The Kingdom That Outlawed Yesterday

There once lived a king who was troubled by the past.

It was not that he regretted anything he had done. On the contrary, he considered himself an exemplary monarch. His difficulty was that other people kept remembering things differently from the way they remembered them yesterday.

A minister who had praised a royal decree would later criticise it. A judge who had declared a law just would sometimes declare it unjust. Even the royal philosophers, who were employed to establish what could be known with certainty, occasionally announced that they had changed their minds.

The king found this intolerable.

“What is the purpose of wisdom,” he demanded, “if it cannot remain wise?”

His ministers consulted the royal philosophers, who met for forty days in the Hall of Immutable Truth. At the end of their deliberations, they presented the king with a proposal.

The kingdom, they advised, should abolish yesterday.

The king was delighted. He issued a decree that from that day forward, the past would have no legal existence. Once a day had ended, all claims concerning what had been thought, said or done on that day would become inadmissible. No citizen could be accused of inconsistency with an earlier statement. No judge could be challenged for reversing a decision. No philosopher could be asked to explain why a conclusion he now rejected had once seemed compelling.

The royal scribes were instructed to remove all dates from the records. The royal historians were dismissed. The royal schools ceased teaching history and began teaching the correct interpretation of the present.

For a time, the kingdom prospered.

The king's ministers agreed with him more frequently than ever before. The judges delivered their verdicts with unprecedented confidence. The philosophers published a new book every morning, each establishing truths that had never previously been in doubt.

The king himself became famous for his consistency.

When an ambassador reminded him that he had once promised to lower the taxes, the king replied that the promise belonged to yesterday and therefore had never existed. When a farmer complained that the king had ordered a bridge to be built across a river that had since changed course, the king explained that the bridge had been entirely appropriate at the time of its construction, which was a time that could no longer be discussed.

The farmer went away satisfied, although he still had to travel forty miles to cross the river.

The king's chief philosopher was especially pleased. He announced that the kingdom had at last escaped the tyranny of experience.

“Experience,” he explained, “is merely the present contaminated by what is no longer present.”

The king awarded him a gold medal.

Then, gradually, peculiar things began to happen.

The roads became difficult to maintain. A road that had once led to a thriving village now ended in a field, but nobody could remember why it had been built. A stone bridge crossed a stream that no longer flowed beneath it. A mountain pass had been closed by a landslide, yet the royal maps continued to show the road as open, since the landslide belonged to a day that had ceased to exist.

The king ordered the roads repaired.

His engineers asked for the original plans.

“Those are records of yesterday,” the king replied. “You know that yesterday has been abolished.”

The engineers attempted to reconstruct the plans from the roads themselves, but the roads had been built in response to circumstances they were forbidden to remember. They could see where the roads went, but not why they went there.

The repairs were abandoned.

Next, the kingdom's orchards began to fail. The royal gardeners had forgotten which varieties of apple flourished in which soils, when the trees had been pruned, and why certain orchards had been planted in sheltered valleys. They retained their knowledge of gardening, but each morning they were required to begin again from the garden as it stood that day.

When the harvest failed, the chief gardener explained that the trees had not obeyed the royal decree.

He was dismissed.

The king was increasingly annoyed. He summoned the chief philosopher.

“Why does the abolition of yesterday seem to have made everything more difficult?”

The philosopher thought for a long time.

“Perhaps, Your Majesty, we have not abolished yesterday thoroughly enough.”

The king ordered a second decree. Henceforth, no one could refer even indirectly to anything that had occurred before the present moment. The kingdom's clocks were redesigned so that they displayed only the hour, without indicating whether it was advancing or retreating. The royal astronomers were forbidden to compare observations. The physicians could no longer consult a patient's earlier symptoms, and the royal physicians began treating every illness as if it had appeared without warning.

The kingdom became quieter.

People stopped telling stories about their childhoods. Friends no longer recalled old quarrels or the occasions on which they had forgiven one another. Craftsmen ceased passing their techniques from one generation to the next, because instruction required the telling of what had been learned before. Even parents found it difficult to explain anything to their children.

The king, however, remained satisfied. His ministers continued to agree with him, and his philosophers continued to publish.

One afternoon, while inspecting the palace gardens, the king encountered a child sitting beside a dry fountain.

“What are you doing?” he asked.

“Waiting for the water.”

“There is no water.”

“There used to be.”

The king stopped.

The child looked up at him. “My grandmother told me that the fountain once flowed all year. She said the water came from a spring beneath the hill.”

“Your grandmother has been speaking of yesterday,” said the king. “That is forbidden.”

“But the spring is still there.”

“How do you know?”

“Because we dug down to find it.”

The king stared at the child.

“You dug?”

“My grandmother and I. We found the old channel. It was blocked by stones. We cleared some of them, but we couldn't finish.”

“Why did you not ask the royal engineers?”

“They said there was no reason to dig, because the fountain had no water.”

The king looked at the dry basin. For the first time, he wondered whether the kingdom had lost something more than its records.

He asked the child to show him the spring.

Together they walked behind the fountain and descended a narrow path into the hillside. There, beneath a tangle of roots, lay a stone channel. The child had cleared enough of it for a thin thread of water to emerge. It trickled between the stones and disappeared into the earth.

The king knelt beside it.

“Can you make the fountain flow again?” he asked.

“I don't know,” said the child. “But we could try.”

The king returned to the palace and summoned his engineers. He told them to restore the fountain, and when they asked for the original plans, he did not forbid them to consult the old records. He asked instead whether they could discover what the builders had understood about the hill, the spring and the channel.

The engineers searched the abandoned archives. They found drawings, measurements, accounts of floods, descriptions of the soil and letters exchanged between the original builders. Some records contradicted others. Some were incomplete. But together they revealed how the fountain had once worked.

The engineers restored the channel.

When the water rose through the basin, the king ordered a celebration. He invited the philosophers, the ministers and the judges, and he announced that the kingdom had achieved a new kind of consistency: one that permitted people to change their minds without requiring the world to forget what they had thought before.

The chief philosopher objected.

“Your Majesty, this appears to reinstate yesterday.”

“Not entirely,” said the king. “Only those parts of it that might help us understand today.”

“But who shall determine which parts are relevant?”

The king looked at the fountain.

“I do not know. Perhaps we shall have to learn.”

The philosopher considered resigning, but he could not remember whether he had already done so.

In the years that followed, the kingdom slowly recovered its roads, its gardens and its craft. The royal historians were recalled, though their salaries remained modest. The schools began teaching children how things had changed, and the philosophers were permitted to disagree with their earlier books, provided that they did not insist those books had never been written.

The king never entirely overcame his dislike of the past. Whenever someone reminded him of an unfortunate decision, he would still mutter that yesterday was an inconvenience.

But he no longer tried to abolish it.

He had learned, at least, that a kingdom could not move forward by making it impossible to discover where it had been.

And the fountain continued to flow, carrying water from the hidden spring beneath the hill, through stones laid by hands no living person had ever seen, into a basin where children gathered to play.

One day, a little girl asked her grandmother who had built the fountain.

The grandmother told her the story.

“And was the king right to abolish yesterday?” the girl asked.

The grandmother watched the water for a while.

“I think,” she said, “he was right to want tomorrow to be different.”

“But why did he abolish yesterday?”

“Perhaps he thought that was the only way to make it so.”

The girl considered this, then dipped her hand into the water.

“I'm glad he changed his mind.”

“So am I,” said the grandmother.

Neither of them asked whether the king would have agreed.

Mr Blottisham Discovers the End of Thought

The Senior Common Room at St Anselm’s was unusually quiet when Mr Blottisham entered carrying a large bundle of papers.

Professor Quillibrace looked up from his book.

“Good heavens, Blottisham. Have you brought your collected works?”

“Not quite, Professor. These are merely the works I have published since breakfast.”

Miss Elowen Stray put down her tea.

“Since breakfast?”

“Sixteen pieces, Miss Stray. Essays, reflections, two definitive accounts of the human condition, and a brief note establishing the limits of philosophical inquiry.”

“Have you finished philosophy, then?” asked Quillibrace.

Blottisham settled into his chair with the air of a man who had been waiting all his life for precisely this question.

“Not merely philosophy. I have finished thought.”

Quillibrace removed his spectacles and polished them.

“An ambitious undertaking.”

“It became unavoidable. I examined my complete body of work and discovered that every question worth asking had already been answered by me. Some questions had been answered several times, admittedly, but that only confirms the thoroughness of the enterprise.”

Miss Stray smiled. “And what will you do now?”

“Retire. There is no point continuing to think when thinking has been completed.”

Quillibrace replaced his spectacles.

“Have you considered the possibility that your archive is complete but your inquiry is not?”

Blottisham looked wounded.

“Professor, I have been publishing for years. My conclusions are extensive.”

“I did not ask about your conclusions.”

“Indeed, and that is precisely the difficulty with modern scholarship. One is continually asked about things one has not been asked about.”

Miss Stray reached for the top paper in his bundle.

“May I?”

“Certainly. It is the definitive account of why nothing remains to be said.”

She read the opening paragraph.

“Mr Blottisham, this appears to be an argument you published three years ago.”

“Naturally. It was definitive then, and remains definitive now.”

“But your conclusion today is different.”

“An elaboration.”

“Yesterday you argued that every genuine question has a single answer. Today you say that genuine questions cease to exist once their answers have been published.”

“Those propositions are perfectly compatible.”

“How?”

Blottisham leaned forward.

“Because the first establishes the answer, while the second establishes what happens afterwards.”

Quillibrace nodded gravely. “A formidable temporal distinction. Have you published it?”

“Four times.”

“Then it must be true.”

“Precisely.”

Miss Stray laid down the paper.

“I wonder whether the trouble is that you have been reading your own work in order to discover what you think.”

Blottisham stared at her.

“Miss Stray, I am the author.”

“Yes. But when you wrote those papers, you were answering questions you had then. Now you are reading them with questions you have since acquired.”

“Questions I have since acquired?” He looked almost alarmed. “Without my permission?”

Quillibrace intervened.

“Questions are notoriously bad at observing academic protocol.”

Blottisham gathered his papers protectively.

“I cannot be expected to revise the entire history of my thought every time I encounter a new question. Nothing would ever be finished.”

“Perhaps,” said Miss Stray, “but finishing something is not the same as exhausting it.”

There was a pause.

Outside, a gardener was attempting to persuade a reluctant wheelbarrow through a narrow gate.

Blottisham watched him for a moment.

“I have always maintained,” he said, “that a sufficiently comprehensive account of a thing renders further investigation unnecessary.”

“Does your account of the wheelbarrow explain why it will not pass through the gate?” Quillibrace asked.

“I have not written an account of the wheelbarrow.”

“Then perhaps the wheelbarrow has escaped your system.”

Blottisham frowned.

“This is exactly the sort of objection that prevents intellectual progress. One cannot include everything.”

“Agreed,” said Quillibrace. “But one should be cautious about declaring everything included.”

Blottisham rose and began pacing.

“I see what is happening. You are suggesting that my work may be incomplete.”

“Not at all,” said Quillibrace. “Your work may be complete in the sense that you have finished writing it. Whether it is complete in every other sense is a separate question.”

“And if it is not?”

“Then it may be worth reading.”

Blottisham stopped pacing.

“Reading it? I have already written it.”

Miss Stray leaned back in her chair.

“Have you ever returned to one of your earlier papers and found that it said something you had not noticed when you wrote it?”

“Certainly not.”

“Not even once?”

“I have occasionally discovered typographical errors.”

“That wasn't quite what I meant.”

“Then you should have asked a more precise question.”

Quillibrace coughed into his hand.

Miss Stray continued. “Suppose that a reader encounters an argument you published ten years ago. She has since learned things you did not know then. She notices a connection between your argument and something you never discussed. Has she found something in your work, or has she added something of her own?”

Blottisham considered this.

“Obviously, if it was not in my intention, it was not in my work.”

“And if it follows from what you wrote?”

“Then I must have intended it, unconsciously.”

“And if it contradicts something else you wrote?”

“Then the other paper must be interpreted in the light of the first.”

“And if they were both written first?”

Blottisham sat down.

“This is becoming unnecessarily complicated.”

Quillibrace returned to his book.

“Most interesting things do.”

For several minutes, nobody spoke. Blottisham looked at his papers, shuffling them into a new order. At last he seemed to reach a decision.

“I shall publish a further sixteen pieces this afternoon.”

Miss Stray raised an eyebrow.

“I thought you had finished.”

“I have. These will be clarifications.”

“Clarifications of what?”

“Of the fact that I have finished.”

Quillibrace closed his book.

“An important matter to establish.”

“And I shall include a warning that no subsequent interpretation of my work may exceed the scope of my original conclusions.”

Miss Stray looked thoughtful.

“Would that warning apply to interpretations you have not yet imagined?”

“Especially those.”

“And how would you recognise them?”

Blottisham opened his mouth. Then he closed it.

The wheelbarrow outside had finally passed through the gate. The gardener continued down the path, apparently unaware that he had resolved a problem that had occupied the Common Room for nearly half an hour.

Blottisham watched him go.

“I suppose,” he said slowly, “that one could write a short paper about that.”

“Only sixteen?” asked Quillibrace.

“Do not tempt me, Professor. I am trying to exercise restraint.”

He gathered his papers and headed for the door. At the threshold, he turned back.

“Of course, this discussion does not alter my conclusion that thought is finished.”

“Certainly not,” said Quillibrace.

Blottisham brightened.

“You agree?”

“I agree that you have finished thinking about it.”

Blottisham considered this distinction for a moment, decided that it was probably a form of wordplay, and departed to prepare his clarifications.

Miss Stray watched the door close.

“Do you think he will read his old papers again?” she asked.

Quillibrace reopened his book.

“Only if he publishes an instruction requiring it.”

“And will that help?”

“My dear Miss Stray, it may help the readers.”

Outside, the gardener had reached the end of the path. He turned the wheelbarrow around and began to push it back through the gate.

The Library That Remembered Everything Except Its Meaning

In a kingdom whose kings had long ago ceased to count their years, there stood a library so vast that no one knew where it ended.

Its oldest rooms had been built by the first kings, who believed that a kingdom could be made eternal if nothing that happened within it were ever forgotten. Every conversation in the market, every dispute in the courts, every song sung at a wedding, every question asked by a child, and every answer offered by a sage was faithfully recorded. Nothing was too trivial to preserve, and nothing too troublesome to be left out.

As the kingdom grew, so did the library. New wings were added, then towers, then underground chambers. When the scribes ran out of parchment, they invented new materials. When the shelves became too heavy, they devised lighter ones. When no one could remember where anything had been placed, they built catalogues. When the catalogues became too numerous to consult, they built catalogues of catalogues.

Eventually, the library employed more people to organise its records than the kingdom employed to govern itself.

Yet the librarians were proud. They could retrieve the words of any king, the argument of any philosopher, the recipe for any forgotten medicine, and the exact sentence spoken by a man who had died three thousand years before.

They called this perfection.

One autumn, a young archivist was assigned to the oldest chamber, where the earliest records had begun to crumble. Her task was to compare several copies of a celebrated dispute between two philosophers, one who maintained that the world was made of enduring things and another who claimed that nothing endured except the relations between things.

The dispute had been copied and recopied for centuries. Its words were familiar to every scholar, and generations of commentators had explained what each philosopher had meant.

But as the archivist read the different copies, she noticed something strange.

In the oldest account, the philosopher of enduring things seemed cautious, almost uncertain. In a later copy, his hesitations had disappeared. By the time the dispute reached the royal edition, he spoke with such confidence that his opponent appeared foolish before he had even begun.

The archivist checked the catalogues. Each copy had been correctly dated. Every alteration had been recorded. Nothing had been lost.

She went to the chief librarian.

“Which account is true?” she asked.

“All of them are authentic,” he replied. “Every word is accounted for.”

“That was not my question.”

The chief librarian adjusted his spectacles. “The earliest account is the earliest. The royal edition is the most authoritative. The commentaries tell you how the dispute has been understood. What more could you require?”

“I would like to know what the two philosophers were arguing about.”

The chief librarian looked at her with the patience reserved for the young.

“That,” he said, “is a matter for scholars. We preserve the records.”

The archivist returned to the chamber. She read the dispute again, this time setting aside the commentaries. Then she read it beside a collection of letters written by the philosophers in their youth. She found a letter in which one had confessed that he could no longer defend the position he had once held. She found another in which the other admitted that his opponent had changed his mind about something important.

The famous dispute had not ended as the official histories claimed. It had continued, in different forms, throughout their lives.

The archivist began to see that the dispute was not a contest in which one philosopher had defeated the other. Each had changed the questions the other could ask. Neither had arrived at the place from which the argument had begun.

She wrote a new account.

When the chief librarian read it, he was displeased.

“You have introduced nothing new,” he said. “Every document you cite has been here for centuries.”

“Then why,” she asked, “does this account say something I have never read before?”

He ordered her to check her sources again.

She did so, and found no errors. He ordered her to consult the authoritative commentaries. She consulted them, and found that they contradicted one another. He instructed her to compare her account with the royal edition. She compared them, and found that the royal edition had omitted the very letters that made the dispute intelligible.

At last, the chief librarian summoned the royal council.

The councillors examined the archivist's work. Some declared that it was not history, because it contained no new facts. Others said it was dangerous, because it made settled questions unsettled. One proposed that the new account be placed in a separate chamber for interpretations that had not yet received approval.

The oldest councillor asked the archivist a question.

“If everything you say is already in the library, what have you added?”

“Nothing,” she replied. “I have only read what was there.”

The councillor considered this.

“And if someone else reads the same records and finds something different?”

“Then we shall have to read them together.”

The council did not know what to make of this answer. They permitted the archivist to keep her account, provided that it was clearly labelled as a personal interpretation and placed where it could not be confused with the official record.

The archivist accepted their terms.

Years passed. She grew old, and other archivists took her place. Some came to her chamber and found her account illuminating. Others found it unconvincing. A few returned to the original documents and discovered questions she had not asked.

The chief librarian, meanwhile, continued to enlarge the catalogues. He had a catalogue made of every interpretation in the library, then another to distinguish interpretations that agreed from those that disagreed. He commissioned a final catalogue, which would show where every question had been answered.

When it was finished, he summoned the archivists to admire it.

The catalogue was magnificent. It listed every document, every copy, every commentary, every correction, every disagreement and every record of a disagreement. Its pages were bound in gold, and its index was so complete that it referred to itself in twelve different places.

The chief librarian stood before it, beaming.

“There,” he said. “Now nothing can escape us.”

A child who had accompanied her mother, one of the junior archivists, wandered up to the great book. She ran her finger down the index, then looked up.

“Where does it say what all this is for?”

The chief librarian laughed.

“Everything is in the catalogue, child.”

“I know,” she said. “But where does it say what it means?”

For the first time in many years, the chief librarian had no answer.

That evening, the archivists left the library in small groups. Some went home to read. Some went to argue with their friends. Others walked through the city, where people were still falling in love, quarrelling over boundaries, inventing songs, making promises and breaking them.

The chief librarian remained in his office, studying the catalogue. He checked its entries, verified its cross-references and corrected a misplaced comma. Then he turned to the first page and read the title.

The Library That Remembered Everything.

He wondered, for a moment, who had given it that name.

He searched the catalogue.

The name appeared in a dozen records, but none identified its author.

He searched again, going further back through the records until he reached the earliest surviving description of the library. There he found a sentence written in a hand so old that it was barely legible:

We build this place so that those who come after us may discover what we could not.

The chief librarian read the sentence twice.

Then he closed the catalogue, took his lamp, and went down into the oldest chamber.

For the first time in many years, he began to read.

The Senior Common Room On Impossible Houses and Other Unscientific Instruments

Mr Blottisham entered the Senior Common Room looking distinctly pleased with himself.

“I have solved the problem of the stories.”

Professor Quillibrace lowered his newspaper. Miss Elowen Stray looked up from her notebook.

“Which problem?” she asked.

“The problem of their intellectual status. The allegories, as we established, can direct attention to features of an argument. They are a kind of imaginative aid. Useful, perhaps, for readers who find abstract reasoning uncongenial.”

“How generous of you,” said Quillibrace.

“But these new stories are altogether different. An impossible house. A kingdom in which mirrors produce an unexplained figure. A machine that generates worlds but cannot find one that fits the evidence. They are not even allegories in the proper sense. They are exercises in fantasy.”

“Have you read them?” Elowen asked.

“Twice.”

“And have you solved them?”

“Of course not. That is my complaint.”

Quillibrace folded his newspaper.

“Then perhaps we should begin with the house.”

“I was hoping you would say that.”

“The Cartographer's House,” Elowen said, “presents us with a peculiar difficulty. Every room can be measured. Every passage can be traced. The stairs lead where they lead. Yet the measurements and connections refuse to assemble into a single, inhabitable plan.”

“An architectural impossibility,” said Blottisham. “The author has cheated.”

“Has he?” asked Quillibrace. “Suppose each measurement is accurate, each passage correctly described, and each observation faithfully recorded. Which observation must we reject?”

“The house cannot exist.”

“That is a statement about the whole, not an identification of the faulty measurement.”

“Precisely. The whole is impossible.”

“And yet the cartographer has observed it.”

Blottisham looked irritated.

“This is a fictional house. It can do whatever the author pleases.”

“An important discovery,” said Quillibrace. “We may now proceed without calling the building inspector.”

Elowen smiled.

“The story makes the relations among the observations the problem. The cartographer cannot simply discard the measurements, because they are individually sound. Nor can he pretend that their accuracy guarantees a coherent whole. He has to confront the difference between getting the parts right and understanding how they belong together.”

“Which is a familiar philosophical problem,” Blottisham said.

“Certainly. But here we cannot solve it by repeating the familiar formulation. We must follow a cartographer who is trying to do his job.”

“And failing.”

“Perhaps. Or perhaps he is discovering something about what his job requires.”

Blottisham turned to Quillibrace.

“Surely you are not suggesting that the house is a scientific model.”

“No. It is a fictional construction that lets us explore a question about the relation between local accuracy and global coherence. A scientific model must answer to the phenomena it models. This house answers to the demands of its own construction. But the difficulty it creates can still be intellectually productive.”

“Productive of what?”

“Questions, for a start.”

Blottisham sighed. “Questions are cheap.”

“Then let us see what they cost.”

Quillibrace took up the second story.

“In The Kingdom That Reflected Itself, the astronomers detect a figure in their mirrors. Their observations are reliable, and the figure helps them account for otherwise troublesome stellar movements. The kingdom gradually treats the figure as an established inhabitant. What interests you about that?”

“The mirrors,” Blottisham said. “Mirrors reflect what is before them. If the figure appears in every mirror, then something must be there.”

“Something is there,” said Elowen. “The reflections are observable phenomena. But what does that establish about the nature of whatever, if anything, stands behind them?”

“That the figure exists.”

“In what sense?”

Blottisham stared at her.

“In the ordinary sense.”

“Which would be?”

“That it is a real entity, rather than an optical effect.”

“Must those be the only possibilities?”

“Surely a reflection is either of something or it is not.”

Quillibrace intervened.

“The story does not establish that the figure is unreal. Nor does it establish that the astronomers have discovered a new entity with all the properties they subsequently attribute to it. It holds open the distinction between the evidence for a phenomenon and the account of what produces it.”

“But this is just the old argument about scientific realism again.”

“Not quite,” said Elowen. “The story also lets us watch the consequences of treating the proposed entity as settled. The figure acquires offices, a statue, a budget, and a place in the kingdom's official descriptions. The commission then tries to discover what stands behind the reflections and finds itself surrounded by more reflections.”

“A warning against bureaucracy,” Blottisham said.

“Among other things.”

“There you go again. Every story apparently means everything.”

“No,” Elowen replied. “But the relations it develops can bear on more than one question. The commission is funny because its procedure intensifies the very difficulty it is supposed to resolve. The mirrors do not become useless simply because they fail to settle the figure's ontology. And the figure's explanatory success does not magically turn the commission's empty pedestal into proof of its non-existence.”

Blottisham looked at the book.

“I dislike the mirrors.”

“An entirely legitimate aesthetic response,” said Quillibrace.

“I mean that they make the question unnecessarily elusive.”

“Would you prefer a story in which the commission opens a door, discovers the entity, and settles the matter?”

“Certainly. It would be much more satisfactory.”

“And if the entity then behaved in ways that contradicted the astronomers' account?”

“Then they would revise the account.”

“And if every revision produced new questions?”

“They would continue revising.”

“And if the problem concerned not merely the properties attributed to the entity, but the assumptions through which the astronomers had decided what kind of entity to look for?”

Blottisham paused.

“That would be a different problem.”

“Precisely,” said Elowen. “And it is the kind of problem the story permits us to approach without announcing in advance that it must arise.”

Blottisham was silent long enough for Quillibrace to turn to the third book.

“Now for the World Engine.”

“I have a great deal to say about that machine,” Blottisham announced. “Its engineering is implausible.”

“Undoubtedly.”

“It generates worlds according to laws engraved on brass rings, tests them against observations, and produces alternatives when they fail. It is a preposterous contraption.”

“Would a plausible contraption be more useful?” asked Elowen.

“Certainly.”

“For engineering?”

“For thinking.”

“Are you sure?”

Blottisham considered the question and decided not to answer it.

Quillibrace continued.

“The Engine is designed to generate every possible world, provided that world can be expressed in its own system. When an astronomical anomaly arrives that none of its worlds can accommodate satisfactorily, the machine reports: ‘NO SATISFACTORY WORLD FOUND.’”

“A software failure,” said Blottisham.

“Perhaps. But what is the failure?”

“The machine cannot do what it was designed to do.”

“Or it cannot do what its designers believe it was designed to do?”

Blottisham frowned.

“The distinction is artificial. If the Engine has been built to generate every possible world, then it should generate every possible world.”

“How does it determine what is possible?”

“According to its laws.”

“And how does it determine whether those laws exhaust the possibilities?”

“By generating the possibilities.”

“According to the same laws?”

Blottisham opened his mouth.

Quillibrace waited.

“An obvious circularity,” Blottisham said at last.

“Is it obvious to the Engine?”

“The Engine is a machine.”

“Which is why,” said Elowen, “it makes such a useful character.”

Blottisham looked alarmed.

“Character? It is a mechanism!”

“Of course. But the story gives that mechanism a task whose success depends on assumptions built into its construction. It can vary the laws, rearrange the symbols, and produce an extraordinary range of worlds. Yet its capacity to generate possibilities does not automatically give it a way to establish whether its own rules delimit possibility.”

“You are comparing it with scientific inquiry.”

“We are exploring a question that scientific inquiry can also encounter,” Elowen replied. “How do we recognise the limits of the framework through which we generate and evaluate explanations? The story does not prove that science is trapped inside an inescapable system, still less that a mathematical theorem about formal systems establishes such a conclusion. It gives us a fictional predicament in which the question becomes difficult to avoid.”

“Then the Engine is a metaphor.”

“It can function as one,” said Quillibrace. “But notice what the story does with it. The Engine does not merely represent a limitation. Its limitation drives the action. The astronomers must decide what to do when their most powerful instrument cannot produce a satisfactory world.”

“They modify the machine.”

“They consider modifying it.”

“A distinction without a difference.”

“Not if they do not yet know which modification, if any, will help.”

Blottisham returned to the book.

“I confess I find the Engine's response rather amusing.”

“‘PLEASE DEFINE “WORLD”’?” asked Elowen.

“Especially that.”

“Why?”

“Because it is an absurdly bureaucratic answer to a serious problem.”

Quillibrace smiled.

“Perhaps the machine has been trained by its Ministry.”

“I see no reason to introduce training into the story.”

“Nor is there any need to. The Ministry is quite capable of producing the same result unaided.”

Elowen laughed.

“But notice what the humour accomplishes. The Engine's response is absurd, yet it exposes a genuine difficulty. Before the machine can determine whether it has generated every possible world, someone must decide what counts as a world and what counts as a possibility. The question cannot simply be handed back to the machine without remainder.”

“Unless the machine is improved.”

“Perhaps. But what would count as an improvement? That is part of the question.”

Blottisham looked from one to the other.

“I begin to suspect that these stories are designed to prevent a satisfactory conclusion.”

“No,” said Elowen. “They are designed to let the difficulty develop before deciding what conclusion, if any, is warranted.”

“Which is a very roundabout way of thinking.”

“Sometimes,” said Quillibrace, “a roundabout is the only road that reaches the place you need to go.”

Blottisham looked unimpressed.

“Let me propose a more straightforward method. The author identifies a philosophical problem, constructs an impossible situation that illustrates it, and then explains the lesson. The reader understands the point. Everyone goes home.”

“And what would be lost?” Elowen asked.

“Nothing essential.”

“Then why did the stories need to be written?”

“To make the point entertaining.”

“Only entertaining?”

Blottisham hesitated.

“The house was rather funny.”

“And the mirrors?”

“Exasperating.”

“And the Engine?”

“Impertinent.”

Quillibrace nodded gravely.

“A promising range of intellectual responses.”

Blottisham ignored him.

“I see your difficulty. You want to claim that the stories discover something rather than merely express something. But surely the author knows what he is doing.”

“Sometimes,” Elowen said. “But constructing a fictional situation can disclose relations that were not fully apparent before the construction. The author may begin with a question and discover, through the demands of the story, another question that the original question did not contain.”

“Then the author has not planned properly.”

“Or the author has learned something while writing.”

Blottisham shook his head.

“Writing is not research.”

“Not necessarily scientific research,” Quillibrace replied. “But it can be a form of exploration. The writer establishes conditions, follows their consequences, encounters difficulties, and revises the construction. The difference is that the fictional world is not evidence about the physical world. Its value lies in the possibilities of thought it opens up.”

“And the reader?”

“The reader takes up the construction and does something with it. Not anything whatsoever, as you feared in our previous discussion. The story constrains interpretation through its details, but it need not dictate a single conclusion.”

Blottisham appeared troubled.

“Then the author cannot guarantee what the reader will discover.”

“No,” said Elowen. “Nor can the author guarantee that the reader will discover nothing.”

Quillibrace reached for his newspaper again.

“I believe we have arrived at the difficulty that distinguishes these fictions from the allegories. The allegories offer us situations in which the central relations of an argument become visible. The experimental fictions give us worlds whose peculiar conditions generate further predicaments. Their function is not simply to restate the original ideas, but to explore what becomes thinkable when those ideas are allowed to acquire fictional consequences.”

Blottisham stood.

“I remain unconvinced that impossible houses and malfunctioning machines have anything to contribute to serious intellectual work.”

“Of course,” said Quillibrace.

“And I maintain that a story should know what it means.”

“An understandable preference.”

“And that the author should remain in control.”

“An admirable ambition.”

Blottisham paused at the door.

“Though I must admit that the Engine's failure to find a satisfactory world was rather interesting.”

Elowen smiled.

“Why?”

“Because it had generated thousands of possibilities, all of them coherent according to its own laws, and yet none was satisfactory. I wanted to know what it would do next.”

“And what do you think it should do?”

Blottisham looked annoyed.

“That is the problem, isn't it?”

He left before either of them could reply.

Quillibrace unfolded his newspaper.

“Do you think he has understood?”

“Part of it,” said Elowen.

“Which part?”

“That a fictional world can become interesting in ways its creator did not entirely anticipate.”

“And the rest?”

She glanced at the three books on the table.

“I suspect he will need another story.”

Outside, somewhere beyond the garden, a staircase descended towards a door that had not existed the day before. In the Kingdom of Vey, a clerk was preparing a new budget for the Custodian. And in Orthe, the World Engine had begun printing a fresh report.

Its first line read:

POSSIBLE EXPLANATION: THE PREVIOUS REPORT WAS INCOMPLETE.

Its second line read:

PLEASE DEFINE ‘INCOMPLETE’.

III. The Observatory of Unfinished Worlds — A Gödel-inspired fiction: Reality reorganises the possibilities

A tale of a machine that could imagine anything, provided it was imaginable

At the highest point of the Kingdom of Orthe stood an observatory whose dome was made of black glass.

It had been built by seven kings, maintained by twelve generations of astronomers, and enlarged so many times that its oldest corridors no longer connected to its newest wings. Its inhabitants regarded this as a minor architectural inconvenience. After all, the observatory existed to study the universe, not to provide convenient access to the lavatory.

At its heart stood the World Engine.

The Engine was a brass cylinder as tall as a tower, surrounded by rings of engraved symbols. Inside it, a thousand glass wheels turned at different speeds. Silver threads passed between them, carrying patterns of light. When the Engine was activated, the observatory filled with the sound of distant bells, and a world appeared in the great chamber.

Not a picture of a world.

A world.

There were mountains, rivers, stars and seas. Clouds gathered over the mountains. Rivers ran towards the sea. Creatures moved through forests. In certain worlds, civilisations arose, built cities, fought wars and composed songs whose melodies had never been heard in Orthe.

The Engine generated worlds according to the laws engraved upon its rings.

Its first designers had been modest. They hoped to produce a world that behaved consistently enough to help astronomers test their theories. But the Engine proved far more capable than anyone expected. By adjusting its parameters, they could generate worlds with different distributions of matter, different stellar histories, different patterns of motion. They could vary a law and observe what followed. They could introduce a disturbance and watch how the world responded.

Most importantly, they could reject worlds that contradicted what had already been observed.

The Engine became the kingdom's greatest scientific instrument.

The Royal Academy established a School of Possible Worlds. The king appointed a Minister for Counterfactual Affairs. The observatory's chief astronomer, Lord Valdemar, received a title, a pension and a staircase named after him.

The staircase, unfortunately, led to a wall.

This was not considered relevant to his appointment.

For many years, the Engine served the kingdom admirably. When astronomers discovered a new pattern in the heavens, they generated worlds in which different explanations were true. They compared the resulting predictions with observations and discarded the worlds that failed.

The process was laborious, but the Engine never tired. It could produce a thousand worlds before breakfast and a million before the afternoon bell, provided the breakfast was not delayed by a royal inspection.

The astronomers came to trust it.

They knew, of course, that the Engine had limits. It could generate only worlds consistent with the laws engraved upon its rings. But those laws had been derived from the best available knowledge, and the Engine's successes seemed to confirm their adequacy.

Then, one winter, a signal arrived from the northern edge of the known universe.

It was not a sound, though the observatory's instruments translated it into a series of tones. It was not light, though it appeared in the astronomers' records as a pattern of brightness. It was a disturbance in the motion of distant stars, a discrepancy that persisted across several independent observations.

The astronomers fed the measurements into the World Engine.

The Engine produced a thousand worlds.

In every world, the disturbance could be accommodated. Some contained unseen concentrations of matter. Others had different distributions of visible material. In a few, the stellar motions had been affected by previously neglected interactions. Each world explained part of the evidence; several explained nearly all of it.

The astronomers congratulated themselves on the Engine's versatility.

Then further observations arrived.

The disturbance was not where the astronomers expected it to be. Its pattern differed from the predictions of the most successful worlds. The Engine was adjusted, recalibrated and supplied with the new data.

A million more worlds appeared.

Again, each world accommodated the disturbance, though some required increasingly elaborate arrangements. One contained an invisible ring of matter that curved around itself without closing. Another possessed a region in which the gravitational influence of an object depended on the route by which it was approached. A third required a new class of celestial body that could be detected only through its effects on other bodies that could themselves be detected only through their effects on the first.

The Minister for Counterfactual Affairs was delighted.

“We are making extraordinary progress,” he declared. “The Engine has shown us a million ways to explain the signal.”

The junior astronomer, Ilyra, raised her hand.

“Has it shown us a million ways the universe could be?”

“Of course,” said the Minister.

“Or a million ways the universe could be described using the laws already engraved upon the Engine?”

The Minister smiled as one smiles at a child who has asked whether the king wears his crown in bed.

“My dear young lady, the distinction is academic.”

Ilyra wrote something in her notebook.

The Minister noticed.

“Are you recording my observation?”

“No, Minister. I am recording the distinction.”

He moved on to the next exhibit.

Ilyra returned to the Engine after midnight.

She had been studying the engravings on its oldest ring. They were written in a notation so ancient that even the Academy's senior scholars could no longer agree on their original meaning. The official translation described them as the laws governing the relations between objects, forces and events.

But Ilyra had noticed something odd.

Every world generated by the Engine contained objects that could be distinguished, forces that could be measured, and events that could be ordered according to the same underlying scheme. Their properties varied enormously, but the relations by which those properties were expressed remained within the limits of the Engine's design.

The Engine could generate a world without stars, but not a world in which the distinction between an object and its relations had no meaning. It could generate a universe with unfamiliar laws, but only by expressing those laws in the language of the existing rings.

Ilyra took a sheet of paper and drew a small circle.

Inside it she wrote:

WORLD

Then she drew a larger circle around it and wrote:

POSSIBLE WORLD

She stared at the diagram.

It was not helpful.

She crossed it out and tried again.

By dawn, she had filled seventeen pages with circles, arrows, and increasingly irritable remarks about circles and arrows.

At last, she wrote a question:

Can the Engine generate a world in which the principles that define its possibilities no longer apply?

She carried the question to Lord Valdemar.

He read it twice.

“An interesting question,” he said.

“Can it?”

“The Engine generates every world consistent with its laws.”

“That was not my question.”

Valdemar regarded her over his spectacles.

“Miss Ilyra, do you wish to know whether the Engine can generate an inconsistent world?”

“No. I wish to know whether its laws determine what can count as a world in the first place.”

Valdemar took off his spectacles.

“My dear, if a thing cannot be described by the laws of the Engine, in what sense can we call it a possible world?”

“I do not know.”

“Then how can we investigate it?”

“I do not know that either.”

Valdemar put his spectacles back on.

“Excellent. You have discovered two reasons not to waste the Academy's time.”

He returned the paper.

Ilyra did not give up.

She began by changing one law at a time. She altered the relations between space and distance, between objects and their properties, between events and their order. The Engine generated unfamiliar worlds, but each remained expressible in the same underlying notation.

She tried removing a law altogether. The Engine refused.

She tried replacing a symbol with a blank space. The Engine produced a world in which nothing could be measured.

She tried instructing the Engine to generate a world in which its own rules were not the rules by which the world was organised.

The Engine emitted a puff of blue smoke and printed a small brass card.

PLEASE DEFINE “WORLD”.

Ilyra took the card to Valdemar.

He laughed so hard that he spilled tea on his waistcoat.

“The Engine has answered you,” he said.

“It has refused to answer me.”

“A distinction without a difference.”

“Is it?”

Valdemar dabbed at his waistcoat.

“Miss Ilyra, the Engine is not a god. It is a machine. It cannot be expected to do what its construction does not permit.”

“Then its successes cannot establish that it has explored every possibility.”

“Certainly not. No sensible person has ever claimed that.”

“But the Academy speaks of the Engine as though it generates every possible world.”

Valdemar looked towards the great chamber, where the silver rings turned in their patient sequence.

“It generates every world that satisfies its laws.”

“Yes.”

“And we have very good reasons to believe those laws describe the universe.”

“Yes.”

“Then what is the problem?”

Ilyra hesitated.

“I think I have been asking the wrong question.”

Valdemar looked relieved.

“Splendid. That is usually the first step towards asking a sensible one.”

Ilyra returned to the Engine.

For several weeks, she stopped trying to generate new worlds. Instead, she studied the histories of the worlds already produced. She examined those that had failed, those that had succeeded, and those that had required the greatest number of adjustments. She traced the relations among their assumptions, the observations they explained and the predictions they made.

A pattern began to emerge.

The most successful worlds did not simply contain the largest number of correct details. They organised the details in ways that allowed new observations to be explained without requiring a separate invention for each one. Their relations were productive. They made further discoveries possible.

The least successful worlds were often the most elaborate. They could accommodate almost anything, but they predicted very little that could not have been anticipated in advance.

Ilyra realised that the Engine's task was not simply to produce worlds. It was to help the astronomers discover which arrangements of relations could sustain an account of the evidence.

But the Engine could only compare arrangements it was able to express.

She took her findings to Valdemar.

“This is a useful result,” he said, examining her tables. “You have developed a better method for evaluating the worlds the Engine produces.”

“Yes.”

“And you have abandoned your question about possibilities beyond its laws.”

“No.”

Valdemar sighed.

“Miss Ilyra, there are limits to how much a person can trouble a machine before the machine begins to seem the more reasonable of the two.”

“I am not asking the Engine to be unreasonable. I am asking whether our definition of a possible world might itself be part of the problem.”

Valdemar folded his hands.

“Then you must tell me what you mean by possible.”

Ilyra looked at the rings of the Engine.

“I think that is precisely what I cannot yet tell you.”

For the first time, Valdemar did not laugh.

That evening, the northern signal arrived again.

It was stronger than before. The disturbance appeared in several new observations, and its pattern was unlike any the astronomers had recorded. The World Engine was activated at once.

It generated worlds containing new forms of matter, altered gravitational laws and unfamiliar relations between distant objects. It generated worlds in which the signal had been caused by a hidden body, worlds in which it arose from a previously unknown interaction, and worlds in which the original observations had been misleading.

Every world was coherent according to the Engine's laws.

None accounted for the full pattern of the new signal without additional assumptions.

The astronomers worked through the night.

At dawn, the Engine fell silent.

A narrow strip of paper emerged from its brass casing.

NO SATISFACTORY WORLD FOUND.

The Academy assembled in the great chamber. The Minister for Counterfactual Affairs demanded that the Engine be recalibrated. The engineers examined its gears. The mathematicians checked its symbols. Lord Valdemar ordered the oldest rings removed and inspected for damage.

Nothing was wrong.

The Engine was operating exactly as designed.

Ilyra stood at the back of the chamber, holding the original record of the signal.

“Could we try something?” she asked.

Valdemar turned.

“What do you propose?”

“I would like to describe the signal without first deciding which kind of thing produced it.”

The Minister frowned.

“How can one describe an observation without describing what caused it?”

“By beginning with the relations we have actually established, rather than assuming that the categories used by the Engine are the only ones in which those relations can be organised.”

“That sounds like a refusal to explain anything.”

“It may be. Or it may be a way of discovering what needs explaining.”

Valdemar studied the record.

“Very well. Show us.”

Ilyra went to the great table and spread out the observations. She marked the relations among the recorded motions, the intervals between the disturbances, and the differences between the predicted and observed patterns. She did not assign the disturbance to an object, a force or a new class of matter. She left those descriptions open.

For several hours, the astronomers examined the arrangement.

Then one of the older scholars noticed that the disturbance correlated with a feature of the observations that the Engine's models had treated as incidental. Another discovered that the pattern persisted across several independent methods of measurement. A third found that a relation previously regarded as a consequence of the disturbance might instead be part of the conditions under which the disturbance occurred.

The room grew quiet.

The astronomers had not discovered a new world. They had discovered that their description of the problem had been organising the evidence in a particular way.

The Minister looked alarmed.

“Are you suggesting that the signal changes its nature depending on how we describe it?”

“No,” said Ilyra. “The signal is not obliged to change its nature. But our descriptions determine which relations we notice and which possibilities we consider.”

“Then the signal could mean anything!”

“Not anything,” said Valdemar. “The evidence has already ruled out many things. The question is whether we have been allowing the Engine's vocabulary to determine the full range of explanations we are prepared to consider.”

The Minister looked from one to the other.

“I shall need a new committee.”

“You will need a new question,” said Ilyra.

The Minister began drafting the terms of reference immediately.

In the months that followed, the astronomers did not abandon the World Engine. They continued to use it, and it continued to prove extraordinarily useful. But they changed the way they approached it. They no longer treated the worlds it generated as the exhaustive inventory of what might be possible. They treated them as possibilities within a particular system of assumptions—powerful, productive assumptions, but assumptions nonetheless.

They also began to investigate whether the Engine itself could be modified. Not simply by adding new laws to its rings, but by altering the relations through which its symbols could be combined.

This work proved more difficult than anything they had attempted before.

Some proposed modifications produced nothing intelligible. Others generated worlds that could not be tested against observation. A few opened promising directions, but it was too early to know whether they would lead to a more adequate account of the universe or merely to a more elaborate machine.

The northern signal remained under investigation.

Years later, when Ilyra had become chief astronomer, a young apprentice asked her whether the World Engine would ever generate every possible world.

She looked at the great brass cylinder, its rings turning beneath the black-glass dome.

“No,” she said.

The apprentice looked disappointed.

“Then what is it for?”

“To help us discover which worlds are worth considering, which possibilities the evidence supports, and which assumptions we must examine when our best explanations fail.”

“But if we cannot generate every possibility, how will we know when we have found the right one?”

Ilyra smiled.

“We may not know that we have found the last possibility. We may know that an account has survived the tests we can devise, that it explains more than its rivals, and that it opens the way to discoveries we could not previously make.”

“And if another signal arrives?”

“Then we listen.”

The apprentice looked towards the northern sky.

“Do you think the universe contains worlds the Engine cannot imagine?”

Ilyra followed his gaze.

“I think the Engine has taught us something important about the difference between generating possibilities and knowing the limits of possibility.”

“That is not quite an answer.”

“No,” she said. “It is the answer that leaves us with work to do.”

Below them, the World Engine continued to turn. Within its chamber, new worlds formed and vanished, each governed by the relations its makers had taught it to express.

Beyond the observatory, the universe remained under no obligation to fit inside them.

And somewhere in the northern darkness, the signal continued.

II. The Kingdom That Reflected Itself — A mirror fiction: When a scientific concept becomes an entity

A tale of mirrors, invisible persons, and a royal commission that found exactly what it was looking for

In the Kingdom of Vey, every important building possessed a mirror.

The palace had seven hundred and twelve. The Hall of Justice had one behind every judge. The Royal Academy of Natural Philosophy had mirrors in its laboratories, lecture rooms and lavatories, where scholars were known to continue their arguments long after their colleagues had escaped.

The kingdom's most celebrated mirror-maker, Master Silas Venn, maintained that a mirror was the most honest instrument ever devised.

“It does not flatter,” he told the Royal Society. “It does not speculate. It shows what is before it.”

A philosopher in the second row raised his hand.

“Does it show what is behind it?”

“Certainly not.”

“Then it is not entirely honest.”

Master Venn considered this objection.

“It is honest within its proper limits.”

The philosopher sat down, having obtained more than he had expected.

For centuries, the mirrors of Vey had served the kingdom well. They helped physicians inspect wounds, judges identify witnesses, and astronomers compare the positions of stars with the marks on their instruments. A person could stand before a mirror and learn something about the appearance of a person standing before it.

No one doubted this.

Then, one winter, the astronomers discovered a reflection that had no owner.

It first appeared in the great mirror of the Northern Observatory. The astronomers had been observing a cluster of stars whose motions seemed inconsistent with the amount of visible matter in the region. They had checked their calculations, repeated their observations and compared their records. The discrepancy persisted.

One night, while examining the instruments, the junior astronomer Mara noticed a figure standing behind her in the mirror.

She turned.

No one was there.

She looked again. The figure remained.

It was tall, dressed in a dark robe, and stood with its hands folded before it. Its face was indistinct, as though the mirror had not yet decided what sort of face to give it.

Mara called the senior astronomer.

He entered the room, looked into the mirror, and saw nothing unusual.

Then the figure moved.

It raised one hand, and the reflected stars shifted around it.

The astronomers worked until dawn. They discovered that the figure appeared whenever certain stellar motions departed from their predictions. Its position varied, but not randomly. Its movements seemed related to the very discrepancy they had been investigating.

By morning, they had given it a name.

The Custodian.

The name was provisional, as all scientific names were supposed to be. The astronomers made this clear in their report, which was entitled Preliminary Observations Concerning a Possible Reflective Agency Associated with Certain Celestial Irregularities.

The Royal Society shortened the title to The Custodian.

Within a month, the discovery had transformed the study of astronomy.

Whenever the figure appeared, its position provided clues about the behaviour of the stars. The astronomers learned to predict where it would stand. They discovered that its apparent influence extended across regions in which no corresponding figure could be seen. Its absence from a mirror did not mean its absence from the calculations.

The Custodian became indispensable.

The king summoned Mara to the palace.

“Can this figure be found?” he asked.

“We do not know, Your Majesty.”

“Can it be measured?”

“Its effects can be measured.”

“Can it be summoned?”

“Not deliberately.”

The king leaned back.

“And yet it has already proved useful.”

“Very useful.”

“Then I shall order a statue erected in its honour.”

Mara hesitated.

“Your Majesty, we do not yet know what it is.”

“Neither do we know what wisdom is,” replied the king. “Nevertheless, we have a statue of it in the Hall of Justice.”

The statue was commissioned.

Master Venn was instructed to design a mirror large enough to display the Custodian clearly. He constructed a chamber of silvered glass, with seven mirrors arranged around a central space. When the astronomers entered, they saw themselves repeated in every direction.

At the centre stood the Custodian.

The figure was clearer now. Its robe was richly folded, its hands finely rendered, and its face almost visible. The astronomers watched it for hours, recording every movement.

They discovered that if they adjusted the mirrors by a small amount, the figure's apparent position changed. If they altered the angle of the light, its outline became sharper. If they placed a candle behind the central mirror, the figure appeared to stand farther away.

The experiments produced a great deal of information.

They also produced a great deal of disagreement.

Some astronomers argued that the Custodian was an independent entity whose presence was revealed by the mirrors. Others maintained that it was a property of the reflective arrangement itself. A third group proposed that the figure was neither simply external nor merely reflective, but belonged to a relation between the instruments, the observations and the phenomenon under investigation.

The king found the third position unnecessarily complicated.

“Surely,” he said, “the figure is either there or it is not.”

The royal philosopher replied, “Your Majesty, that depends on what you mean by there.”

The king dismissed him from the room.

Meanwhile, the Custodian's reputation grew.

The Royal Academy published a treatise describing its distribution throughout the kingdom. The Royal Survey Office marked its territories on maps. Apprentices learned the equations governing its effects. The king's ministers discussed its influence on trade, navigation and the security of the realm.

Soon the Custodian was no longer spoken of as a possible explanation for the stars' behaviour.

It was spoken of as something that gathered, moved, influenced and persisted.

A grammar grew around it.

The astronomers no longer said, “Our calculations require a contribution from the Custodian.” They said, “The Custodian gathers beyond the northern stars.” They no longer said, “The observations are consistent with an unseen influence.” They said, “The Custodian's influence extends beyond the visible heavens.”

The change was gradual, almost imperceptible. No decree announced it. No astronomer could identify the day on which the figure had become a fact of ordinary speech.

But the change had consequences.

When the Royal Academy established a Department of Custodian Studies, it needed a director. When the king commissioned a census of the Custodian's territories, he needed surveyors. When a young astronomer proposed a new experiment, she was asked how it might illuminate the Custodian's nature.

The figure had acquired offices, maps, budgets and a professional constituency.

It was, by every administrative measure, flourishing.

One morning, Mara received a letter from the Royal Commission on Unseen Entities.

The Commission requires you to identify the original Custodian, distinguish it from its reflections, and establish its location within the physical kingdom.

Mara read the letter twice.

Then she wrote back:

I can identify the circumstances under which the Custodian appears. I can describe its observed effects and the predictions that follow from them. I cannot yet identify an original figure independent of the reflective arrangements.

The Commission replied that her answer did not satisfy the terms of reference.

She was summoned to testify.

The chairman, Lord Pellifer, was a man who believed that every difficulty could be resolved by placing it under the appropriate heading.

“Dr Mara,” he began, “do you believe the Custodian exists?”

“I believe that we have strong evidence for the phenomenon we call the Custodian.”

“That was not my question.”

“It is the question I can answer responsibly.”

Lord Pellifer consulted his papers.

“Are you denying the existence of the Custodian?”

“No.”

“Are you affirming it?”

“I am distinguishing what we have established from what remains uncertain.”

The chairman looked pained.

“Dr Mara, this Commission cannot proceed on distinctions.”

“Then I fear it cannot proceed very far.”

A murmur passed through the room.

Lord Pellifer ordered the mirrors brought in.

The Commission had commissioned a new apparatus, consisting of one hundred mirrors arranged in a circle. At its centre stood an empty pedestal. The theory was that the original Custodian would appear in the central space, while the surrounding mirrors would display its reflections.

The experiment began.

One by one, the mirrors were adjusted. The astronomers altered the angles, changed the lighting and recalibrated the instruments. The figure appeared in the seventh mirror, then the nineteenth, then the thirty-fourth. Each time, the surrounding reflections changed.

But the central pedestal remained empty.

“Perhaps the figure is too distant,” suggested one commissioner.

“Perhaps it requires a different arrangement,” said another.

“Perhaps we need more mirrors,” said Lord Pellifer.

Master Venn, who had been invited as an expert, cleared his throat.

“My lord, more mirrors will provide more reflections.”

“And what will provide the original?”

Master Venn glanced at the apparatus.

“That, my lord, is the question.”

The Commission ordered two hundred more mirrors.

The experiment continued for another six months. The apparatus grew until it occupied an entire wing of the palace. The Custodian appeared in more places than ever. Its movements could be predicted with increasing accuracy. Its relation to the stars was mapped in greater detail.

Yet no original figure was found.

The Commission published its report, which concluded that the nature of the Custodian remained uncertain and recommended the establishment of a permanent institute to investigate it.

The institute was opened with great ceremony.

Above the entrance, a sculptor placed a statue of the Custodian, its face hidden beneath a hood. Beneath the statue, an inscription declared:

To the Unseen, whose presence is known by its effects.

Mara attended the ceremony. She stood at the back of the crowd, watching the mirrors in the entrance hall multiply the statue into an endless procession of dark figures.

A young apprentice approached her.

“Do you think we shall ever find the original?” he asked.

“I do not know.”

“But the evidence is so strong.”

“Yes.”

“And the predictions are so successful.”

“Yes.”

“Then surely the original must be somewhere.”

Mara looked into the nearest mirror.

The Custodian stood behind the apprentice. Its robe was dark, its hands folded, its face indistinct.

She moved to the left. The figure moved with her.

She moved to the right. It followed.

The apprentice turned around. There was nothing behind him.

“Tell me,” Mara said, “what would you expect to see if the Custodian were an independent entity? And what would you expect to see if the figure arose from the relations among the mirrors, the observations and the things we are trying to understand?”

The apprentice thought for a long time.

“I am not sure.”

“Good. Then you have found the beginning of the investigation.”

“But if we cannot distinguish the two possibilities, what are we to do?”

“We continue to test what can be tested. We seek evidence that can discriminate between the accounts. We do not pretend that uncertainty makes the evidence worthless. Nor do we pretend that successful predictions answer every question we might ask.”

The apprentice looked again at the mirror.

“Then the Custodian may be real, but we may not yet know what kind of reality it has?”

“Precisely.”

“Or we may discover that our present account needs to change.”

“Precisely.”

He frowned.

“You seem remarkably comfortable with not knowing.”

Mara laughed.

“I have spent thirty years among astronomers. One learns to make oneself comfortable wherever there is a chair.”

They left the chamber together.

Behind them, the mirrors continued to reflect the Custodian. Its image stood in the central mirror, in the outer mirrors, and in the mirrors reflected by the mirrors. It appeared at different distances, in different postures, under different lights. Each reflection was real as a reflection. Each revealed something about the arrangement that produced it.

But none, by itself, settled what stood beyond the glass.

Years later, the Royal Academy replaced the original apparatus with a new instrument. The old mirrors were removed, polished and distributed throughout the kingdom. Some were installed in laboratories, others in observatories, and several were purchased by wealthy citizens who admired the Custodian's likeness.

The figure continued to appear in certain arrangements. In others, it did not. New methods revealed effects that the old instruments had missed. Some predictions were confirmed; others failed. The question of the Custodian's nature remained open, though the evidence for the underlying astronomical phenomena grew stronger.

In the great hall of the institute, a young scholar discovered an old mirror hidden behind a cabinet. It was smaller than the others, its silvering worn, its frame made of dark wood.

When she looked into it, she saw herself.

Behind her stood a figure in a dark robe.

She turned.

Nothing.

She looked again.

The figure remained.

The scholar felt the old excitement rise in her chest. She reached for her notebook, then stopped.

Instead, she moved the mirror towards the window. The figure shifted. She turned the mirror slightly, and its face disappeared. She placed a candle behind the glass, and the figure seemed to recede into the darkness.

She repeated the experiment.

The results were consistent.

That evening, she entered the observatory and wrote in her notebook:

An unexpected figure has appeared in a reflective instrument. Its appearance varies systematically with the arrangement of the instrument and the conditions of observation. Further investigation is required.

She paused, then added:

It would be premature to name it.

The next morning, she found that someone had placed a small brass plaque beneath the mirror.

It read:

THE CUSTODIAN

The scholar sighed.

Somewhere in the palace, a clerk was already preparing the paperwork for a new department.