The evidence for the native British naming system must be untangled from
the Latin of its recording. The names of non- citizens often feature
patronyms, such as Nectovelius f[ilius] Vindicis
(nom. Vindex). Given later Celtic habits, it would be fair to assume
that patronyms were a distinctive feature of British names -- but
recall that Roman names could include patronyms too. Several inscriptions
of British names use the relational term nepos which in Latin can
mean grandson, nephew, or descendent in general. Without a clearer context
it is hard to tell whether this is indicating a significant ancestor (as
Gaelic ó came to do) or perhaps alluding to the known
practice among some British tribes of matrilineal descent where the
mother's brother, rather than the father, is referenced. In addition to
relational bynames, non-citizens' names might include a tribal byname, as
in the case of Lucco Treni f[ilius] Dobunn[us] "Lucco, Treni's
son, the Dobunnian" .
Old Welsh Names
The single best source for personal names of the Old Welsh period is probably the collection of charters and land-grants in the Book of Llandav. I worked from the group covering the 6th through 10th centuries. The people recorded here fall into two categories: those connected with the church, either as witnesses or as recipients; and the laymen, both witnesses and donors. The two groups display a notable amount of overlap in the name patterns they use, although some patterns are more characteristic of one group than the other. The basic elements that appear are given names, patronyms, titles/occupations (which may be modified by a place), and a small handful of other bynames, the recognizable ones being adjectival.
Names using patronyms are primarily those of lay people. There are examples of filius, filia, and nepos. The word order is fairly free, since the Latin inflections remove most of the ambiguity. For the most part, the father is referred to by given name, but there are two examples of X filius episcopi "son of the bishop". (Take into consideration that the Celtic church differed with Rome on the subject of married priests.)
The greatest variety of occupational or titular bynames are found among the churchmen. There we find episcopus (bishop), magister (master, teacher), doctor (probably a teacher of some sort), presbiter (a religious functionary), scriptor (writer, scribe), lector (reader, lecturer), sacerdos (priest), abbas (abbot), archidiaconus (archdeacon). On the secular side we find rex (king), princeps (prince -- although this may instead be a religious title of some sort), and faber (smith). Sometimes these might have a location added: rex Gleuissicg, abbas Nant Carban, sacerdos Ilduti, lector Catoci. In the last two cases, the given name of the dedicatory saint stands for the church. Where in one case we find abbas Lannildut, elsewhere we find abba[s] Ilduti.
There are no entirely clear examples of a locative byname standing apart from a title or occupation. One possible exception is Berdicguent which might possibly be a given name Berdic run together with the place name Guent. The descriptive bynames all belong to laymen. Two famous kings show up with their nicknames: Huwel Da (good) and Morgan Hen (old). Besides those, we have one example that is undoubtedly description rather than a name per se, Dulon virgin, and one that is probably a Latin translation of mawr (great): magnus.
A handful of individuals were recorded with combinations of the above: a title plus a patronym, a patronym including the father's title, a patronym including the father's descriptive byname, a descriptive byname plus a patronym. It is sometimes difficult to draw the line between the recording of a name and the recording of a detailed description of who and what a person was. One borderline case is the longwinded Lifris filius episcopi, archidiaconus et magister Sancti Catoci, "Lifris, son of the bishop, archdeacon and master of St. Cadog's".
Four women show up in the records. They appear with the same types of
bynames as the men, with the exception that none bears a title or
Roman-Era Given Names
So what are we looking at in the way of given names? For the Roman period I use "given name" to refer to a native British name, whether used as such or incoporated in a triple name as a cognomen. A glance over a far-from-complete list of the available British names tells us that they follow the standard Indo- European habit of di-thematic compounding. (In this respect, it should be noted, the Roman naming system is very anomalous.) We can spot prothemes like Cuno- (dog) and Dumno- or Dubno- (deep, world), or deuterothemes like -vellaunus (meaning uncertain) or the ubiquitous -rix, cognate with Latin rex "king". The definitive study of early Celtic compound names has yet to be written, but a good place to start is D. Ellis Evans' Gaulish Personal Names. While Gaulish and British were clearly different languages, at the time of the Roman invasion of Britain some of the southernmost tribes were only a couple of generations away from having lived on the continent and were probably still more the former than the latter. Also, between the conservatism of personal names and the close relationship of the two langauges, the differences can be fairly subtle. At this point in the language the similarities with Gaelic cognates can be a lot easier to spot, too. Consider the case of British Dumnovall[os] and Old Irish Domnall.
In addition to compounded names, the same sorts of elements were used in one-element, uncompounded names. These might include a suffix of some sort that had no independent meaning, and they might be indistinguishable from shortened forms of compounded names. There is reason to believe that compound names may have been a mark of the noble class among the Celts, with craftsmen and lower classes being more likely to bear uncompounded names, but the evidence is mostly statistical and circumstantial . The simple structure of these names combined with the similarity of the sound systems of Latin and the Celtic languages can make it impossible in some cases to tell the origin of a particular name. Cassius, for example, is both a Latin nomen and a Latinized Gaulish given name .
As usual, if we were to work simply from documentary evidence, we would
conclude that women were very scarce in early Britain. There aren't really
enough examples from which to draw any firm conclusions about gender
differences in names. Latin names took their gender from inflectional
endings. The root had no inherent "gender" and could be used as
easily in a feminine as a masculine name. British feminine names are
inflected as Latin first- declension feminine nouns
 -- whether that
reflects how they were treated in British or not. But given the very small
number of feminine names we have from this period, the evidence is fairly
strongly against the proposition that men's and women's names were
strictly defined. Take, for example, the woman's name Bodicca which,
in addition to the famous queen, shows up in an inscription as the name of
a centurian's wife. But it also shows up as a nomen Bodiccius and as
the Old Welsh masculine name Budic. Another possible
masculine/feminine pair are the mid-6th century masculine Avitori
(probably from Avitorix), and late 5th century feminine
Avitoria (probably from Avitoriga), suggesting a way to
feminize any name with the -rix deuterotheme. But even more
intriguing is the example of the name Tancorix appearing in a
context in which it clearly belongs to a woman!
. The possibilities are
tantalizing but too uncertain for much usefulness. Note that the infamous
-wyn/-wen distinction doesn't exist yet at this point; the element
is spelled the same whether it appears in a woman's name Cuniovenda
or in a man's name Barrivendos.
Old Welsh Given Names
In my discussion of the names of Roman Britain I have distorted somewhat the variety of languages that could be found. People from all over the Roman Empire ended up in Britain: merchants from Syria, cavelrymen from Spain, soldiers from Tunisia. But my interest here is not simply in names found in Britain but in names with some British content, and I have not yet found examples of British elements combining with anything other than Latin.
In the Old Welsh period, the case was different. Here we find Welsh elements combined with Anglo-Saxon names, Biblical names (via the Church) and some decidedly altered Latin names. Although it is sometimes difficult to untangle them, there are also some fairly Irish-looking names mixed in with the Welsh -- not too surprising considering that the Welsh and Irish churches communicated closely.
Looking at the Welsh names, we find that the inflectional endings have disappeared in many cases, and where they appear it is certainly due to the Latin context of the document rather than any continuing tradition of inflection. (Note that the borrowed Anglo-Saxon names show an occasional Latin inflection too.) In addition, the names have undergone a bit of syncope (i.e., loss of unstressed syllables), most often in the last syllable of the protheme. Thus *Cunovellaunos becomes Cingualan. (Another miscellaneous sound change is British /v/ > Old Welsh /gu/.)
It is perhaps entertaining to take a particular name through the changes from British to Medieval Welsh. (The British form is highly conjectural.)
|British||Tancogeistla [filia] Marocanti Glasobrin[?]|
|Old Welsh||Tancoystl merch Morgant Glasvryn|
|Medieval Welsh||Tangwystyl verch Morgant Glasvryn|
|Modern Welsh||Tangwystl ferch Morgan Glasfryn|
Go to appendices
Editted and published by Arval Benicoeur