* * *
When you look across history and across cultures, one of the most common ways to identify people in names, once you’ve established their “given name” (i.e., the principal label that identifies the individual), is by indicating whose son or daughter they are, that is, by using a patronym. (These names may sometimes refer to the mother rather than the father, but we’ll use “patronym” for the general category.) When you look at modern inherited surnames, you can see the traces of this type of name in many languages.
It is also the case that words for close kinship are extremely stable across time. One of the classic ways of illustrating the relationships and differences within the Indo-European language family is by tracing the words for “mother” and “father” in the various branches.
Given these two observations, it is somewhat peculiar that within one closely-related language family, the Celtic languages, we find some curious discontinuities in the words for “son” and “daughter” that appear in the various recorded languages. (Patronyms may also be formed without overt words meaning “son” or “daughter”, for example, using a possessive form, but this article is only looking at names with the overt elements.)
While the various Gaulish languages (whose relationships are debated) are not exactly direct ancestors of the modern Celtic languages, they represent the oldest examples from that family.
I have yet to find a Gaulish name formula that includes an independent word meaning “son”, however both men and women may bear patronyms indicated with a suffix -gnos (masculine) or -gna (feminine) that has a sense of “child of, progeny of”. (In fact the -gn- part of it is directly cognate with the “gen” of “progeny”.)
Bratronos Nantonignos (Bratronos son of Nantonos)
Severa Tertionigna (Severa daughter of Tertiu)
A related word that shows up in Gaulish inscriptions is geneta, meaning “a young woman”, quite possibly with the implication of “a daughter”, although it isn’t found in name formulas (that I know of). Remember this word; it becomes relevant later.
There’s a name that shows up in Gaulish inscriptions, Maponos, which is based on the root word mapos meaning “son, young man”. I have no examples of mapos being used in patronymic name formulas, but keep it, too, in mind for later.
Daughters show up in one set of names indicated by the word duxtir (which, in fact, is cognate with English “daughter”).
Severa duxtir Valentos (Severa daughter of Valens)
Goidelic Langauges (Irish, Scottish, and Manx Gaelic)
So you remember that Gaulish word geneta I told you to remember? In the Goidelic branch of the Celtic family, the cognate of this becomes the standard word for “daughter” in name formulas. In the earliest records, written in Ogham letters, it appears as inigena.
Inigena Cunigni Avitoria (Avitoria daughter of Cunignos)
By the Old Irish period it had become ingen, then modern Irish inghean, continuing to appear in name formulas. The Scottish Gaelic form is slightly different but still clearly from the same origin.
Lasarfina ingen Chathail (Lasairfhiona daughter of Cathal)
Remember that Gaulish name based on the word mapos? In Irish, sons are now using a word cognate with this that appears in the earliest Ogham inscriptions as maqas.
Andagellas maqas Caveti (Andagellas son of Cavetas)
This is, of course, the familiar mac of later Irish and Scottish surnames, which are derived from literal patronyms.
Aodh mac Maghnusa (Aodh son of Maghnus)
But that’s not quite the end of the story. Remember the Gaulish word duxtir? It still shows up in names, but as an internal part of given names. Early Irish given names are formed in a number of interesting ways, one of which involves a noun (like “son”, “daughter”, “devotee”, “hound” etc.) followed by a possessive name. (In fact, this would be an interesting topic for another name squib.) And the word used for “daughter” in this context is a reduced form of the same root as Gaulish duxtir.
Dar-Erca (a feminine name)
Mac-Erce (a corresponding masculine name)
Brythonic Languages (Welsh, Breton, Cornish)
I’ll use Welsh as the example here because it has the largest set of early name records. Sons use the word map, cognate with Gaulish mapos and Gaelic mac. This shows up in the earliest records in the regular form map, but by the later medieval period has been reduced to ap, although the ordinary noun for “son” remains map.
Higuel map Caratauc (Hywel son of Caradog)
Hywel ap Caradog (the same)
Daughters, on the other hand, use an entirely different word from any of the ones seen before: merch. This shows up in the earliest records in the regular form merc. It sometimes became reduced in later centuries to ach, similarly to how map was reduced, but for various socio-linguistic reasons, women’s names were more likely to retain the full, regular form of the word.
Elen merc Loumarc (Elen daughter of Llywarch)
Elen ferch Llywarch (the same)
But remember that other daughter word, Gaulish geneta and Gaelic inghean? That’s still used in Welsh as geneth, but only as an ordinary word meaning “a girl” without any sense of meaning “daughter”. And there is no equivalent word to Gaulish duxtir and Gaelic der used in any sort of “daughter” sense in Welsh.
Although we think of relationship terms as being fairly conservative, and despite the fact that the Celtic languages have consistently used some sort of formula indicating parentage in names, the formulas that we see in use seem to have been re-invented regularly over the centuries, showing traces of connections between the languages but finding new ways of expressing the same concepts.