty-one had a date centuries later, which meant those ships had been heard from, which meant they had arrived somewhere or had at least been able to say something. And four had a date and then a second field which the working groups argued about for six years and finally agreed, on a majority of three panels to one, meant ended.
Nobody had ever told Sol that the Ambit had arrived. Nobody had ever told Sol that four hundred and six people died in year sixty-nine, that a world was closed, that the ship had gone on to a junk star out of the tertiary catalogue and stopped there, that there were three hundred and eleven thousand people living at four degrees on a world called Ouma with a chemistry in the ground that gave them an old age at seventy-one.
Sol had gone on listing them for a thousand years, four bits a second, in a cycle of four hundred and forty-one days, into a sky that had not once answered.
The beacon was not addressed to them. That was established early and repeated often, in sessions and in the year-nine text, because it was the load-bearing fact and it was the one people kept mislaying. It was omnidirectional. It was addressed to nobody, which meant it was addressed to everybody, which meant it was the largest and longest and most patient act of unrewarded bookkeeping in the recorded history of the species, and it had been running for at least a thousand years and had been carried out by people none of whom could have expected to see it answered.
There was a phrase for that, on Ouma. It was on a plate at the west lock of the forty-degrees-south station and it was on page nineteen of the text, and it had been said first by a woman who kept a list she could not give to anyone: shame is not a plan. But there was another one, older and less quoted, and it turned up now in every second session for a decade, because it fitted and nothing else did.
Section 18
Keeping the list is how you go on being a person who could use it.
Section 19
The argument about whether to answer took four years and two months.
It began before the catalogue was decoded and it did not wait for it, and it was, by universal agreement afterwards, the argument that the whole thousand years had been rehearsing for.
The technical part was nothing. A reply cost power, and power was reactor time, and reactor time on a ship with three point one percent in her tanks was a matter of thermal budget and radiator margin, which meant it was Bakri Ouma-Vaan's department. A hundred kilowatts into a nine-hundred-metre array would reach Sol at a signal-to-noise anybody there could work with, provided anybody there had an instrument as good as the one Ouma had spent nine years and four point four hours per capita per year building, which was an assumption the panels published with a wide interval and a note.
Thirty years to a reply, if there ever was one. Fifteen out and fifteen back and nothing at all in between, and everybody who voted would be reading the outcome at seventy or not at all.
The real question was what to say, and it took four years because it turned out to be a question about who they were, and nobody had had to ask that out loud for eight hundred years. The technical part was nothing and the four years were spent on the rest of it, and the first thing the four years produced was a piece of arithmetic that a fifteen-year-old did in an afternoon and that nobody afterwards was able to get out from under.
Four bits a second is what reaches. You can put more power into the array, and they modelled that to death, and the honest conclusion of three panels forbidden to share instruments was that a hundred kilowatts into nine hundred metres at fifteen light-years bought you four bits a second and no argument was going to buy you forty.
Four bits a second is a hundred and twenty-six million bits a year.
The year-nine text, entire, in plain language, with its figures: nine months.
The Register of Consequential Omissions, four thousand one hundred and eleven entries: four years.
The ledger — every name in the book from Bay 2 forward, four hundred and six and then a thousand years of everybody else, with dates and no cause: eleven years.
Iyabo Ouma's twenty-nine years of handwriting, in facsimile, unedited, including the pages that repeat: sixty-one years.
The transcripts of the argument they were currently having: a hundred and sixty years, and rising by about eleven months per year of argument, which somebody put on the wall in the Long Gallery as a joke in the second year and which stopped being funny in the third.
The whole holding of two worlds, everything either of them had ever measured or written down: something on the order of a hundred and forty thousand years.
So you could not send everything. That was established in the first month and it never went away, and it meant that the Ambit and Ouma, having spent a thousand years constructing an entire civilisation around the proposition that a fact's purpose is not to be chosen for you, were now obliged to choose an order.
"It's Verrat," said a Bureau instructor of thirty-four in the ninth session, and she was not being provocative; she was reading. "Not all true things are load-bearing. The purpose of a fact aboard a closed vessel is not to be known but to be useful. That is the sentence that killed four hundred and six people and it is also, word for word, the design brief for a four-bit-a-second link, and I would like everybody in both halls to sit with that for a minute before we start being clever."
Eight seconds. Nobody used them.
"The difference," she said, "the only one, and I have looked for a bigger one for two years — is that the framework was applied to us by forty people in a room, and this will be applied by us to strangers, in the open, with the ordering itself published and votable and revisable forever. That is a real difference. It is not as large as I would like it to be."
They quoted Hana Merrick constantly that year. She had been dead for six hundred and sixty years and her Register entry — number one thousand four hundred and nine, five lines, filed by a woman who had been overruled and had published her own dead afterwards with the losing recommendation attached — was read out in session eleven times in four years, and the settlements' schools taught it that autumn to children who had previously known her only as an exam question.
I have spent five years unable to decide whether I was right, and I have concluded that this is the correct condition for a person in my office and that any successor who feels certain should be watched.
Section 20
The positions sorted themselves into four, as positions did.
Send the warning. Kesh-3's coordinates, the precursor chemistry, the reflectance signature at nine hundred and forty-one nanometres, the gate at four degrees, the panels' fit with its error bars and the word never still honestly at the far end. Nothing else. It was the only part of the holding that could keep anybody alive, it fitted inside four hundred days of transmission, and it was addressed to the only audience that mattered, which was the next nine people who would otherwise take their helmets off on day sixty and cheer.
Send the method. The warning, and then everything required to check the warning: the assay archive, the forty-one dead landers with their dates, Sarr's nine weeks, Achebe-Roque's gate, Ouma's twenty-nine years. Because a conclusion you cannot audit is a rumour, and a rumour is what a fact becomes when it has travelled fifteen light-years and lost its working.
Send nothing. Nine hundred and forty-one entries, four of them ended. We do not know what ends a ship. We have been silent for a thousand years by pure administrative accident and we are alive and there are three hundred and eleven thousand of us, and the proposal on the table is to spend nine years of fabrication to become audible.
Send the whole thing, in an order, forever. Start with the warning, because the warning is the part that saves lives; then the method, because a warning that cannot be checked is a rumour; then the Register, then the ledger, then the transcripts; and go on transmitting for as long as there is a reactor and somebody willing to run the radiators, in the certain knowledge that nobody alive will see the end of the transmission and that the ledger alone will still be going out when every person who voted is on the wall on Deck 5.
The case for sending nothing was made best in the second year by a man of fifty-one from the far south named Osei Vaan-Ouma, who ran a cold tunnel and had never spoken in a joint session before.
"I'm not frightened of the four," he said. "I want that first, because everybody expects the far south to be frightened and I'd rather disappoint them early. Four out of nine hundred and forty-one is a rate and I can read a rate. Ships end. Ours nearly ended nine times and four of those were us.
"Here's mine. That beacon has been shouting for a thousand years without an answer and we are about to be the answer. Not an answer. The first one, so far as we know. And the moment we are, we are a fact in somebody else's holding — a line in a catalogue in a room fifteen light-years away, with a mass and a drive class and a last-information date, and a place. And they will do with that line whatever people do, and none of it will be checkable by us, ever, and I would like this hall to notice that we are proposing to hand our position and our chemistry to an institution about which we know precisely one thing, which is that it keeps very good lists.
"So does the Morale Directorate."
Eight seconds.
"I know what happens next. Somebody will tell me that they cannot come here in less than centuries and cannot do anything to us that we could not do to ourselves, and that is true and it is not what I am afraid of. I am afraid that in four hundred years there will be a generation on Ouma that has been listening to Sol for four hundred years and has begun to hold its own arguments with one eye on what Sol will think of them. That's it. That's the whole of my objection. Not invasion. Deference. We are the only people in the universe who have ever had to be honest with nobody watching, and it made us, and I would like somebody to argue that we won't lose it."
Nobody managed to, in four years, to his satisfaction or anyone else's. His statement went onto the plate in the same size letters when it was over.
Section 21
What decided it was said in the third year by an orchardist of forty-four in Ag-1 on the ship, up on a nine-year rotation, with pruning shavings in her hair, and it was not an argument anybody had prepared for because it was not about Sol at all.
Her name was Delphine Sarr-Achebe and she had asked for four minutes.
"I've listened to two years of people saying what will they do with it," she said, "and I want to turn it round, because I think we've been holding the wrong end.
"For a thousand years every single thing this ship and this world has said out loud has been said to an audience that could check it. That is not a virtue of ours. It's a condition. We built the condition on purpose after it nearly killed us, and then we spent nine hundred years congratulating ourselves on a discipline that we have never once had to keep without a hundred and ninety thousand fifteen-year-olds standing behind us with pencils.
"Now, for the first time since year eighty, we are going to speak to somebody who cannot audit us.
"Fifteen years out. They will have the assay data and they will not have the world it came from. They will have our numbers and no way on this side of a century to catch us in a single one of them. We could send the warning and leave out Bay 2. We could send Kesh and leave out the twelve years. We could send a thousand years of a well-run civilisation and simply not mention that it began with forty exhausted people in a room deciding that eleven thousand others could not survive the truth, and nobody out there would ever know, and — this is the part I would like said slowly — nobody in here would ever be able to prove that we hadn't done it for good reasons.
"So the message isn't about them. It's the only exam we have ever been able to set ourselves. We are going to find out, in public, at four bits a second, what this ship says when nobody is checking.
"And I think we know the answer. I think we've known it since a maintenance engineer walked onto a stage with nineteen pages of pencil and put them on the floor unread and told eleven thousand people the first thing out of her mouth was that they shouldn't believe her.
"Send the lie first. Not the warning. The lie. Put Bay 2 and the twelve years at the top, in the header, in the part that repeats every cycle forever, before the chemistry, before the coordinates, before anything useful — so that the first four hundred bits anybody out there ever receives from us is the worst thing we ever did, and everything after it arrives from a source that has already told them it once decided that four hundred and six deaths were not load-bearing.
"That costs us nothing. That's what I'd like the far south to hear. It costs us four hundred bits and a fortnight of transmission a year. And it makes every number that follows it worth having, which the numbers otherwise are not, because a warning from a stranger with a clean record is worth about as much as a fuel margin in a year-nine textbook."
Eight seconds, and then, from the ground, eleven separate people trying to speak at once, which the lag rendered as a noise the transcript records only as (overlap).
Section 22
The ballot ran eleven days in the autumn of year one thousand and forty-four. Turnout was ninety-eight point two percent, joint, which broke the record set four years earlier by the ballot that had built the array.
There were two questions.
The first — shall the vessel transmit — carried with seventy-nine point one percent.
The second was the order, and it was four options and a write-in, and what carried, with sixty-one point four percent, was the header-first proposal, whole, exactly as the orchardist had said it and as the drafting committee had subsequently spent nine months failing to improve.
Bakri Ouma-Vaan was sixty-six and read the figure in the Bureau room at 0400 on the twelfth day and laughed until he had to sit down, and could not explain the joke to any of the eleven people present, all of whom had been born on the ground, and had to send them to the archive, where the explanation is filed under an index line, class open, petitionable, and has been consulted nine thousand times.
There was a third thing on the paper, added late by amendment on the floor of a settlement hall in the far south and carried on the indicative before anybody on the ship had had time to think about it, and it was the shortest item and the one that took the longest to draft.
A question.
A girl of sixteen from the ninth camp had asked for it and had been told she would have to ask twice and had asked four times. "Everyone keeps saying we're not asking them for anything," she said, "and everyone says it in the voice people use when they're being dignified. I'd like to know why. We have a chemistry in the ground that gives us an old age at seventy-one and four candidate approaches that are all wrong and eleven hundred people working on it, and there might be a laboratory fifteen light-years away that solved it four hundred years ago and doesn't know we exist. Thirty years for an answer. I'll be forty-six. My mother will be dead and I'll be forty-six and I would like to have asked."
They put four questions in, at the end of each cycle, after the warning and before the ledger, and the drafting committee wrote them in the plainest register two worlds could manage, and the fourth one is the one that ended up on the plate:
Do you know how to stop a fold?
Section 23
The transmission began at 0217 on a Fourthday in year one thousand and forty-five, because Astro had a sense of occasion that a thousand years of institutional humility had entirely failed to kill.
Nobody felt anything. There was no burn and no tilt and no fastener in the hull was asked any question it had not been asked for a hundred and forty years, and the Long Gallery had a hundred and thirty-one people in it, which was almost the entire standing complement, and eleven thousand more on the ground in forty-one assembly halls at four seconds' lag, watching a number.
The number was radiator margin. That was what there was to watch: a hundred kilowatts going into an array is a hundred kilowatts of heat coming out of a ship, and the only visible fact of the greatest act of speech in the history of either world was a thermal load curve on the fourth position of a wall, climbing through its transient and settling, and Bakri Ouma-Vaan sat at a bench on Deck 3 with his feet on a crate and watched it settle and said, to nobody, "That's it, then."
Heat is public. Heat is the one thing that leaves.
Eleven weeks later a girl of fifteen at the fourth camp on Ouma, doing the autumn reconciliation with a duplicate radiometer under a beans rota, folded four hundred days of her own instrument's noise floor and found the ship in it: a bump, four millikelvin high, at the Ambit's orbital period, creeping against the precession, narrowband, unmistakable. She published it on Fourthday in the ordinary weekly reduction with a four-line covering note and the observation that her method was Ouma-Vaan's, that she had got the sidelobe model wrong the first time, and that anybody could check it.
It is the most-taught page of that century. Not because it was hard. Because they had answered a beacon and then, without anybody planning it, had immediately gone and measured their own answer from outside, with the same instrument and the same arithmetic, and had published the residual.