2084: An Algorithmic State — Act XVI: Distribution, Chapter 86: Aisha, Inside

Aisha files three formal minutes over eighteen months about one clerk who copies more than four hundred referrals a year by hand. One sentence of the Pranaali changes. A journalist she has never met gets the credit, and she cannot prove he is wrong.

In Brief: Aisha spent four years arguing in the courtyard. Now she works inside a Ministry records office, where she finds that nobody decides anything, so she works through a clerk’s transcription burden instead. Later, a journalist’s piece in a publication the Ministry reads explains the change a different way.

The thing nobody had told Aisha, in four years of arguing about this from a courtyard, was that the inside of a records office is not a place where anybody decides anything.

She had gone in expecting rooms with decisions in them. Over those four years she had built a fairly detailed private architecture of where the harm was made. In it there was an office and a person. There was a moment when somebody with a name looked at a paragraph and chose the version that would cost a woman in Nizamuddin her medical tier. By roughly her fifth month she had understood that no such office existed. The paragraph had been drafted in 2061 by a committee of eleven. She eventually located its surviving members and interviewed them. Two of them could not remember the paragraph at all. One remembered it as a compromise struck to protect a different clause, which had itself been repealed in 2068. The harm was not decided. It was inherited, the way a family inherits a bad staircase. Everyone who lives in the house learns to take the third step differently, and nobody thinks of it as a decision anyone made.

By the end of her first year she had worked out that this meant there was no argument to win. She had been very good at arguments. For four years she had been the person in a courtyard who could take the honest position and hold it without a clean name for it. She brought that whole apparatus in with her, and found that there was nothing in the building for it to grip.

What there was, instead, was work.

Every operative criterion in the Pranaali generated administrative burden somewhere. Every unit of administrative burden landed on a desk belonging to a person who had never read the criterion and did not care about it. That person wanted, above all things, less work. The variance criterion in Section 5 was the one the Cultural Contribution board scored by. It governed how far a determination could sit from its modelled band before a human being was required to look at it. In its 2061 wording it generated a class of referral that arrived at the drafting section’s own clerks in a format three of their systems could not index. So four hundred and something of them a year had to be transcribed by hand, by a Grade II named Bhavna. She was a broad quiet woman who wore her reading glasses on a cord. She had been doing it for six years, and she mentioned it, unprompted, to anyone who sat near her at lunch.

Aisha did not argue that the criterion was unjust. By then she had established that this would refer her to the policy directorate, which was a room where minutes went to be acknowledged.

She filed a minute about Bhavna’s transcription.

It took eighteen months. It took a second minute, about the indexing, and a third, about the cost of the transcription in officer-hours. She was able to state that cost to two decimal places because she had spent four evenings with Bhavna’s own logbook, and Bhavna had been pleased that somebody wanted it. It took a redraft of the referral format. She did not write it and was not consulted on it, and when it arrived it was attributed in the file to a working group she had never sat on. And it took, in the end, a single amended sentence in the variance criterion. It was adopted because the amended version generated a referral in a format the systems could index. So it generated no transcription, and so it made a Grade II’s afternoon shorter. It also had a consequence that nobody in the drafting section discussed at any point in eighteen months. It widened, by a small margin, the class of determinations a human being was required to look at before they took effect.

The referrals stopped arriving at Bhavna’s desk on a Monday. She had four left over from the week before and did those. On the Tuesday there were none, and on the Wednesday there were none. On the Thursday she asked the section officer whether the format had changed, and he said he believed a working group had done something, and went back to his screen. Nobody told her what. She kept the logbook open on the corner of her desk for another fortnight, with the pen beside it, and then put both in the drawer. At lunch she said to Aisha that six years was a long time to do a thing and then have it stop without a note. Aisha said nothing about the minutes. She asked what Bhavna would do with the afternoons.

She did the arithmetic afterward, because she had never in her life been able to resist doing the arithmetic. Somewhere between nine hundred and thirteen hundred determinations a year would now be read by a person who would not previously have read them. Of those, on the historical rate, perhaps forty would be changed. Of the forty, some unknowable number would be the difference between a woman keeping her tier and a woman losing it. There was no instrument anywhere in the building that could tell her which ones, or how many, or whether any at all.

She was thirty-six. She had given the Ministry four years to move one sentence, and the sentence had moved for a reason that had nothing to do with anything she believed.

The piece appeared in the kind of publication the Ministry read and did not admit to reading. It was written by a man called Ashwin, a journalist who stood next to documents for a living, whom she had never met and never would.

It was very good. It was better than good. It was the best piece of administrative journalism she had read in her four years, because he had done the thing almost nobody did. He had found the two actual people the old wording had happened to, and put them on the page with their own names and their own sentences. She could tell from the piece itself that this had cost him considerable time, and had cost them something too. He had gone to the sector twice. He had sat with them. He had made it impossible to read the paragraph without seeing them.

The piece said that the variance criterion had changed because he found them.

Aisha read it twice at her desk and then once more that evening, and could not fault a single fact in it. She understood that he had reasoned honestly to the causal claim in its final section, from everything available to him, and had not invented it. What was available to him included the date of the amendment and the date of his own earlier reporting. It did not include an eighteen-month sequence of minutes about a transcription burden, which had never been published anywhere and never would be.

She thought about writing to him.

She composed the letter twice, in her head, walking to the metro, and did not write it either time. There were two reasons, and she was careful to keep them separate in her own accounting. The first was that she could not spend the capital. She was one officer, four years in, with a small store of the particular credit that comes from being useful and quiet. She had a next paragraph in view, Section 8, the tier-reallocation clause, a much larger animal. Correcting a journalist’s attribution would buy her nothing toward it. It would cost her the reputation of not being the sort of person who wants credit. That reputation, she had learned, was the single most valuable asset available to anyone in that building.

The second reason was harder and she came back to it for years.

She did not actually know that he was wrong.

His first piece on the criterion, the one with the two people in it, had appeared four months before the amendment. It had been read, she later established, by two people in the drafting section, one of whom had forwarded it to a third. Her minutes had been in the system for seventeen months by then. She could construct, without straining, a version in which the transcription argument moved a format and the piece moved a will, and neither alone would have been enough. And no method existed for saying which had done what, or how much, or whether either had done anything at all. Perhaps the paragraph had simply arrived at its own moment, the way paragraphs do. No such method existed in her hands, or in his. It did not exist anywhere in the four hundred and twelve pages a panel had already produced on a much larger version of the same question.

She filed the piece in her own folder, the private one. It was not a ledger. For four years she had refused to keep one, because a ledger is a thing you keep when you have started needing to be able to prove something.

She kept the folder anyway.

Section 8, she wrote that night, on the first page of a notebook she bought on the way home and did not tell anyone about, generates no burden on anyone with the power to move it. Find one.

About this serial: 2084: An Algorithmic State is a dystopian novel by Nithin Palal, published online one chapter at a time.

Previous: Chapter 85: The Man Who Came Looking

Next: Chapter 87: Kavya, Unfunded