What the Drawer Knows and the Grandchildren Do Not

We met in a café near the Retiro that Marías had chosen and Cusk had, upon arrival, silently judged. It was the kind of Madrid establishment that had been operating since before the transition — dark wood, marble-topped tables, a television behind the bar tuned permanently to a football channel at zero volume. The waiter knew Marías, or at least knew to bring him a coffee without asking. Cusk ordered tea and received it in a glass cup with no saucer. She did not comment on this. I ordered water and immediately wished I had ordered something that gave me something to do with my hands.

“The question I keep circling,” I said, “is which sibling is the protagonist. There are three of them. They find these documents — execution orders their grandfather signed. And I think the story wants to be about the one who wants to burn the papers.”

“You said ‘the story wants,’” Marías said. “Already you are crediting the story with desires it does not possess. You have desires. The story is inert. It is a series of decisions you have not yet made.”

He was right, of course. But I had hoped to get further into the premise before being corrected.

“What interests me,” Cusk said, “is that you’ve identified the burner as the central figure. Why not the one who wants to publish? Or the one who can’t decide?”

“Because the burner is the one whose position is hardest to hold.”

“Hardest for whom? For you, or for a reader?”

“Both.”

“Those are different problems.” She turned her glass cup a quarter turn on the table. “You’re worried the reader will think the burner is a villain. But the more dangerous possibility is that the reader will think the burner is the only sane one in the room, and then the story has no tension at all. Just a sensible person surrounded by sentimentalists.”

Marías laughed. It was not a laugh of amusement, exactly. “She is right that this is the danger. But she is wrong about why. The danger is not that the burner appears sane. The danger is that you — ” he looked at me — “will be so afraid of making the burner a villain that you will bend the entire narrative to justify the burning. And justification is the death of fiction. We do not write to justify. We write to delay the moment of judgment until the reader cannot remember which side they were on when they began.”

“You want ambiguity,” I said.

“I do not want ambiguity. Ambiguity is a word people use when they mean they have failed to choose. I want the judgment to arrive — I want it to arrive fully formed, with confidence — and then I want it to dissolve. Not because the narrator has undermined it, but because the next document, the next text message, the next fragment of the grandmother’s memoir, rearranges everything the reader thought they understood. The judgment keeps arriving. It keeps dissolving. The reader is never allowed to rest.”

Cusk had been listening with the particular stillness I was beginning to recognize as her form of disagreement. “That’s a very elaborate way of saying you want the reader to be confused.”

“Confused is not the same as—”

“I know it’s not the same. I’m being unfair to you deliberately, to see if you’ll defend the position properly.”

A brief silence. The television behind the bar was showing a replay of a goal. No one in the café was watching it.

“What I mean,” Marías said, and I could hear him choosing his words with a precision that bordered on hostility, “is that the documents themselves must do the work. The execution orders. The grandmother’s three memoirs — each one abandoned. The text messages between the siblings. These are not evidence in the legal sense. They are evidence in the Sebaldian sense. They accumulate. They create a pressure that is not the pressure of plot but the pressure of — I want to say accrual, the way silt accrues in a river delta, you do not see it happening until the river has moved.”

“And time,” I said. “The same document means something different in 1975 than in 2003 than in 2024. The grandfather signs an order. At the time it is bureaucracy. Forty years later it is history. Sixty years later the grandchildren find it in a drawer and it becomes — what? A grenade? A confession?”

“Not a confession,” Cusk said. “A confession requires a confessor. The grandfather is dead. The documents are not confessing anything. They are simply there. That is what makes them unbearable — they do not explain themselves. A confession at least has the decency to be addressed to someone. These papers are addressed to no one. They performed their function decades ago and now they are orphaned. Purposeless. The purpose has migrated entirely to the grandchildren, who did not ask for it.”

This was, I realized, the closest the three of us had come to agreeing on anything, and none of us had noticed. We were talking about the same object — the document as an inert explosive — from three directions, and for a moment those directions converged.

“But I am concerned,” I said, “about the grandmother. Her three abandoned memoirs. Why three? Why does she keep starting and stopping? I’ve been imagining that each draft gets closer to the truth and then flinches.”

“No,” Marías said. “That is too clean. Each draft closer to the truth — that is a trajectory, and a trajectory is a form of reassurance. I would prefer that the three memoirs approach the truth from entirely different angles, as though the grandmother kept losing her nerve not because the truth was too painful but because she could not locate it. She knew what happened — she lived in the house with the man — but knowing and narrating are such different operations that she could begin the narrative three times and produce three incompatible versions of what she knew.”

“So the memoirs contradict each other.”

“They do not contradict each other because that would imply they are trying to say the same thing. They are trying to say three different things, and the grandmother abandoned each one when she realized it was becoming the wrong thing.”

Cusk shifted in her chair. “I want to push against this idea of the grandmother as the keeper of the real story. It’s a common move — the old woman who knows everything, the repository of family truth. I find it sentimental.”

“You find it sentimental because English literature has made it sentimental,” Marías said. “In Spain the grandmother who knows is not a comforting figure. She is terrifying. She held the family together during forty years of silence. Not because she was brave, but because the alternative to silence was the destruction of everything she had built on top of the silence. Her house, her children’s careers, the marriages, the grandchildren — all of it was built on the foundation of what she did not say. When she tries to write the memoir, she is not unburdening herself. She is performing a controlled demolition. And she keeps aborting the demolition because she can see what will fall.”

“Then the memoirs are not about truth at all,” I said. “They’re about damage assessment.”

“Closer. They are about the impossibility of separating the truth from the structure that grew around its concealment. You cannot extract the secret without destroying the family, and you cannot preserve the family without perpetuating the lie, and the grandmother understood this with a clarity that the grandchildren, when they find the papers, do not possess.”

Cusk was quiet for a long time. The waiter came and cleared her empty glass. She asked for another tea and he brought it in the same glass, which I noticed she examined for traces of the first cup.

“I’ll tell you what I think the story needs,” she said, “and it is not the grandmother. It is the text messages. The siblings texting each other in real time as they go through the drawer. The present-tense, reflexive, unedited responses. ‘Have you seen this one?’ ‘Don’t show Mamá.’ ‘I think we should call a lawyer.’ These are people processing something they have no framework for, and the text message is the most honest document in the story because it is the least considered. No one redrafts a text message.”

“People do redraft text messages,” I said.

“People delete text messages,” she corrected. “Which is a different thing. The deleted text — the one typed and then erased before sending — that might be the most important document in the entire story. The thing one sibling almost said to another and thought better of.”

Marías was nodding, which he did slowly and rarely. “This is where the structure must bear the weight. You have three kinds of documents: the official papers — the execution orders, the court records — which are the language of the state performing its function. You have the grandmother’s memoirs, which are the language of a private person trying and failing to construct a narrative. And you have the text messages, which are the language of people who have not yet had time to construct anything. Three registers. Three relationships to time. The official documents are frozen — they mean exactly what they meant in 1943 and also they mean something entirely different now, and both of those things are true simultaneously. The grandmother’s memoirs are recursive — they loop back on themselves, revise, abandon. And the text messages are instantaneous, evanescent, a conversation that is happening in the same present tense as the reader’s act of reading.”

“So the structure of the story is the juxtaposition of these registers,” I said. “Fragments. An execution order next to a text message next to a page from the second memoir. The reader assembles the narrative.”

“The reader attempts to assemble the narrative,” Marías corrected. “Whether they succeed is not our concern.”

Cusk leaned forward. “But here is where I disagree with you, Javier. You’re comfortable with a reader who never assembles the narrative. I am not. I think the story owes the reader a shape — not an answer, but a shape. An emotional trajectory, even if the factual trajectory remains fractured. By the end, the reader should feel something has shifted, even if they cannot name what.”

“An emotional trajectory is still a trajectory.”

“Yes. And I am comfortable with that. The outline novels are built on the idea that other people’s stories, told in their own words, in their own structures, accumulate into a portrait of the listener that the listener never directly provides. The same principle applies here. The siblings never tell us what they feel about their grandfather. The documents tell us. Not because the documents contain the feelings, but because the selection — which documents to include, in which order, the white space between them — that is the emotional argument.”

“The white space,” I repeated. I was writing this down in my notebook, which was the first time either of them had seen me write anything.

“The silence between the fragments is where the story actually lives,” Cusk said. “The Pacto del Olvido is not a pact to forget. It is a pact to leave a very specific kind of silence in the place where speech would normally go. The story should reproduce that silence structurally. Not by being vague. By being precise about what is absent.”

“And the sibling who wants to burn,” I said. “Where does that sibling live in this structure?”

“That sibling is the one who understands the structure best,” Marías said. “Burning is not denial. Burning is the recognition that these documents have a half-life — they will continue to irradiate everyone who touches them for generations. The sibling who wants to burn is not saying ‘this didn’t happen.’ The sibling is saying ‘knowing this will not undo it, and keeping these papers will cause more suffering than the suffering they document.’ That is a morally serious position. It may be wrong. But it is not ignorance.”

“And the sibling who wants to publish?”

“Equally serious. Equally potentially wrong. The sibling who publishes is saying that the victims’ silence has lasted long enough, that the grandson’s discomfort is a trivial price for historical reckoning. But that sibling may also be performing something — performing righteousness — in a way that substitutes the gesture of exposure for the harder work of actually sitting with what the grandfather did.”

“And the third sibling?”

Neither of them answered immediately. The football on the television had given way to a panel of men in suits gesturing at a tactics board.

“The third sibling,” Cusk said, “is the story. The third sibling is the one who is still reading the documents. Who has not yet decided. Who is, in that sense, the reader’s surrogate — not because the reader will identify with indecision, but because the third sibling is the one still inside the act of interpretation. Still turning the pages. Still arranging the fragments. The story ends — or it fails to end, which is more honest — while the third sibling is still reading.”

“Still reading their own family,” Marías said. He said it like he was hearing it for the first time. “Yes. That is the correct verb. They are reading their family. And the story is the record of that reading. Incomplete, biased, full of gaps where the binding has come apart.”

I wanted to ask about the title but something in the room had shifted and asking felt like an interruption of a thought that had not finished arriving. Cusk was looking at her tea. Marías was looking at the television, which was now showing advertisements. Outside, a siren passed — close, then far, then gone. The kind of sound Madrid absorbs without registering.

“One more thing,” Marías said, without looking away from the screen. “The execution orders must be described with absolute precision. The paper quality. The typeface. The ink. The way the signature looks — whether it is hurried or careful, whether the pen pressed hard or lightly. These details are not decoration. They are the moral content of the story. A careful signature on an execution order is a different moral object than a hurried one. Both are damning. But they are damning differently.”

“And the reader will have to decide which is worse,” I said.

“The reader will not be able to decide,” he said. “Which is the point.”

Cusk was already putting on her coat. “I want to add one thing about the text messages,” she said, standing. “They should include the time stamps. Not because time stamps matter narratively, but because a text message sent at 2:14 a.m. is a fundamentally different document than a text message sent at 10:30 in the morning. The time tells you whether the sibling is thinking clearly or lying in bed unable to sleep. It tells you whether the message was composed in daylight or in the dark. These are the kinds of details that feel minor and are not.”

She left. Not abruptly — she simply finished what she had to say and stood and the leaving was part of the sentence. Marías and I sat for another minute. He ordered a second coffee. I looked at my notebook, which had three sentences in it, none of which I could fully decipher.

“The grandmother’s third memoir,” he said. “The third attempt. It should be the shortest. Perhaps only a page. Perhaps only a paragraph. And it should be the one that comes closest to saying the thing directly. Not because the grandmother finally found the courage. Because she finally ran out of ways to avoid it.”

He drank his coffee. I closed my notebook. The siren had come back, or it was a different siren entirely — there was no way to know, and it did not seem important to determine which.