One of the most frequent mistakes of anyone who begins to investigate their own family is to believe that the surname worked in the past as it works today. People find two siblings with different surnames and assume there is an error, or that they are not really siblings. They find a great-grandmother with no surname at all and assume the document is incomplete. They find the same name spelled several ways and suspect fraud. None of this is usually an error. It is simply how a system with a very different logic worked, a system that, to modern eyes, can look like chaos.
The absence of a rule
Let us begin with the essential. In the old Portuguese world, before the nineteenth century, there was no rigid rule for the transmission of surnames. A child could receive the father’s surname, the mother’s, a grandparent’s, or even a godparent’s. The choice was highly flexible, and historians and genealogists sometimes describe this long period as one of onomastic indiscipline, because people used surnames different from their parents and siblings, drawing them from the father, the mother, the grandparents, the uncles, or wherever they pleased. The historian Nuno Gonçalo Monteiro observes that, there being no legal norm on the matter in Portugal, the first surname was as a rule the paternal one, but one could freely choose among those used by the parents or the four grandparents, and for that reason cases of siblings who did not use the same surname were frequent.
This means that the continuity of the paternal surname, which today seems natural and almost obligatory to us, was merely a tendency, not a law. A wish, a tribute, a convenience was enough, and the line broke.
The godchild and the godparent
One of the clearest examples involves godparents. Even when a male child might otherwise have been expected to continue the father’s surname, it happened that the family wished to honor the godparent, and the child ended up taking on the godparent’s surname instead. Families often chose the surname of the most important, richest or most influential family, the one whose name brought the child the most benefit. There was nothing irregular in this. It was a social strategy, and researchers need to account for it, because it is what makes a child appear, suddenly, with a surname that comes neither from the father nor from the mother.
Women were a different matter
If among men there was already variation, among women the picture is looser still. Many simply carried no family surname. As I explain in The Woman Who Did Not Change Her Name, it was common for a woman, especially of the common people, to bear only a name of devotion in place of a surname, a Maria da Encarnação, a Maria de Jesus, a Maria do Perpétuo Socorro, names of saints or patrons to whom she had been consecrated. That is not a surname. And whoever takes it for one sets off in pursuit of a lineage that never existed.
Among those who did have a surname, there was a tendency, not a fixed one, for daughters to follow the maternal line and sons the paternal. Let me illustrate with the kind of case I find. Imagine a Maria Balbina da Silva, daughter of a Francisco Mascarenhas. Note that she does not carry the father’s Mascarenhas. She carries Silva, because she was the daughter of a Hermelinda da Silva. And this Balbina’s daughter is once again a Maria José da Silva. The surname Silva passes through three generations along the maternal line, ignoring entirely the man’s Mascarenhas. Whoever searched for the daughter by the father’s surname would never find her. She is on the other side, in the line of the mother and the grandmother.
The same name, three faces
The point where this Babel becomes most visible is when it piles up in a single document. I have handled passports, which are official documents, in which the same woman is recorded under three different names. A Maria José da Moita, also known as Maria José Caranguejeira, also known as Maria José da Silva. Three recorded names for a single person, declared side by side on the same paper. And this is not an exception, nor a rare curiosity. It is common. It happens with women and it happens with men too. The same person could be recorded differently depending on who was writing the name, the parish priest, a neighbor, a registrar or a notary, and each fixed it in a different way.
Where the name settled
There is a practical reason for so much variation, and understanding it is half the battle in a genealogical search. At baptism, the priest usually recorded only the given name, without much concern for surnames, and it is common for the parents of the same person to appear with different surnames from one child’s record to the next, above all the mother. The name only settled, in fact, at the time of marriage. That is why marriage records are almost always the most complete and most reliable documents for identifying the surnames a family actually used. Researchers learn to anchor the analysis in marriage records and to treat the baptism with caution, knowing that there the name was still fluid.
The name of devotion that became a surname
All this fluidity, besides confusing the reader, hampers the research itself, because it often leaves us unsure of what to look for. Let me give the clearest example. Imagine a woman baptized in Portugal as Maria dos Anjos. There, on the far side, dos Anjos is not a surname. It is a name of devotion, like de Jesus or da Encarnação. She crosses the Atlantic carrying no family surname at all. And if she is an unmarried mother, what will she pass on to her children? The only thing she has. So the child becomes Manuel dos Anjos, Francisco dos Anjos, and what had been born as religious devotion becomes, now on Brazilian soil, a family surname.
The same happened everywhere the Portuguese migrated, in Brazil, in the United States, wherever it might be. The name of devotion, on crossing the border, hardened and became a surname. And here is the trap for the non-specialist. He takes the dos Anjos of the Brazilian grandchild and goes looking for it in the Portuguese records as though it were a hereditary surname, searching for an Anjos lineage that never existed in Portugal. The lead, which seemed solid, dissolves at the origin. A case like this can only be solved by understanding the naming conventions of that period, knowing that the same name may have been devotion on one side of the sea and a surname on the other.
Reading the chaos with method
Faced with all this, the worst a researcher can do is demand of the past the coherence of the present. If I search only for the father’s surname, I lose the daughter who followed the mother. If I discard a document because the surname does not match, I throw away the right piece. The Babel of Portuguese names is not gratuitous disorder. It is a system with a logic of its own, in which a name reflected less a fixed rule of descent and more the bonds, devotions and conveniences of a life. That is why I follow every line at once, the father’s, the mother’s, the grandparents’, the godparents’, and I read the variations not as errors to correct, but as clues to interpret. A name, in the old Portuguese world, was not a fixed label. It was a portrait of the place a person held among their own. And to learn to read it is to learn to see the whole family behind a single written line.
Another family of names also leaps from the records: Manuel Exposto, Raimunda Exposta, José Exposto, names that carry a history of abandonment and care in the word itself. But that is a subject for another article.
Mnésis, documentary, genealogical, and historical research. From your roots to your future.