The Museum of Things That Had Never Existed

I. The Donation

The letter arrived on a Tuesday, although the museum's correspondence register recorded it as having arrived on the previous Thursday.

The curator, Dr Elian Voss, noticed the discrepancy while checking the morning post. She assumed that someone had made a clerical error and placed the letter beneath a paperweight shaped like a small bronze horse. The horse had been excavated from a site that, according to the archaeological survey, had never been inhabited.

The letter was brief.

Dear Dr Voss,

I wish to donate a collection of objects to your museum. None of them has ever existed, but I believe they belong in your care.

The collection will arrive tomorrow.

Please ensure that the receiving room is empty.

Yours sincerely,

A Friend of the Museum

There was no signature, no return address, and no indication of how the sender knew that the museum had a receiving room.

Dr Voss read the letter twice. She had spent twenty years acquiring objects whose origins were disputed, whose authenticity was uncertain, or whose existence had been reported by only one unreliable witness. The proposed donation was unusual, but not entirely outside her experience.

She wrote to the donor, accepting the offer provisionally and requesting a catalogue.

The reply arrived before she had finished sealing the envelope.

The catalogue is enclosed.

There was no catalogue in the envelope.

There was, however, a small brass key.

Dr Voss examined it under a magnifying glass. It had no maker's mark. Its teeth were arranged in an unfamiliar pattern, and its bow bore an inscription in a language she could not identify.

She placed it in a specimen bag and entered it in the register.

Object 2026.104.1: Brass key. Provenance unknown. Intended use unknown. Authenticity undetermined.

She hesitated over the final field.

Existence: confirmed.

Then she locked the key in her desk.

The next morning, a wooden crate stood in the receiving room.

No delivery vehicle had entered the museum grounds. The security cameras showed the room empty at midnight and the crate standing in the centre of the floor at six in the morning. The night watchman denied having seen anyone enter.

The crate bore a single inscription:

THE MUSEUM OF THINGS THAT HAD NEVER EXISTED.

Dr Voss summoned two assistants, photographed the crate, recorded its dimensions and authorised its opening.

Inside were four objects, each wrapped in grey cloth.

The first was a keyhole cut into a small square of oak, with no door attached.

The second was a portrait of a woman whose face seemed familiar to everyone who looked at it, though nobody could say whom it resembled.

The third was a map showing a country surrounded by sea. The country had no name, but its borders were drawn with extraordinary precision.

The fourth was a clock without hands.

Dr Voss examined each object, then returned to the crate.

At the bottom lay a folded sheet of paper.

It contained a single sentence.

You will need these when the museum begins to remember.

She read it aloud.

One of her assistants laughed.

The other asked what it meant.

Dr Voss said that it was probably an attempt at humour, although she was not entirely certain whose.

She ordered the objects transferred to the conservation laboratory.

By the end of the day, the museum's database had assigned each of them an accession number.

The database rejected none.

II. The Keyhole

The key was the first object to change.

Three days after its arrival, Dr Voss found that the keyhole in the square of oak had become larger. Its dimensions no longer matched the measurements recorded on the day of accession.

She checked the specimen bag, the photographs and the conservation notes. All agreed that the opening had originally been narrow and vertical. Now it was wider, with a shallow curve at the top.

The change was small enough to suggest measurement error.

She placed the object beneath a scanner.

The scanner produced an image of a door.

Not a representation of a door, but a door: oak panels, iron hinges, a brass handle and, at its centre, the keyhole. Beyond the door was a corridor.

Dr Voss ran the scan again.

The result was identical.

She asked the museum's imaging specialist to examine the file. He found no evidence of manipulation, no software artefact and no reason to doubt the scanner's operation.

“Perhaps the object contains more information than we can see,” he suggested.

“Information about what?”

He shrugged. “A door, presumably.”

That evening, Dr Voss returned to the laboratory alone. She took the brass key from her desk and carried it to the oak square.

The key fitted.

She turned it.

Somewhere beyond the laboratory wall, a lock clicked.

Dr Voss withdrew the key and examined the room. Nothing had moved. The laboratory door was closed. The corridor outside was empty.

She turned the key again.

This time, she heard footsteps.

They approached from the other side of the wall, stopped directly in front of her, and waited.

Dr Voss put the key down.

The footsteps receded.

She spent the next two hours examining the laboratory's architectural plans. There was no room behind the wall. The museum's foundations had been surveyed repeatedly, and the building stood on solid ground.

She requested the original plans of the building.

The following morning, the archives delivered a set of drawings that showed a corridor running behind the laboratory.

The corridor had been removed from the plans during construction because the architect had decided that it was unnecessary.

A handwritten note in the margin read:

The door has been retained. It may prove useful.

The note was signed by an architect whose name did not appear in the museum's employment records.

Dr Voss checked the catalogue of the museum's architectural archive.

The architect's name appeared in an entry dated forty years earlier.

The entry described the architect's death.

The death certificate stated that he had never been born.

III. The Portrait

The portrait attracted visitors.

At first they were museum staff, then their families, then people who had heard rumours about an unfamiliar painting in the conservation wing. Dr Voss permitted a small number of visitors to see it, provided that they completed a questionnaire afterwards.

The results were extraordinary.

Nearly everyone recognised the woman.

Some said she resembled a childhood friend. Others remembered a schoolteacher, a neighbour, a relative who had died before they were born. One visitor insisted that she was the woman who had once saved him from drowning, although he had never been in danger of drowning.

No two visitors identified the same person.

Yet every visitor was certain that the resemblance was exact.

Dr Voss commissioned a specialist to examine the pigments. The painting contained no unusual materials. Its technique was consistent with a portrait produced in the late nineteenth century, although no known artist had used the distinctive combination of colours.

The specialist asked whether the museum had a photograph of the sitter.

Dr Voss showed him the painting.

“No,” he said. “A photograph of the person.”

“There is no known person.”

“Then we cannot establish the identity.”

“That is why I asked you to examine the painting.”

He looked at the portrait again.

“I think I know her,” he said.

Dr Voss waited.

“She was my mother's sister.”

“Did your mother have a sister?”

“No.”

He left without completing his report.

The next day, Dr Voss noticed that the portrait had acquired a small inscription on its lower edge.

It read:

For Elian, who will remember me.

She called the conservation team. No one had touched the painting since its last examination. The inscription was not painted over the existing varnish; it appeared to belong to the original work.

Dr Voss searched her family records. She found no reference to the woman, but she did find an old photograph of herself as a child, standing beside a garden wall.

In the background, half concealed by a tree, stood a woman whose face was indistinguishable from that in the portrait.

Dr Voss did not remember the photograph being taken.

She did not remember the woman.

She enlarged the image until the pixels began to separate. The woman was still there, smiling at someone outside the frame.

On the reverse of the photograph, in her own handwriting, were the words:

I promised I would never forget.

She had no recollection of writing them.

That night, she dreamed of a woman standing in a garden. The woman told her that she had been waiting a long time.

“For what?” Dr Voss asked.

“For you to become someone who could remember me.”

When she woke, she wrote down the dream and placed it in the museum's records.

She labelled it Unverified personal testimony.

Then she crossed out the word personal.

IV. The Country on the Map

The map was the most difficult object to catalogue.

Its geography was coherent. Rivers flowed towards the sea, mountain ranges followed plausible formations, and the settlements were positioned along trade routes that appeared to have developed over centuries. The map included roads, bridges, ports, agricultural regions and the ruins of cities that had been abandoned.

But no geographical database contained the country.

Satellite imagery showed only ocean where the map placed its coastline. Historical records contained no reference to the settlements. The language used for the place names was unfamiliar, although some words bore a resemblance to languages spoken on several continents.

Dr Voss asked a cartographer to examine the map.

He spent a week comparing it with every known projection and historical atlas he could access. At the end of the week, he returned with a curious observation.

“The map is not wrong,” he said. “It is impossible.”

“What is the difference?”

“A wrong map misrepresents a place. This map represents a place that cannot be located.”

“Could it be fictional?”

“Of course. But fictional maps usually reveal the purposes of their makers. This one is too exact in places that have no narrative significance. It contains drainage systems, minor roads, soil classifications and obsolete administrative boundaries. Someone has mapped this country as though they expected people to live there.”

Dr Voss asked him to identify the country.

He pointed to a small island near the northern coast.

“This is where I grew up.”

She looked at him.

“You grew up on an island that does not exist?”

“Not this island. But the harbour is identical. So is the road that climbs the hill. There was a church here, and a school here. The names are different, but the layout is the same.”

He turned the map towards her.

“Except my island was much smaller.”

The cartographer left the museum and returned to his home. That evening, he sent Dr Voss an email describing a childhood memory: a harbour, a church, a road climbing a hill, and an island that had vanished from every map he had consulted since leaving it.

He attached a photograph of himself as a child.

Behind him stood the same harbour shown on the map.

Dr Voss contacted the cartographer the following morning. His email account no longer existed. The museum's visitor records showed no evidence that he had ever entered the building.

She searched for his name.

No results.

She searched for the photograph.

No results.

She returned to the map and noticed that the island had grown.

The harbour was now larger, and a new road led inland. Several settlements had appeared along the coast. The country seemed to be developing.

She measured the changes and entered them in the object's record.

Observation: geographical configuration not stable across examinations.

Then she added:

Possible correlation between observation and observer.

The next morning, the map showed a new city.

It was named Voss.

V. The Clock Without Hands

The clock had not changed since its arrival.

It was a plain circular instrument, with a white face and a narrow black rim. There were no hands, numbers or markings. Its casing contained no mechanism that Dr Voss could identify, and its weight was inconsistent with the materials of which it appeared to be made.

She had placed it in a sealed display case, partly because she did not know how to preserve it and partly because she had begun to dislike the feeling that it was waiting for something.

Visitors generally ignored it.

Those who noticed it sometimes complained that the clock made them feel late.

One man became so distressed that he demanded to know what time it was. Dr Voss told him that the clock did not display time.

“Then why is it here?”

She had no answer.

A week later, the man returned.

“I have worked out what the clock measures,” he announced.

Dr Voss invited him to explain.

“It measures the time between things.”

“What things?”

“An event and its occurrence. A question and its answer. A promise and the moment it becomes impossible to keep.”

He seemed satisfied with this explanation.

“Can you demonstrate it?” she asked.

He shook his head.

“The interval disappears when you try to measure it.”

He left the museum and was never seen again.

Dr Voss began observing the clock at regular intervals. She recorded the time of each examination, the conditions in the room and the duration of her observations.

The clock's appearance remained unchanged, but the records showed a pattern.

Whenever she returned to the case, the interval since her previous visit seemed to have been different from the interval recorded by the museum's clocks. Sometimes an hour had passed in what felt like a minute. Sometimes a minute had occupied the better part of an afternoon.

She checked the clocks, replaced the batteries and commissioned an independent timekeeping instrument. The discrepancies persisted.

Finally, she asked a colleague to observe the clock while she remained elsewhere in the building.

He reported no unusual experience.

When Dr Voss returned, she found that he had left a note beside the case.

The clock has no hands because it does not tell the time. It waits for the person who needs to know when something happened.

She asked him what he meant.

“I don't remember writing that,” he said.

She examined the note. The handwriting was hers.

That evening, she stayed in the gallery until the museum closed. She sat opposite the clock and tried to imagine what it would mean to encounter an event without being able to say when it had occurred.

She thought of the woman in the portrait.

She thought of the country on the map.

She thought of the door behind the laboratory wall.

Then she remembered a conversation she had once had with her father, many years before his death. She could not recall his words, but she remembered the feeling of having asked him a question that he had been unable to answer.

She had spent years trying to reconstruct the conversation.

Now, sitting before the clock, she wondered whether she had been trying to recover not the words but the moment in which she had become someone who needed an answer.

The clock remained still.

For the first time, she was not certain that it had ever been an object.

VI. The Missing Room

Dr Voss returned to the laboratory with the brass key.

She turned it in the oak square.

The lock clicked.

This time, the wall opened.

Beyond it lay a corridor lined with glass cases. The cases contained objects she recognised from the museum's collections: a broken cup, a child's shoe, a stone carving, a manuscript with missing pages. Each object had been catalogued, conserved and displayed elsewhere in the building.

But here, each was different.

The cup was whole. The shoe had no wear. The carving showed a face that was absent from the original. The manuscript contained the missing pages.

At the end of the corridor was a room with a desk, a chair and a large mirror.

On the desk lay a catalogue.

Dr Voss opened it.

The entries described the objects in the room, but the dates were wrong. Some objects were recorded as having arrived before they had been made. Others were described as having been destroyed before they had been discovered. Several entries referred to people who had never existed.

She turned to the final page.

It contained an entry for the museum itself.

Museum of Things That Had Never Existed.

Date of foundation: to be determined.

Founder: Dr Elian Voss.

Purpose: to preserve objects whose existence depends upon their being remembered.

Current status: incomplete.

Beneath this entry was a second one.

Object 2026.104.1: Brass key.

Provenance: the collection.

Intended use: access to the missing room.

Authenticity: confirmed.

Dr Voss closed the catalogue.

She had never founded a museum. She had inherited neither the building nor its collections. She had been appointed curator after the previous director retired, and the records of that appointment were unambiguous.

Yet she knew the handwriting in the catalogue.

It was hers.

She looked into the mirror.

At first she saw only her reflection. Then she noticed that the room behind her was different. The glass cases were arranged in another order, and the desk stood beside a window overlooking a garden.

A woman was standing in the garden.

Dr Voss turned around.

There was no window.

She looked back into the mirror. The woman had come closer. Her face was now clearly visible.

It was the woman in the portrait.

The woman raised her hand and placed it against the glass.

Dr Voss did the same.

The mirror was cold.

She thought she heard a voice, but it might have been the sound of the museum's ventilation system.

Then the woman spoke.

“You took your time.”

Dr Voss tried to answer.

The woman smiled.

“You always do.”

The glass darkened, and the reflection disappeared.

Dr Voss stood alone in the room.

On the desk, the catalogue had opened to a new page.

Object 2026.104.5: Mirror.

Provenance: unknown.

Intended use: reflection.

Observation: the reflected space may not correspond to the space occupied by the observer.

Additional note: the observer has not yet arrived.

Dr Voss read the entry again.

Then she looked at the mirror.

Her reflection was no longer there.

VII. The Catalogue

The museum's director was concerned when Dr Voss failed to attend the morning meeting.

He sent an assistant to find her. The assistant searched the galleries, the conservation laboratory and the archives. He checked the staff rooms and the basement stores.

Dr Voss was nowhere to be found.

Her office was locked. Her desk was tidy. The specimen bag containing the brass key lay beside the keyboard.

The assistant noticed that the keyhole in the oak square had disappeared.

He informed the director, who ordered a full inventory of the museum's collections. The inventory revealed several discrepancies. Objects had been moved without documentation. Some records described items that could not be found, while other items had no records at all.

The director suspected theft.

He called the police, who inspected the building and found no evidence of forced entry. The security cameras showed no one entering or leaving the museum during the relevant period.

The investigation was closed.

The museum remained open.

Visitors continued to come, although some reported that the galleries seemed larger than they remembered. Others complained that familiar objects had been replaced by unfamiliar ones. The director dismissed these reports as errors of recollection.

He commissioned a new catalogue.

The work took three years.

When it was complete, the director announced that the museum's collections had been fully documented. Every object had a number, a description, a provenance and a location. Every discrepancy had been investigated. Every unexplained absence had been recorded.

The catalogue was a triumph of scholarship.

The director placed it in the entrance hall, where visitors could consult it before beginning their tour.

On the first morning, a young woman approached the desk and asked to see the Museum of Things That Had Never Existed.

The attendant searched the catalogue.

There was no entry for it.

“Are you sure you have the right museum?” he asked.

The woman showed him a letter.

It was brief.

Dear Dr Voss,

I wish to donate a collection of objects to your museum. None of them has ever existed, but I believe they belong in your care.

The collection will arrive tomorrow.

Please ensure that the receiving room is empty.

Yours sincerely,

A Friend of the Museum

The attendant looked up.

“Who is Dr Voss?”

The woman smiled as though she had been asked a familiar question.

“I believe,” she said, “that she is the curator.”

“But we have no curator by that name.”

“Not yet.”

She placed a small brass key on the desk.

The attendant examined it. Its teeth were arranged in an unfamiliar pattern, and its bow bore an inscription in a language he could not identify.

He opened the catalogue to the first page and began to write.

Object 2026.104.1: Brass key.

Provenance: unknown.

Intended use: unknown.

Authenticity: undetermined.

He hesitated over the final field.

Then, without knowing why, he wrote:

Existence: confirmed.

The woman thanked him and left.

That evening, the attendant went home and dreamed of a room in which every object had a different history. He saw a portrait of a woman he had never met, a map of a country that had never existed, and a clock that measured the interval between a question and its answer.

At the end of the dream, he found himself standing before a mirror.

His reflection was holding a catalogue.

The catalogue was open to a page bearing his name.

Beneath it was a single sentence:

The museum will remember you when you are ready.

He woke before he could read anything more.

The following morning, he arrived early and went to the receiving room.

It was empty.

He locked the door, checked the windows, and returned to the entrance hall.

At six o'clock, the security cameras recorded a wooden crate standing in the centre of the receiving room.

No one had entered the building.

The crate bore a single inscription:

THE MUSEUM OF THINGS THAT HAD NEVER EXISTED.

Inside were four objects, each wrapped in grey cloth.

The attendant opened the first.

It was a small square of oak, with a keyhole cut into its centre.

He turned it over in his hands.

On the reverse, in handwriting he recognised but could not place, was a short instruction:

Begin here.

He looked towards the door.

For a moment, he thought he heard footsteps approaching from the other side of the wall.

Then the sound stopped.

He put down the oak square, took a pen from his pocket, and opened the catalogue.

There was a great deal to record.

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.