Featured post

Gender in the Merovingian World

Friday, 5 August 2011

Help required

From linguists/philologists - maybe from Alex Woolf in particular!

I went for a drink with an old friend - Steve Brohan - whom I hadn't seen for ages and since we have a shared interest in post-imperial Britain discussion turned to that subject.  Now, Steve proposed a potentially radical idea, which I really like, concerning linguistic change in Britain.

The problem is this: There are pretty well no Brythonic borrowings in Old English.  There are, however, many hundreds of Latin ones, which (alongside other evidence) is leading to some people (including me) suggesting that the language of lowland Britain, encountered by the Anglo-Saxons, was a Romance low Latin, rather than a late Brythonic/proto-Welsh.  However, if this was the case, why did the 'Saxons' not end up speaking a Romance dialect, like (say) the Franks?  Hitherto, answers have focused on the scale of migration.  This is (I think) half right.  Right in that it might be down to different numbers of incomers, but wrong in that I think that that 'scale' is to be measured in chronological terms (that is to say, large numbers but over a long time-span, topping up the cultural value of English).  Only half-right too in that it ignores the other element of the equation, which is that Roman culture had collapsed more severely in lowland Britain than anywhere else (other than the far north of Gaul, where there was similar linguistic change).  Thus there was possibly much less cachet attached to Latin and Roman culture.  That's the aspect I have focused on in the past.

Yet, as Steve pointed out, this can't be entirely satisfactory, because Roman identity did matter so much to the Anglo-Saxons when we know about them from written sources - and it also mattered in the highlands.  He suggested that the dominance of Old English might have become because it was used as a politically dominant 'lingua franca' in areas where British and Latin were spoken.  That would work in that highland/lowland border zone - along the edge of the villa zone, where the greatest prosperity in late Roman period is attested, and where the most powerful AS kingdoms emerge.  The scenario (if I remember correctly) would run like this.  Pre-Anglo-Saxon British highlanders would know some Latin but not much - enough to be able to make transactions with lowland villa-owners etc, especially to pay taxes and so on.  The villa owners, by contrast, would know no British.  When an Anglo-Saxon military elite came to power, however, both would need to learn Old English to communicate with these warrior aristocrats, and knowing this language would enable them to communicate with each other in the new set up. 

Thus Old English would become a dominant language first on the frontier between the villa zone and the highland zone.  As this was an economically prosperous area in the late Roman period, and a politically dominant one in the middle Saxon period, some 'continuity' of dominance might be suggested.  The written sources suggest that the period around 600 was one wherein economic and other changes allowed kingdoms in the south-east briefly to acquire dominance before this was reclaimed by the 'border-zone' kingdoms (Mercia/Northumbria/Wessex) - the earliest action of Aethelberht of Kent is to defeat his predecessor as over-king, Ceawlin of Wessex...

I wonder if yet another aspect of the tyranny of the invasion narrative is that we always think in terms of Anglo-Saxon cultural influence moving 'of necessity' east to west.  But is it possible that it was first established in the west of the lowlands and then spread back east on the back of political dominance?  The archaeological traces of 'Saxon' culture are after all found right across the 'Saxon' zone from the start.  If they are less numerous on the 'frontier' then it is time we started seeing that as a result of the fact that furnished inhumation (symptom of instability) was less 'necessary' (for want of a better word) in regions of greater social stability, then 'Saxon' material culture would obviously show up less often.

This is an idea that I find very attractive but I am not a linguist or philologist.  I argued with Steve as long as I could before he convinced me.  Are there other crucial problems with this that I might have missed?  Or is it one to float or even push?  Thanks.


  1. The choice of a lingua franca can certainly be a matter of convenience more than affection or ideological commitment. For example when I have introduced friends, here in Oxford, from Palestine or Iraq to other friends from North Africa, they have peferred to converse not just to me but to each other in English rather than Arabic. This is because local versions of Arabic seem too different in pronunciation and slang to make chatting comfortable.
    It may be that there were always different very regional versions of Brythonic, or maybe the Roman occupation had killed of any high status Belgic version, so that English became convenient even for Britons talking to each other.
    David Hillman

  2. This is quite an intriguing idea. Posts like this make the wheels spin in my head, even if I have nothing much to add. It certainly reinvigorates my thoughts and understanding on the exciting possibilities of the late/post imperial period in Britain.

  3. The current thinking on the Old English - Celtic linguistic relationship moves towards postulating a diglossia in the Anglo_Saxon period; the surviving OE is a literary, elite dialect, thus relatively free of Celtic influence. On the other hand the spoken language in Anglo-Saxon England would be influenced by local Celtic varieties as their speakers shifted to Old English. In a language death situation you do not expect lexical borrowings appearing in the dominant language, but rather features of syntax - and these are identifiable in Middle English, which is a continuation of the spoken (and unattested) variety rather than written construct we call "Old English". The use of do, periphrastic constructions and a number of other syntactic features of Modern English are attributed to Celtic influence this way.

    As to the Latin influence, it is substantial, but mostly connected with (a) Christianisation of the Anglo-Saxon kingdoms in the 7th century, and (b) learned loanwords into the written (again) variety. The number of early post-settlement loanwords from Latin is rather low.

  4. I must say that this certainly sounds like an interesting and potentially viable idea to me, though I am no expert in early Anglo-Saxon history. I presume (not least from your mentioning him by name) that you're acquainted with Alex Woolf's fine 'Apartheid and Economics in Anglo-Saxon England', which I also find a an interesting (if inevitably tentative) attempt to solve the problem.


  5. Another possibility is the view advocated by geneticist Stephen Oppenheimer. In brief, the Belgae of Southern England were a Germanic people and their language was possibly mutually intelligible with that of the Anglo invaders. Modern English is descended from a blend of the two dialects.
    Here's a link to an early Prospect Magazine piece on the theory.

    more here

  6. The problem I see with Oppenheimer's hypothesis is that as presented in the article it simplifies things to a large extent and ignores a lot of relevant lignuistic evidence; maybe the book presents a more extensive argument?

  7. The book certainly goes into the argument in much more depth. Possibly still not enough depth to satisfy critics.


  8. There is a brand new article just out by David Parsons in the Transactions of the Philological Society 109 (2011), 113-137, which deals with the relationship between Latin and British in Roman and early post-Roman Britain. This is now a must read for this topic.

    For the argument Guy posts above I am not comfortable with the idea that elite languages become widespread vernaculars in this way though I ahve wondered whetehr the dominace of Romace in western Europe may not be partly because it was the lingua franca for bi-lingual Celts and bilingual germans (after all Celtic was still spomen in the Trier region when Jerome was there).
    As for the earliest English settlements not being in the East then that is fine. We ought perhaps be looking for them in the north (I favour the fifth-century phase at Traprain as proto-English).
    The argument cited above, largely depndent on the work of Poussa and Tristram that OE is British-free because it is an acrolect is problematic because it clearly emerged in the sventh and eighth century from regional dialects and does have distinctly insular peculiarities (some of which, morphosyntactically, may be derived from British or Romance.

    Alex W.

  9. Here is Geoffrey Sampson's take on Oppenheimer: http://www.grsampson.net/QOppenheimer.html

    The critique seems pretty devastating.


  10. I think I'd be cautious with that Jerome reference, Alex.

  11. Why in particular, Guy. Obviously all such anecdotes need to used cautiously but I have often wondered why people assume they know that Celtic disappeared so early from the Continet. My guess is it is under researched as a topic because of prevalent presumptions which have little basis in hard evidence.

  12. My own guess is that the linguistic map of the Roman Empire would have been quite a complex patchwork, sometimes at a very local level. Incidntally though Gregoy of Tours identifies soem words as Celtic he seems to be using the word in the Geographical sense since all the words he descibes thus are actually Romance. there is on the other hand a reasonably good evidence for Galatian still being spoken alongside Greek in Late Antiguity in Anatolia.

  13. I have been thinking about Trier overnight. Are you sure that the evidence you present does not point rather to a larger gap between the rich an the poor than elsewhere in Gaul? If this were the case then one might expect a larger cultural, and perhaps linguistic, gap between the rich and the poor also.

  14. 1. GT didn't think those words were 'Gaulish' the point is Celtic is a regional not a linguistic term in Latin of that period because part of Gaul was Celtica (as opposed to Belgica and Aquitainica).
    2. Since Jerome did speak Latin/Romance then what can the language he didn't understand in Trier have been but Celtic.
    3. I don't think Greek writers confused Gaul and Galatia - Galatia was the normal Greek term for Gaul (cf ptolemy).

  15. Your point 1: I'm not sure what point you're making. I need to go and have a look at GT (and C8th charters) again.
    Your point 2: You know that but did he? You aren't taking the source on its own terms. I repeat: all he is saying is that one language he didn't understand sounded (to him) like another language he didn't understand. Now, he doesn't say whether he understood either, but it seems unlikely that he did. If he did understand, then what I expect he was saying was that these demotics (of Latin and Greek) were similar - maybe that meant he could understand bits but they were heavily dialectal. We just don't know what he meant and we can't interpret it by projecting back onto him our (or more technically accurately, *your*) knowledge of linguistic taxonomies etc.
    Your point 3: Fair enough. But it it doesn't evade the point I'm making, that an educated late Roman ought to know that, historically, the Gauls and Galatians were related (at least according to the classical histories) - the empire was littered with statues of dying Gauls/Galatians - and that that might very well have coloured anything he might have observed.

  16. Galatas except sermon Graeco, quo omnis in Oriens loquitur, propriam linguam eamdem pene habere quam Treviros,

    This seems reasonably explicit and remember Jerome lived in Trier for nearly five years.

    Also don't confuse archaeologically observable Romanitas with Romance language.

  17. Your auto-correct has gone a bit mad there! But my point stands - unless you think that he could speak either (or both) Treverian or Galatian Celtic dialects, what does it amount to? One language I don't understand sounds a bit like another language I don't understand. Assuming that sermo means speech/language, rather than words/vocabulary. I know he lived in Trier - as a palatine official - and he travelled through Asia Minor where I suppose he heard Galatian. No. Sorry, Alex but I'm still not convinced that this will carry any weight. After all we don't even know that this was his own observation. It's probable but it needn't be. I dare say a bit of rooting through possible classical sources might be helpful, if only to rule out possibilities.
    Was I confusing archaeologically observable Romanitas with Romance? I'm not sure I was...

  18. Jerome goes on to compare the differences between Treveri and Galatian with the modification of Afro-Phoenician [i.e. Punic] and Latin over time.

  19. The idea that English first became the lingua franca in the border regions and then spread back East,filling in the gaps But what
    is the evidence that they were more stable and prosperous than the South East?
    It is usually argued that from Edwin's time on dominance moved to Northumbria, Mercia, and Wessex because they had more opportunity of expanding at the expense of the Britons, which begs several questions.
    I think the Chronicle actually claims that Aethelbert,usually understood as the later king, was driven back into Kent by Ceawlin.But before Ceawlin, whose reign was probably quite short, neither Bede nor the Chronicle name any over-kings except the enigmatic Aelle of Sussex.I what kind of polities you see as exercising dominance before this from the border regions.
    It could be that many of the Saxons in regions later called Wessex and Mercia arrived at different times and with little connection with the Saxons settled somewhere in the Eastern region by the proud tyrant.I see them in the fifth and sixth century as mixed Celtic Saxon regions,but of small polities, where alliances and emnities cut across ethnic divides, and where a kings fighting men included exiles,hostages, foster children, and young adventurers from many parts.
    Who in these regions was politically strong enough to dominate the rest of England?
    I'm not trying to argue against your scenario, just seeking clarification.
    By the way, thank you for your posts on current matters. I will always think of Starkey, from now on, as a former historian.
    David Hillman

  20. If we assume that the Britons who migrated to Brittany came from eastern and/or central areas of Britain (i.e. they moved south to avoid the changes associated with the Anglo-Saxons - current or impending) then we have to conclude that those areas of Britain included people who spoke Brythonic. The old-fashioned French nationalists who wanted Breton to derive from Gaulish have given up, and there seems little doubt that it derives from what some call Common neo-Brythonic. If we accept this, and we accept that the Anglo-Saxons were a significant part of what drove these people to cross the sea, then we have to allow room for Brythonic speakers in areas quite a long way East.

    Of course the chronology of Breton settlement (leaving aside Riothamus and his 50,000 aiding Anthemius), is obscure, as is the reason for the departure from Britain and where in Britain they came from, but it is an easier sell to associate this with the Anglo-Saxons than anything happening in the far west of Britain.

    As for Trier - 1000 latin inscriptions from the 4th-8th centuries often using strikingly post/non classical orthography to represent the new sounds being used in spoken latin, leaves little doubt (in my mind) as to what language was spoken in that particular part of Gaul.

    Mark Handley.

  21. I think that most stories tell of the Breton immigrants coming from South Wales and Cornwall. Saints Lifes and place names seem to imply this too.
    David Hillman

  22. I think linguistic analysis of Breton also points to a close relationship with Cornish (and so what we expect to be SW British -- though, of course, as it happens, all our useful examples of British are more western than not!). Still, if British survives more vibrantly in western Britain (or the highland regions) in the post-Roman period, then it might well have already been surviving there more vibrantly in the Roman period too.

    I suppose its difficult to say how different (or not) were the pre-Roman languages of, say, the Iceni and the Dumnonii, or when a putative "Eastern Brittonic" of the Iceni may have died out.

  23. As I recall Gregory of Tours makes a comment that he knows people still speaking Celtic and others speak Frankish and others Latin.

    As for Jerome, he was at least bilingual, knowing Greek and Latin. As you all know, he would later go on to learn Hebrew and Aramaic. It seems to me that it goes against type for Jerome to have spent 5 years in Trier and if some of the population was speaking a Celtic language not to take some note of it. Not evidence, I know, but nonetheless I think we need more than a minimalist rejection of the statement.

    One of the things I'm always surprised at is how seldom other situations where various Indo-Europeans come into an area, maintain their language, and borrow very little from the native linguistic stock. Or, as in the case of the Normans, adopt a new language and very little of their native Norse is borrowed into Anglo-Norman French. Further, are not the majority of Latin loans from the pre-Christian era names of trade-goods? And Celtic loans largely of geographic features? Both are indicative of something, I just don't know what.


Note: only a member of this blog may post a comment.