THE MAFIA BOSS WATCHED HIS BRUISED MAID DRAG HER BROKEN LEG ACROSS THE MARBLE FLOOR, BUT WHEN HIS ELEVEN-YEAR-OLD DAUGHTER HANDED HIM A SECRET PAPER, HIS FIANCÉE TURNED DEATHLY PALE.

The anonymous photograph that had started our investigation carried a date on its back: April 2009. A second copy arrived at the archive in the spring, submitted by a former volunteer after the hearing. This version showed the same children outside the green doorway, but the printed date said September 2008.

I noticed the discrepancy while checking the public exhibit inventory. The children stood in the same places. The striped awning cast the same shadow. Even the loose corner of a poster matched. They were copies of the same exposure, but at least one date had been added later.

June stared at them side by side. “Which one is real?”

“The photograph is real. One or both dates may be inaccurate.”

“Does it matter?”

“It might. The timing affects whether the picture documents access before or after a particular renovation.”

We consulted the original donor, who turned out to have retained a sleeve of negatives. The negatives had no written date, but a contact sheet showed the doorway with a temporary banner advertising a meteor shower. Public program calendars placed that event in 2008. It was stronger evidence than either handwritten date.

The discovery didn't overturn the case. It did something more humbling. It reminded us that even our most persuasive image could carry a mistaken label.

I proposed correcting the exhibition catalog and recording the basis for the revised date. One volunteer objected.

“People will think the whole archive is unreliable,” he said.

“They're more likely to trust us if we show how we corrected a mistake,” I answered.

He wasn't convinced. Neither was I entirely comfortable. But comfort was not the purpose of the correction.

At home, Sophie was studying the planets when I explained why the photograph needed a new date. She laughed.

“That's like when I wrote Neptune on the wrong model and Amelia said my whole solar system was fake.”

“Did you fix it?”

“Yes. Then we argued about glitter.”

“Did the glitter have a correct answer?”

“She thinks more is always better. That's obviously false.”

For a few minutes we talked about her project, and the Meridian photograph belonged only to me. Then she asked a question I hadn't expected.

“Can I stop helping with the exhibit forever if I want?”

“Yes.”

“Even though Mom started the program?”

“Even then.”

She nodded. “I think I want to come to star nights, but not the history meetings.”

“That sounds like a good choice.”

Later Luca thanked me for not trying to change her mind. He said he had been tempted to tell Sophie that helping preserve Sarah's work was a family responsibility.

“She already carries enough,” I said. “Her mother doesn't need to become homework.”

He leaned his head against the chair. “I wish somebody had told my father that family loyalty wasn't homework either.”

The conversation left us both thoughtful. He was still untangling inherited contracts. I was still learning to draw boundaries between my work and our private life. Neither process ended when the lights went out.

The conflicting photograph dates also forced us to reconsider a caption we'd prepared for the historical panel. The first version had described the six children as attending the reopening of the east entrance. The contact sheet indicated the picture belonged to a meteor-shower event, and we couldn't establish whether any reopening had taken place that day.

Carmen read the corrected caption aloud: CHILDREN AT THE MERIDIAN EAST ENTRANCE DURING A NIGHT WINDOW EVENT, 2008. DATE ESTABLISHED FROM PROGRAM CALENDAR; NAMES WITHHELD.

“It's less dramatic,” she admitted.

“It's more useful,” I replied.

We invited two trained volunteers who had not worked on the original wording to review the evidence. One asked whether the sky-event banner might have been reused in another year. We checked photographs of the program decorations and found that the banner had indeed appeared twice. The calendar alone was not enough. A poster visible at the edge of the contact sheet included a sponsor logo used only during the earlier season, which strengthened the date but still didn't prove the exact day.

Naomi suggested recording a range rather than a single date. It was humbling to discover that our correction itself had initially been too confident.

June spent an afternoon updating the metadata and documenting the successive interpretations. She left the original labels preserved in the internal research history, clearly marked as superseded, so a later researcher could see how our understanding had changed. No public visitor would need to read the entire technical history, but the institution would no longer conceal that its description had evolved.

That evening a volunteer asked whether we were wasting time on one old photograph while the city faced urgent housing decisions. I understood the concern. I explained that accurate archives would not build homes, and that the hearing team—not our cataloging group—was responsible for moving the planning questions forward. Different people could do different forms of necessary work at the same time.

After everyone left, I stood in the library courtyard and watched June lock the display cabinet. She was tired and proud of the work, though she would never call a corrected date heroic. I thought about the child who had once written CAN THE MOON HEAR US? The answer mattered less to me than whether the child would feel safe asking another question after getting something wrong.

When Sophie later heard about the correction, she asked whether we should put an eraser next to the star wall. I laughed and said we already had one. She suggested adding a sign that read: YOU CAN CHANGE YOUR QUESTION WHEN YOU LEARN MORE. Anita approved it the next day.

A week later, Iris Vale contacted the independent privacy reviewer directly. She did not ask for an interview or publicity. She wanted confirmation that the old participant packet concerning her had been identified, and she wanted advice about how she could exercise whatever rights were available to her.

The reviewer handled the matter confidentially. I learned only that a request had been received, because Anita needed to report program-level totals without identifying applicants. That was as much as I was entitled to know.

Teresa called me the next evening. “Iris says the process was respectful,” she said.

“I'm glad.”

“She's still angry.”

“She has every right to be.”

“Good. I didn't want you to think one correct form would settle the past.”

I smiled despite myself. “I promise I know better than that.”

At the commission's continuation hearing, a representative from Larkspur presented the missing compliance condition as a historical notation that had been superseded by newer design standards. Maya disagreed that the record supported such a definitive conclusion.

“The condition may have continuing effect,” she said. “Even if it doesn't, the commission must explain how it considered it.”

A commissioner asked why the project team had not flagged the issue earlier. Rourke said they had relied on available city checklists. Maya calmly answered that the earlier record had been available in the municipal archive, though not properly cross-referenced.

I recognized the kind of failure. No villain needed to erase a page if two filing systems could lose sight of each other for a decade.

After the hearing, Nolan Pierce published a follow-up article acknowledging the corrected photograph date. The headline still implied that the investigation was unraveling. I sent a concise correction to his editor, explaining what the dating revision meant and what it did not mean. I did not write an angry public thread. Sometimes the durable record mattered more than the immediate argument.

The university archives director then offered me a temporary role helping design cross-reference standards for community agreements. It would extend beyond my placement and pay enough that I could reduce my weekend shifts.

I was thrilled. I was also worried that accepting the role would deepen the conflict between professional obligations and Luca's family history. Professor Ellis suggested I disclose the relationship and request a supervisor for any Meridian-related material.

“You don't have to reject an opportunity because you have connections,” she said. “You have to manage the connections transparently.”

I accepted the job with that condition in writing.

After I received the university offer, Luca came by with takeout dinner. We sat on the floor because my table was covered with folders. I told him about the disclosure requirements and my fear that people would think I had earned the position through him. He asked what evidence supported that fear. I laughed at hearing my own favorite question returned to me. There was none beyond the old anxiety of being judged through my proximity to a wealthy man. I had completed the training, submitted the work, and received an independent evaluation. He didn't tell me to stop worrying; he helped me name what was actually true.

The same afternoon, Marisol notified Luca that an old corporate correspondence file had been found. It contained a memo from his father's office asking Northbank to revise its drawings to show the eastern corridor as a service route.