Every family keeps a sentence no one ever questioned. Grandfather came over as a stowaway on a ship, thirteen years old. It was told at Sunday lunch, at the wake, at the great-grandchild's christening, and each time someone added a detail: the name of the ship, the port, the reason for the flight. No one lied. The story was simply told so many times that it hardened, and what hardens stops being a memory and becomes truth. It is the same mechanism as the Greek myths and the old stage plays, told and retold until the final version overrides every other. I call it the crystallization of the family myth, and it is one of the first things I learn to recognize in every case that reaches my hands.
Two ideas that tend to travel together are worth separating. Crystallization is the process, the way a narrative settles through repetition and accretion, from one generation to the next. False family memory is the result, what the family comes to hold as recollection and what, set against the document, does not hold up. One produces the other. But there is a subtlety that changes everything: not every crystal is false all the way through. Almost always there is, at the bottom, a true core. A real origin, real blood, a crossing that genuinely happened. The noise gathers on top; the truth stays underneath.
I like to think of it as a forgotten photograph. At first the portrait has a full name written on the back: Roberto Antonio, Joaquina's brother. A generation later, the portrait is only Roberto, grandmother Joaquina's brother. In the next, it loses its own name and is merely grandmother Joaquina's brother. And at last it is nothing more than an old photograph at the bottom of a drawer, a face no one can name. Information does not vanish all at once; it degrades, losing one datum with each passage. And where the datum falls, the family fills the gap with narrative. The myth, in the end, is the dressing that memory invents to cover what loss laid open. Crystallization and erasure are the same phenomenon seen from two sides: what hardens on top is exactly what was lost underneath.
That is why the error almost never lies in the material the family keeps. It lies in the conclusion the family draws from it. There is a sentence I hear at nearly every first meeting: grandmother was Portuguese, fair-skinned, light eyes. The trait is true; European blood really does run in the family. The mistake is to fasten that blood to the most vivid figure of affective memory, the one reached with one's own eyes, rather than to whoever actually crossed the sea. Often it was not the grandmother; it was her mother, or her father, a generation above. The same holds for accent. But grandmother spoke with an accent, they say. Of course she did. A child born in Brazil, daughter and sister of Portuguese, raised in a house or a colony where Portugal was heard all day long, without the mixing that school and the street bring today, spoke the way she heard at home. Accent is the fingerprint of the childhood environment, not of the birth certificate. The family reads it as a passport, and it is only the echo of the house.
I had a case that sums this up better than any theory. A client had always known his father was Portuguese, born in Portugal; he even believed the Brazilian paperwork had been forged, something that did in fact happen, to make life easier in Brazil. It was the baptism that dismantled his certainty, and it is worth understanding why. A birth registration is made on the word of the declarant; the child need not be present, which makes it a document relatively open to tampering. A baptism is another matter. The child is there, before the priest, at the font, and the priest records what he sees. He recognizes a newborn of a few days and tells it apart from a child of one or two years. You cannot falsify the age of a body present before a witness. When I located that baptism, it spoke of a child of a few days, and the civil certificate, drawn up a day after the birth, said the same. The father's passport showed the family entering Brazil with the older daughters and a babe in arms, and the interval between that arrival and my client's birth proved they were all already here when he came into the world. He was born in Belém, not in Portugal. And yet his memory was not pure invention: he really had spent part of his childhood in Portugal, there really were Portuguese sisters, some of whom later returned to Portugal. The recollection was true in what he had lived, and false only in the point where he had pinned it, the birth. The baptism was the nail that returned that memory to its right place in time.
It was to avoid erring in this way that I built, over the years, an order of work I try never to abandon. The first step is always the document, before the conversation. I ask the family to send what they have, I hunt down what is missing, and I read it all alone, in silence, until I form a reading of my own, raw, resting on the evidence alone. Only then do I sit down to hear the oral history. The reason is simple and it is hard: if I hear the family first, I enter the documents with the myth already assembled in my head, and the myth has a perverse power, that of making me see confirmation where there is none. The interview, when it comes, is done without leading. I do not ask whether the grandfather was from such and such a town, because the question itself plants the answer. I ask about sensory and seemingly incidental details: whether he complained of the cold, whether he spoke of a river, of a food, of a brother, of an uncle, of some relative he had over here. These are triggers that make the true memory come back on its own, without my putting anything into it. It was a question like that, about a sister, that solved a case I had been working for months. An elderly cousin mentioned, in passing, that the grandmother had an older sister in Rio de Janeiro. The woman I was looking for appeared in no passenger list because she had entered as a child, on her older sister's passport, through the port of Rio, though she went on to live in the Northeast. The fragment no one thought important was the one that opened the door.
I must confess, however, that the historian himself can be contaminated, and the costliest lesson I learned bears the name of a town: Lalín. It was an informal case, a face-to-face conversation right at the start. The family's certainty was enormous; the client was already preparing to travel to the town just to fetch the document, so sure was he. And I, faced with such conviction, with people who had lived alongside the man, took it as fact. I inverted my own rule: I went first to what the client indicated and only afterward to the sources. I spent months in parishes without pausing to notice what the paperwork already had to tell me. When at last I reconstructed the origin of that certainty, I discovered it had been born from the mouth of a cousin, and never from a document, never from the mouth of the immigrant himself. The town had never once been named by him. Crystallization, I realized then, does not stop with the family; it has a final link, and that link can be me, if I lower my guard. That secondhand certainty convinced me because it dispensed with the test, and that is exactly why it deceived me.
And there is a second form of contamination, quieter, that comes not from outside but from within me. Over the course of a difficult case, I accumulate hypotheses, and each one that failed leaves a residue. There comes a point where I no longer read the documents with clean eyes; I read them through everything I have already tried. It is what some call reading fatigue. It is not that I stop looking at the document; it is that I stop seeing it. From rereading so often, I no longer read the paper, I read my memory of it, certain there is nothing left inside to extract. And it is precisely in that certainty that the small detail hides, a word in the corner, a date that does not match, a note in the margin, because I have decreed that I know the whole document. Familiarity blinds. When I realize I have stalled in this way, the remedy is not to insist; it is to set the case aside for fifteen or twenty days, let the residue settle, and then return to the documents as though I had never read them. Distance is what restores the raw eye. At bottom, it is the same discipline as seeing the document before the conversation, now turned against myself. The three forms of certainty that deceive me, the client's certainty, my own hypotheses, and the belief that I have already exhausted the document, share the same root: certainty dispenses with testing, and it is that dispensing that lets the thing slip through. Method is, above all, an engineering built to force me to doubt what I think I already know.
There is, finally, a part of the craft that no manual teaches, and it is the most delicate of all: the weight of what one finds. That is why, at the first meeting, I ask a question that hands the client back control of his own story. Do you want to know everything that is discovered, or only what matters for the citizenship? Some want only what serves the process; others enter into the spirit and want it all. I ask because the drawer sometimes holds a wound. There are marriage certificates that note, in a corner, that the union took place by judicial decision, and not every judicial-decision marriage hides the same thing. But some, in turn-of-the-century Brazil, carry a hard story behind them: violence followed by marriage as the aggressor's way of escaping prison. It was neither the rule nor routine, but it happened, and it appears in the records with some frequency. To reach a descendant and translate that into what it truly was is a weight one learns to carry from no book. The same holds for the old categories of the registers, the natural child, the adulterine, words that today sound like an insult and that many receive with anger, as if I were accusing them, when I am only reading the legal language of a time that has died.
Faced with this, what I do is cut and polish. I remove the sensitive information when it sits in a corner of the certificate and is not needed for the process. I pass the document straight to the lawyer when the client does not need to see it. I give context to what cannot be hidden, so that the young woman of the nineteenth century is not judged with today's eyes, so that a descendant understands that “natural child” is a historical label, not a verdict on the person. To cut and polish is not to lie or to erase; it is to choose the face by which the truth is delivered, and its measure. But there is a fine line, and it must be seen clearly: cutting and polishing cannot become falsifying. There is the duty to protect the living from the brutality of the document, and there is the duty not to betray what the document says. The historian lives in that tension. He decides how far to polish the stone, but he has no license to invent a stone that was not there. Care for the living, faithfulness to the dead.
That, in the end, is the meaning of the work. To disprove Lalín, to prove that the grandmother was not the Portuguese one, to show that the boy was born in Belém, none of this is to destroy a family's memory. It is an archaeology: to scrape away, carefully, the layers of noise each generation laid down, until reaching the crystal that lay at the bottom. The myth, false to the letter though it may be, served its purpose, gave identity and belonging, and deserves respect for that. But the real Roberto, with a name, a town, and a document, is always more than the anonymous face forgotten in the drawer. The historian does not steal the myth. He gives back to the family what time had already taken from it.
From your roots to your future.