Sephardic Naming Traditions

Naming customs are the most useful piece of Jewish tradition a genealogist can know, because they are the only one that lets you reconstruct a generation you have no records for.

They are also the reason your Sephardi family tree contains eleven men called Abraham.

How does Sephardi naming differ from Ashkenazi?

Sephardim traditionally name children after living grandparents. Ashkenazim name only after the dead.

That single sentence is the most quoted fact in Jewish genealogy and it is broadly correct, but it needs a small refinement to be useful.

Ashkenazim have a taboo against naming a child after a living relative. The reasons given vary — that it invites the Angel of Death to take the wrong person, that it uses up a name still in service — but the practice is consistent and strong.

Sephardim have no such taboo. Naming a child after a living grandparent is an honour to that grandparent, and it is the normal and expected thing to do. Sephardim also named after the dead; the difference is that they were not restricted to it.

The asymmetry matters for research, and it is easy to get backwards:

If you find a child named after a relative	Ashkenazi family	Sephardi family
What it tells you about the relative	They had died, usually shortly before. A strong inference, useful for dating deaths.	Nothing. They may have been alive or dead.
What it tells you about the family structure	Little, because there is no fixed order.	A great deal, because the order is regular.

So the Ashkenazi custom is better for dating deaths, and the Sephardi custom is better for reconstructing families. Use each for what it does.

What is the pattern?

Grandparents first, in a set order, alternating between the paternal and maternal sides.

The classical Sephardi order runs:

First son — the paternal grandfather
First daughter — the paternal grandmother
Second son — the maternal grandfather
Second daughter — the maternal grandmother
Thereafter great-grandparents, then the parents’ siblings, then other relatives

Regional variation exists — some traditions alternate the sides, so that the first daughter takes the maternal grandmother — and you should establish which convention a particular community used before relying on it. But the strong preference for the paternal grandfather in the first son is close to universal.

Two consequences follow immediately. The eldest son’s name usually tells you his paternal grandfather’s name, so if you have a marriage but no earlier records, the first child’s name is a hypothesis about the previous generation. And a sibling group is a partial reconstruction of two sets of grandparents: given four or five children in order, you can often name three or four grandparents, and then go looking for them.

What is it actually for?

Four uses, in order of value.

Reconstructing a generation. Where records end at a marriage, the children’s names propose the grandparents’ names. That is a hypothesis, not a finding — but it is a hypothesis that tells you exactly what to look for.

Testing an identification. You have two candidate fathers for the same man. One of them has a father whose name matches the eldest son’s; the other does not. That is real evidence, and it is independent of everything else you have.

Distinguishing families with the same surname. In a community where forty men are called Cohen, the pattern of given names across a sibling group is a fingerprint.

Detecting an Ashkenazi line in a Sephardi community, or the reverse. A family that never names after living relatives, in a Sephardi congregation, is worth a second look — and the reverse holds too.

Use it as corroboration, never as proof. The pattern is a custom, not a law. Families broke it constantly, for reasons ranging from a beloved uncle to a name they simply disliked.

Why do two names complicate everything?

Because the pattern operates on the Hebrew name, and most of the records you will meet give the civil one.

This is the point that most often defeats people, and it is specific to the Western Sephardim.

Most Western Sephardi men carried two names: a Hebrew or synagogue name, and a civil or commercial name. Both were real, and they follow different logics.

The Hebrew name obeys the naming pattern — Isaac’s eldest son is named for Isaac’s father. The civil name may be a rough equivalent, such as Moses and Moisés, Jacob and Jacques, Isaac and Isaque, or it may be entirely unrelated, particularly among families who had lived as New Christians. Manuel, Duarte, Diogo, Álvaro, Fernão and Gaspar carry no Hebrew content at all.

So the pattern can be perfectly intact in the Hebrew names and completely invisible in the civil ones.

This is the same problem described on our surnames and India pages, where Jaques de Paiva was Moses Zagache and Fernão Mendes Henriques was Isaac Ergas. If you are working from commercial records, notarial acts or shipping documents, you are seeing the civil name and the pattern will look absent when it is not.

Where possible, do the analysis on the Hebrew names. Circumcision registers, ketubot, congregational registers and gravestones give them; civil records generally do not.

What given names should you expect, by region?

Given names are far more informative than surnames, and unlike surnames they vary sharply by region. A name can often place a family within a generation or two.

Western Sephardim — Amsterdam, London, Hamburg, Bordeaux, Livorno, the Atlantic world — used a tight biblical core: Abraham, Isaac, Jacob, Moses, David, Solomon, Joseph, Samuel, Aaron, Daniel, Benjamin, Elias, Raphael, used relentlessly, alongside Iberian civil names.

Eastern Sephardim, in the Ottoman lands, used the biblical stock plus a distinctive layer: Behor or Bohor, literally firstborn and used as an actual given name; Nissim, Haim, Shabetai, Sinyoru, Yomtov, Menahem.

Maghrebi Sephardim, in Morocco, Algeria, Tunisia and Libya, used Messod, Makhlouf, Yamin, Shalom, Mimun, Amram, Aziza, Freha, Zahra — names shared with the surrounding Judeo-Arabic-speaking world and largely absent from the Western communities.

A name like Behor or Makhlouf in a family tree is a geographic marker, and a more reliable one than any surname.

Why are women’s names more useful than men’s?

Because there are far more of them, and because they are distinctively Sephardi in a way men’s names are not.

Sephardi women’s names are markedly more varied, more Romance and more secular than either Sephardi men’s names or Ashkenazi women’s names.

Alongside the biblical Sarah, Rachel, Rebecca, Esther, Leah, Judith, Hannah and Miriam, you find Gracia, Reyna, Luna, Estrella, Sol, Bienvenida, Palomba, Clara, Bella, Donna, Rica, Ledicia, Simha, Mazaltov and Buena.

These are Spanish, Portuguese, Ladino and Hebrew words meaning grace, queen, moon, star, sun, welcome, dove, joy, good fortune. They carry no biblical reference and they are distinctively Sephardi.

Two practical consequences. A woman called Luna, Reyna or Bienvenida is almost certainly Sephardi — these names have no Ashkenazi currency. And because the female stock is wider, women’s names discriminate between families far better than men’s do. In a congregation with thirty men named Abraham, the women’s names are what will separate the households.

What about patronymics?

Ben in Hebrew, bar in Aramaic, ibn and abu in Arabic all mean “son of”, and appear constantly in Hebrew documents: Yitzhak ben Avraham.

In Hebrew records the patronymic is often the only identifier a man has, which is why circumcision registers and ketubot — which name the father as a matter of course — are so valuable.

Many Iberian surnames are frozen patronymics: Rodrigues is “son of Rodrigo”, Fernandes “son of Fernando”, Lopes “son of Lope”, Nunes “son of Nuno”, Gomes “son of Gomes”. These were surnames in the Christian population too, which is one reason they do not distinguish Jews. See our surnames page.

What do double names mean?

Often, that somebody was seriously ill.

Double given names are common — Abraham Israel, Isaac Haim, Moses Raphael, Jacob Israel — and the second element is often doing specific work.

It was and is customary, when someone is gravely ill, to add a name in the hope of changing their fortune. The names used are Haim, meaning life; Raphael, God heals; Hizkiyah, God strengthens; and Alter or its equivalents.

This is a genealogical signal. A man who appears as Isaac in early records and as Isaac Haim in later ones has very probably had a serious illness in between — which dates something, and which explains an apparent name change that might otherwise look like two different people.

A posthumous child was frequently named for the father who had died before the birth.

Did the tradition die out in the West?

We think it weakened substantially among the Western Sephardim, and we think it is the most interesting open question about Sephardi naming. Here is the evidence on both sides.

Reasons to think it weakened

Integration. The Anglo-Dutch Sephardim were the most integrated Jewish communities in Europe, and by the eighteenth and nineteenth centuries their given names were changing. The London Sephardi elite of the Victorian period was using English names — the progression from Moses Montefiore to Claude and Leonard Montefiore is a family history in three names.

Intermarriage. From about 1810 Sephardim and Ashkenazim in Amsterdam intermarried regularly, and the two naming customs are directly incompatible: one requires naming after living grandparents, the other forbids it. In a mixed family one tradition has to give way.

Secularisation. The nineteenth-century Western Sephardi communities were among the first to secularise, and secular naming follows fashion rather than family.

Reasons to be cautious

The pattern may be hiding in the Hebrew names. As set out above, the Western Sephardim used two name systems, and the pattern operated on the Hebrew one. A study based on civil records would find the tradition had died out even if it were completely intact. This is the most likely source of error in the whole question.

Small endogamous communities recycle names anyway. In a congregation of a few hundred interrelated families, the same names circulate regardless of any rule — which can look like a pattern where there is none, or obscure a real one.

And nobody has actually measured it.

Which means this is a question that can be answered

We hold the data to test this, and as far as we know nobody ever has.

Our database holds nearly two hundred thousand people across communities including Amsterdam, London, Curaçao, Charleston, Livorno, Rome, Tunis and Mogador, with parent-child links. That is enough to ask, for each community and each period: what proportion of eldest sons carry their paternal grandfather’s name?

If the answer is 70 per cent in Mogador in 1850 and 25 per cent in London in 1850, the tradition demonstrably weakened in the West and we will be able to say when. If it turns out to be 65 per cent in both, we will have learned something more surprising.

Either way it would be the first empirical study of Sephardi naming patterns, and we intend to do it. If you would like to support that work, please get in touch.

Why is everybody called Abraham?

Because the tradition works, and there is a cost to a strong naming tradition that falls entirely on the genealogist.

In a small, endogamous community that names children after grandparents, the same names recur relentlessly — and not just across generations but within them. Four brothers each name a son after the same grandfather, and now there are four living first cousins with identical names in the same town at the same time.

This is not a hypothetical problem. It is the single commonest cause of error in Sephardi genealogy, and it produces confident, plausible, completely wrong family trees.

Two real examples of what it looks like in the records. In the circumcision registers of Mogador, a name-based reconstruction produced households in which the sons’ births spanned more than thirty years, and one man apparently fathering children across a hundred and three. And in the marriage registers of Tunis, which run from 1788 to 1881, a single “Isaac Lumbroso” appeared to have acquired twenty-one children — because at least three men of that name, from a major rabbinic family, had been merged into one.

The practical defences:

Never treat a name as an identity. Two Abraham Cohens are two men until you have evidence otherwise.
Always apply a generation window. A register spanning ninety years contains three generations of the same names.
Check the spans. If a man’s children are born across more than about thirty years, or if he is fathering children at seventy-five, you have merged two people.
Use the women. The wives’ names are the most efficient way to separate identically-named men.

Research services

If you are trying to establish which of several identically-named men is yours, or to reconstruct a generation from a sibling group, that is exactly the kind of problem the naming pattern is good for — and exactly the kind where it is easiest to go wrong. If you would like help, including for a Portuguese or Spanish nationality application, you can hire a professional Sephardic genealogist.

If you have found this page useful, please consider making a small donation to support this site.

On this page