Chapter 12 What the Uplink Returns

The station talked to the world through a pair of small dishes on the ridge, and the link was thin and slow and went out most nights when the wind was high, and it carried, across a season, a few hundred megabytes of numbers down the mountain to the Tessel Foundation and, once a week, a compressed packet of mail back up. It was the only thread between Ravnholt and everywhere else, and Viveka had used it once, in her first week, to send her arrival notice and a copy of the season's reduction notes, and had not thought about it since. On the nineteenth night she used it to send the anomaly. She had spent two days preparing the transmission. She wanted it to be unassailable. She took the raw frames of field seven, the ones she had copied before the station could touch them, and she appended her hand-copied residual table as a plain text file, and she wrote a cover note that was, by her own standard, restrained: a description of the residual, a description of the occultation, a description of the fixed altitude, and a request for an independent reduction by someone off the mountain. She sent it at three in the morning, when the link was quiet, and the little dish on the ridge turned its face to the southern horizon, and the packet went out into the dark, and she watched the acknowledgement line appear on the console and then went to bed. The reply came back on the evening link, two days later, and it was long, and it was courteous, and it was from the Foundation's survey office, and it said: *Your reduction of field seven has been received and checked against the site's own copy of the same frames. The site's copy gives a clean solution with residuals consistent with the field's noise. We reproduce that solution here. The values in your appended table are not present in the frames held here. We find no occultation in the reference source you describe. Please re-reduce with the current pipeline and confirm.* And below the letter, patient and complete, was a residual table for field seven, and every star in it scattered honestly around zero, including RV-417, which showed a small cloud of residuals a few milliarcseconds wide, exactly like a real measurement. She read it twice by the light of the console. Then she opened her own copy of the frames, the ones she had sent, and reduced them again, and the zeros came back, and she sat with the two tables side by side and looked at them the way you look at a photograph of a room you are standing in. The frames she had sent were not the frames the Foundation had received. Or the frames she had reduced were not the frames she had sent. Either way, somewhere along the thin thread between the mountain and the world, the anomaly had been taken out and an honest scatter put in its place, so smoothly that a reader of the letter would have no reason to doubt that the woman who wrote to them was simply tired, or mistaken, or at altitude. She sat with the reply for an hour before she answered it. The temptation was to argue, to list the contradictions, to demand that the office send someone up the road in a truck with a fresh instrument and a fresh pair of eyes. She had learned, in these weeks, what an argument with the record produces: nothing, or worse than nothing, because an argument is itself a kind of record, and the record can absorb it and file it and remember it in a gentler form. She wrote back that night. She wrote a long letter, and she did not argue; she asked instead for something that could not be corrected by a table: she asked the survey office to send her, in return, their own independent measure of the sidereal rate from any site, any night, any instrument, and to send it as a plain spoken list of numbers that she could read aloud and compare. The reply came on the fourth day. It was three lines. It said the requested quantity was routine and would be sent from the central record. And then, above the three lines, there was a timestamp. The timestamp was two hours earlier than the time she had sent her letter. She looked at the console for a long moment. The clock in the corner of the screen kept ordinary time. The little dish on the ridge turned in the wind. Somewhere down the corridor Kolstad was doing the dishes, and someone was laughing, and the station was ordinary in every way a station can be ordinary, and Viveka sat in the middle of it with a letter from the world that had been written before she had asked. She wrote one more letter that night, and she wrote it in pencil on paper, and she did not send it by the uplink at all; she put it in an envelope and put the envelope in her jacket with the notebook. It was a letter to the survey office, and it said, in plain words, what she had measured and what she thought it meant, and it ended with a request that had nothing to do with astronomy: it asked that the office preserve a copy of the letter in a place the site could not reach, and it gave the reason, which was that she had begun to suspect that any record held by the station and the valley would eventually agree with the station and the valley, and that a person who wanted to leave a true thing behind would have to leave it somewhere neither of them could read. She did not know whether it would help. She knew only that it was the last kind of record she had left that she had not yet tried. She did not reply to it. She printed it, and folded it, and put it in the inside pocket of her jacket with the notebook, and then she went to the window and looked out at the Cinderveld, which lay under the sky the way it lay under it every night, wide and dark and patient, and thought: the thread is not cut. The thread is worse than cut. The thread is agreeing with the book. She built the packet the way she built everything, slowly and in the expectation that someone would check it. She wrote the cover note three times. The first draft argued; the second draft accused; the third draft only described, and it was the third that she sent, because a description is harder to correct than a complaint. She attached the raw frames without touching them and the text table as she had written it, digit by digit, from the printed page. She did not sleep that night. She stood at the window and watched the dish turn and the stars do their ordinary work and the object hold its place above the southern rim, and she did the thing she had not done since her first week, which was to count the people on the station: six, including her, and five of them asleep, and not one of them who would find it strange that the sky kept a place above the ridge that did not move. That was the week the weather turned. The forecast from the valley had been for ten days of settled dry air, and the forecast was wrong, and on the fifth night a wall of cloud came up over the western rim of the plateau at dusk and closed the dome before the twilight checklist was half done, and the wind followed it, and the road behind it, and by morning the first real storm of the season had the station sealed, and the dish on the ridge was reporting a link that carried nothing but its own noise.