By Mike Haseler

Thanks go to David Petts whose questions I will try to answer in this article.

This is a third article in a series. The first two article were:

What languages at the eve of the Roman Invasion

Gaelic (Green), Welsh (Red),
Pictish (Blue) Wikipedia

Let's start by looking at the typical idea of the languages of the UK prior to the Roman invasion as shows to the left. This show the typical unsubstantiated assertion that we get. This is really just a "nationalist" wish list: Ireland is Irish, Scotland is Pictish and England is Welsh. The English are nowhere to be see.

Grey areas of Welsh like language
around 6th century

The next stage in the classic idea of the evolution of British languages is that of the genocidal Anglo Saxons who ethnically cleansed all of England in the 6th/7th century. But strangely without leaving a sign in the archaeological record. After this event the map now looks more like the one to the right with the "Welsh" who Nationalists call "Britons" pushed to the West.

But notice how well the area occupied by the Welsh fits in with the Historia Brittonum description for the Picts:

After an interval of not less than eight hundred years, came the Picts, and occupied the Orkney Islands: whence they laid waste many regions, and seized those on the left hand side of Britain, where they still remain, keeping possession of a third part of Britain to this day. (Historia Brittonum 12)

And there is some support for the idea that the "Welsh" were invaders because the word used for them did not refer to a place. Instead the English "Welsh" derives from the Old English Wilisc, Wylisc (West Saxon), Welisc, Wælisc (Anglian and Kentish), from Wealh, Walh: non-Germanic foreigner. However that could result from Welsh invasion from Wales to a Germanic area as much as any idea that the Welsh invaded Britain.


So, one could just as equally make out a case for a Welsh genocidal group invading from the West as an Anglo-Saxon genocide from the East. Except for the small problem that there is no genocide recorded in the archaeological record.

However, broadly speaking there does appear to be a DNA divide which suggests that Welsh was being spoken to the west of Oppenheimer’s DNA divide, whilst a proto-English was being spoken to the east of the Pennines. So, this would appear to be have been the situation in Roman Britain because there is little reason to think there was any change as the idea of the Anglo Saxons ethnically cleansing the East of Britain has been discredited.

But it was probably not that simple. The lack of obvious major change in population at the Anglo Saxon invasion together with place names makes it fairly certain that there was a significant area where the language before this invasion was of a Germanic type possibly closely related to Anglo Saxon. Just to be clear Anglo-Saxon is a Germanic-type language which is also called "Old English".

However, it does not appear possible to explain the Roman names just by Germanic origin. So, I am forced to consider other intrusive languages. But even Welsh area may be problematic as I'm not entirely happy that the Welsh areas had Roman tribal names derived from Welsh. But it is difficult to explain the presence of Welsh unless it were here at the Roman invasion. So, I am not confident as to when Welsh arrived. It does not fit in the linguistic map. Except for the close relatives of Breton & Cornish (and perhaps Pictish) it is a "language out of place" with no close relatives (Irish is not a close relative).

A Working Hypothesis

A "best guess" working hypothesis of the main language
influences in Britain at the Roman invasion

It appears likely there were more than two languages. As a "working hypothesis" I would expect:


A form of (Old) French to be spoken by the elite in the SE of Britain from around 100BC (arriving with cremations and other related archaeology).

Yorkshire Chariot Burials

I also think there would be a "Yorkshire" group of Germanic speakers dating from about 500BC associated with Chariot burials . Would 500 years be enough for this to become more of a regional accent than a different language by the Roman invasion?


In Scotland, we have a Pictish language with records of the Picts from 297AD, but it is proving difficult to reconcile the evidence as to where this was spoken and the Celtic myth and lack of rigour in Scottish archaeology makes this all the more difficult. The best guess is that it was spoken in SW Scotland as St.Ninian went to evangelise the Picts and this is where he set up his base. The Northern limit is very difficult to place and may well have extended much further north at later periods.


My best guess is that Gaelic is in some way related to the Romance Languages on the Iberian Peninsula as Gaelic appears to have some similarity to Spanish. The names Hibernia and Iberi also hint at a connection. So very tentatively I am attracted toward the idea of a very early emigration from the Iberian Peninsula to (parts of?) Ireland and NW Scotland which correspond to the present areas where some Gaelic is Spoken.


At a guess, it would seem likely that some Norse settlers were present in NE Scotland from an early period.


To clarify the situation. The evidence points to Gaul being a proto-French or Proto-Romance (i.e. the language group with French, Italian and Spanish) speaking area with the Celtic area corresponding with the area where the French regional accent known as Languedoc was spoken as this appears to correspond with the area given by Caesar for the Celts.

Just as England was invaded by the Anglo Saxons which had been blamed for completely removing all traces of a former language, so France had its own invasion of the Franks which it was assumed had removed any traces of Gaulish. However that assumption has to be questioned as there is very little evidence for a language change and accounts suggest a relatively small number of people in a ruling elite which is unllikely to have completely changed the language.

So "Celtic" was probably one of the Romance languages which was probably similar to the proto-romance langauges spoken North of Italy and in Eastern Spain. We have to assume that the Gauls (other than Celts) all spoke a similar Romance language, but with considerable regional differences leading to the major divide in French between the Celtic and other areas of Gaul.

Belgae in Britain?

The Roman writers do not speak of any Celts in Britain. However we know there were Belgae Gaul (not Celtic Gaul) invaders in the SE of Britain as we have Roman texts referring to them. Could these be described as Celtic? They were "Belgae". The Belgae were one of the subdivisions of the Gauls and another was the Celts. It is likely they all spoke not too dissimilar languages. However, it appears to be a ruling elite, and given the time it took to assimulate the French/Norse invaders after 1066, it is likely this group was also more of a regional accent by the time the Romans left.

Latin Britain?

Over the course of Roman occupation we see Latin the language of governance impacting on proto-English in Britain just as it must have had some impact on proto-Romance languages in Gaul and Spain. But Latinisation does not appear to be that strong.

Bede writing about 731 talks of five languages: English, British, Scottish, Pictish, and Latin. As Latin was used by the Church it may not have been in use outside, but there is the possibility of some residual Latin speaking elements remaining from Roman times. If however this is a residual population from the Roman empire, it does suggest that Latin retained a separate identity and wasn't integrated. Again this reduces the chance there was considerable "Latinisation" of indigenous languages making it more likely what we see now is the result of what was there before the Latin language arrived with the Roman empire.

What does a possible Germanic origin of Avon mean?

The fact that Avon could be derived from Anglo Saxon: hæfen meaning 'haven, port' which in turn possibly derives from "Efen" meaning "calm", suggests a possible root for Rivers named "Avon" of *æfen meaning an area of calm used as a harbour.

However the fact it is possible to give a germanic language origin only disproves that the river name must be Welsh. It could still be Welsh, but it does not have to be Welsh and could also be Germanic/Anglo-Saxon. So more evidence needs to be taken into account. If this other evidence suggested Germanic - I would opt for the Anglo-Saxon derivation. If other evidence suggested the area was proto-Welsh, then I would opt for the Welsh derivation.

Britanny and Magnentius’s armies

We have an account that the reason Britanny is so Called is because a Roman Emporer from Britain took his men there. Britanny also happens to be a Breton speaking area. Breton is like Welsh.

I'm not sure about how the Welsh-like language got to this area. David has a point if as he says they were stationed in areas that the above map suggests would be Germanic. Several possibilities present themselves:

  1. For some reason Magnentius army was mainly Welsh speaking ... and Welsh was taken to Brittany as an intrusive language given the area both the Breton Language and the Britanny name.
  2. For some reason Magnentius army was mainly (or partly) Welsh speaking ...but they stayed in Breton because it was already Welsh-like. And only the name "Britain" came with the army
  3. Magnentius army was not mainly welsh ... it just happened to end up in Brittany which just happened to have spoke a Welsh like language. Perhaps because the area had links to Wales. So, Magnentius only brought the name Brittany and didn't have anything to do with introducing Welsh like Breton to the area.


It was stated there are several hundred Gaulic inscriptions and the question was asked whether these showed clear signs of being a proto-romance language.

However, despite there being reported as several hundred inscriptions, they do not provide a large lexicon as the Gaulish glossaries only contain about 160 words. Many are fragmantary and/or only partly in Gaulish. An example of one of the better ones is the Pillar of the Boatman. This is written in Latin with some Gaulish language features and reads:

Tib(erio) Caesare /
Aug(usto) Ioui Optum[o] /
Maxsumo /
nautae Parisiaci /
publice posierunt //
Eurises // Senan[t] U[s]e[t]lo[n][-] //
Iouis // Taruos Trigaranus //
Volcanus // Esus //
[C]ernunnos // Castor // [---] //
Smer[---] //
Fort[una] // [--]TVS[--] // D[--]
(CIL 13, 3026; RIG L2-1)

Much of the text is missing, what is there appears to be largely names. Another of the better inscriptions is the Bern zinc tablet found in the 1980s in Bern. It is inscribed with an apparently Gaulish inscription, consisting of the four words, each on its own line, the letter formed by little dots impressed onto the metal:

(Dobnoredo Gobano Brenodor Nantaror)

The longest known Gaulish text was found in 1983 in L'Hospitalet-du-Larzac. It is inscribed in Latin cursive script on both sides of two small sheets of lead. Probably curse tablets (defixio), they contain magical incantations regarding one Severa Tertionicna and a group of women (often thought to be a rival group of witches), but according to Wikipedia: "the exact meaning of the text is unclear".

Given the paucity of inscriptions and the mix of languages invovled, it is extremely difficult to draw any firm conclusions regarding these inscriptions. However from the lexicon of words that have been identified the following can be said:

Taking Welsh first ... they aren't welsh-like (i.e. Breton). As I showed that very few words in assumed Gaulish texts can be derived from Welsh.

Taking Irish ... as I said above. I find some credence in the idea that Old Irish is related to a proto-Romance language in spain. I would therefore be interested to compare possible roots in Old Spanish, Irish & Old French

Celti-Iberian Inscriptions

I am unable to find any accepted translation of Celtiiberian "texts". E.g. there are four Botorrita plaques of which only the one in Latin has a translation. It is therefore not possible to say anything about these except that it is possible that they have wrongly identified the alphabet so that the letters are being incorrectly read.

6th/7th Century changes

David highlighted the fact that therew was a radical transformation in the way in which place names are constructed in the 6th/7th century and something similar with personal names and asked why this was. Two things spring to mind. The first is that Roman Christianity came in which if it used a Latin Bible would have brought latin rather than Norse names into common use. The second is that the ruling elite changed. They would likely have "owned" places and so constructed names according to their own fashiion.


It has long been fasionable to reject Nennius as a source of history because he recounts a story about some dragons. However, this is a second hand story and it is abundantly clear it is not intended to be taken literally as it appears to be a form of divination from dreams:

"I will now unfold to you the meaning of this mystery. The pool is the emblem of this world, and the tent that of your kingdom: the two serpents are two dragons; the red serpent is your dragon, but the white serpent is the dragon of the people who occupy several provinces and districts of Britain, even almost from sea to sea: at length, however, our people shall rise and drive away ;the Saxon race from beyond the sea

David also highlighted the fact that Nennius mentions Troy. But again let us read what Nennius said:

Respecting the period when this island became inhabited subsequently to the flood, I have seen two distinct relations .... AEneas, after the Trojan war, arrived with his son in Italy;

The word used for "relations" is "experimentum" which also translates as "proof". So it seems Nennius is saying he has seen two possible derivations of the origin of the islands. He does not assert they are both true, indeed, they cannot both be true. So he is very clearly saying they are dubious derivations.

Also we must remember that "Nennius" was writing about 400years after Maximus in a relative nearby area. Homer & Troy was 2000years earlier is signalled as "dubious" because it is only one of two known accounts and it was much further away. It is therefore quite reasonable to give a lot more credence to the account of Maximus than to Homer.