Read the opening.

Arriving at 7:45 AM, a registrar named Ito Keiko badged through the staff entrance on the north side and crossed the ground-floor atrium on her way to the stairs. Family name first, in the Japanese convention she had answered to for all of her forty-three years. She was senior enough that colleagues called her Ito-san, and sufficiently precise that they called her nothing else. After eleven years at NACT, she no longer registered the morning light refracting through the concave glass along Roppongi. The building had long since ceased to be an architectural landmark to her; it was a tool. She knew its blind spots, its corridors, and its habits by heart.

The main atrium was not yet open to the public. Down on the level below, the cleaning team was already at work, their noise carrying up through the polished concrete along with the distant cycling of the environmental systems in the ceiling. Underneath both, the sound the atrium made before the crowds arrived: a faint, live resonance that the building’s ten thousand daily visitors absorbed without knowing it existed. Her footsteps rang at this hour in a way they would not by nine. She took the stairs to the fourth floor. The retrospective galleries were dark behind their temporary partition walls, awaiting the objects that would populate them for four weeks and then leave, as everything eventually left, in crates.

NACT held no permanent collection. This was, architecturally and institutionally, the point. Unlike the Tokyo National Museum in Ueno, or any of the other national museums in the city, NACT was not a repository. It was a venue: roughly 150,000 square feet of flexible exhibition space, organized around a soaring central atrium, available to rotate through loan exhibitions and public art showcases on an unceasing schedule. Everything came in and went out. The building’s job was to provide the environment, and Keiko’s job was to document what passed through it precisely enough that objects left in the condition they arrived.

Nineteen years as a registrar, eleven of them at NACT, had produced an immaculate record. Across ceramics, lacquerwork, textiles, photographic prints, and large-scale installations, not a single object had ever been lost. That included a seventeenth-century folding screen so delicate that the humidity protocols for its two-week exhibition had required three separate briefings with the loan institution. Keiko considered this her professional baseline rather than an achievement, in the same way that a surgeon considers not killing anyone a minimum, not a distinction.

She dispatched the pre-inspection email to the lender’s documentation contact just before nine. Twenty minutes later, she gathered the condition report forms from the registrar’s office, clipped them to her board, and, at the inspection room door, verified the environmental log. The climate control had held steady since the night before: sixty-five degrees, forty-five percent humidity. She noted the readings, stepped inside, and closed the door behind her.

The moon jar occupied the center of the white inspection table, betraying no sign that it was something worth dying for. She photographed it.

The jar was the size of a large pumpkin and the color of a winter sea: a pale, muted celadon, gray-green where the glaze lay thick and translucent where it thinned over the clay at the edges. Joseon dynasty, circa 1685, according to the accession paperwork: the Korean ruling dynasty whose ceramics are among the most prized in Asia. Its silhouette was not quite symmetrical: one shoulder slightly higher than the other, the way moon jars always were, made in two halves and joined at the equator, the seam invisible under the glaze but present in the body’s gentle lopsidedness. Under the macro lens, the glaze resolved into a fine network of crazing, the web of surface cracks that develops in aged glaze, a faint map when the light was right. Its surface was otherwise unmarked except for a hairline crack along the upper shoulder, barely visible at this angle, which she had already logged in two photographs, one from each side. It was three hundred and forty-one years old, the centerpiece loan from the Korean Cultural Heritage Foundation for a retrospective of Joseon ceramics that would open in eleven days. She had six more photographs to take, four measurements, and fifteen minutes of dictation.

She set the camera to its macro setting and got on with it.

The inspection room at the National Art Center Tokyo was not designed with mood in mind. It was designed for light — four overhead panels, daylight-balanced, no shadows — and for surfaces that could be wiped down. The table was white. The walls were white. The pale gray linoleum floor showed up every scuff and grain of dirt. Keiko preferred rooms that hid nothing. It was why she found NACT a reasonable arrangement. Everything moved through. The paperwork was always fresh.

Eyes on the jar, she kept her voice unhurried as she spoke into the small recorder clipped to her lanyard, cataloging the object piece by piece: glaze, body, foot rim, and the visible interior. For the inside surfaces, she positioned the camera at the vessel’s mouth and tilted the lens downward at a sharp angle, careful never to brush the ceramic rim. Nearly two decades of practice had given her an eye that was accurate to a sixteenth of an inch, producing images that could be perfectly aligned across years of inspection cycles. Good practice, she called it. Her colleagues called it slightly obsessive. They were simply underestimating the value of consistency.

The door opened behind her at 10:23 AM.

“Good morning. I am Dr. Yun Se-bin, accompanying courier from the Korean Cultural Heritage Foundation. I was told to present myself here.”

Keiko straightened and turned. Yun Se-bin was thirty-something, slender, in a dark blazer over a white shirt, with reading glasses pushed up into her hair and a soft-sided document case under her arm. She looked around the room the way Keiko had seen researchers look at new archives: measuring the space, absorbing its logic, not yet at ease with it. As she looked at the jar, her expression shifted into something intense and proprietary. Keiko recognized that look immediately. She had seen it before in scholars who had spent years studying a reproduction, only to find themselves standing in front of the original for the first time.

“Senior Registrar Ito.” Keiko set the camera down and gestured toward the table. “Please come in.”

“Beyond the courier role, I have some independent research interest in the object.” Se-bin moved to the other side of the table and looked at the jar without touching it. She had good instincts, or good training: she kept her hands at her sides. “Your condition report protocol — do you permit annotations from the courier representative?”

“Standard practice. Lending institution observations are co-signed at the bottom.” Keiko retrieved the report from the clipboard on the side table. It was a twelve-page document. She was two pages in. “I would ask you to wait until I have completed my photographic record before writing. A pen near a glazed surface in this light makes me nervous.”

Se-bin almost smiled. “Of course.”

They worked in companionable quiet for the next twenty minutes. Keiko photographed and dictated. Se-bin stood at a respectful distance and watched, occasionally making small notes in a hardbound research journal she had produced from her document case and laid open on the side table. She did not crowd the jar. She did not offer unsolicited commentary. Keiko revised her read of the woman upward.

“I’m going to turn it now,” Keiko said. “If you want to observe the base markings, this is the moment.”

Se-bin came to stand beside her. Keiko lifted the jar in both hands, her nitrile gloves gripping the smooth surface where cotton would have slipped, and felt the cold density of its two days at sixty-five degrees radiate through them. It was not the superficial chill of metal or glass, but the deep, even coldness of fired earth, which held temperature the way it held its glaze: patiently, and for a long time. It was heavier than its size suggested; they always were. As she rotated it slowly, a shallow, incised mark appeared on the underside: two characters, cut before firing, partially obscured by a run of glaze. Keiko photographed it from three angles, reading out the dimensions into her recorder. Se-bin leaned in close to study the mark, her face unreadable. She said nothing.

Keiko set the jar down and took the clipboard from the side table. She initialed the first three pages, turned to the annotation section at the bottom of page four, and passed the clipboard and her pen to Se-bin.

“Condition observations from the lending institution,” she said. “Date and sign at the bottom of the annotation. I’ll countersign.”

Se-bin took the clipboard with both hands, set it on the side table beside the open journal, and sat down to write. She wrote in Korean, which was not unusual. Keiko’s condition reports were accepted in Korean, Japanese, or English: provenance work crossed borders by nature, and a researcher’s sharpest observations surfaced in her native language first.

Keiko was wholly absorbed in her final exterior measurements, her attention entirely focused on the task. Still, at the edge of her attention, she noticed that Se-bin was writing extensively — a distinct anomaly, as courier annotations rarely ran past three lines. A researcher with an institutional system of her own would naturally have more to record. She mentally filed the observation away, the way she handled any data that was anomalous but not yet legible.

When Se-bin returned the clipboard, Keiko glanced at the annotation: a neat, regular Korean hand, the characters evenly spaced. She initialed the page without reading it. Her Korean was limited to signage and numbers. The annotation would be filed with the report; if it required translation for institutional purposes, the curatorial team would arrange it.

Se-bin had dated and signed at the bottom, as asked. Keiko countersigned below. The inspection record was complete to that point.

“Thank you,” Keiko said. She meant it as a formality, but something in Se-bin’s manner made her add: “You’re here for the full duration of the loan? The retrospective runs for four weeks.”

“Until the retrospective closes.” She paused for a moment, as if deciding something. “There is something in the acquisition file I would like to look at more carefully. A detail in the provenance record that does not align with what I expected to find. It is probably an administrative inconsistency, but I prefer to confirm it myself.” Se-bin set the pen down on the side table. “I have been to your archive before. Dr. Fujiwara arranged access for my earlier visit.”

“She’ll have set you up with our archivists, then. The acquisition records run back to the building’s founding.” Keiko looked at her recorder and checked the time marker. “I’ll need another hour in here. If you need anything before you leave, the main reception can direct you to the curatorial offices.”

“I know the building a little,” Se-bin said.

Keiko nodded. Se-bin picked up her document case and left, closing the door behind her with the care of someone conscious of sound.

Two minutes later, the door opened again. Keiko did not look up from the photograph on the camera’s rear screen.