THE LAST ROMANIZED BRITONS

There are many books and essays about Sub-Roman Britain and the Romanized Britons. One of the best in my opinion is related to King Arthur and was written by the famous John Morris in 1973. His "The Age of Arthur" was the first attempt by a professional historian to build a picture of what used to be Roman Britain during the period 400–650 AD, when King Arthur (whom Morris accepts as an authentic historical personage) was supposed to have lived. The book is not, however, exclusively about Arthur, but rather about the history of Romano-Celtic Britain during that era. 

For Morris nearly 4 million inhabitants -nearly all of them fully Romanized- were living in Roman Britannia south of the Hadrian Wall when the Romans withdrew to the continent in 410 AD, but one century later the Britain population was reduced to just two millions (including nearly 200000 german invaders). However the remaining Romanized Britons were able to defeat around the year 500 AD these Anglo-Saxon enemies, obtaining the control of most of Britannia for half a century..... and only after 577 AD were definitively defeated by the Teutonic invaders. 

Ken Dark ("Britain and the End of the Roman Empire", 2000) uses archaeological evidence to demonstrate that the Romanized Britons were alive and well inside 'Anglo-Saxon' territory for several hundred years at least (until the start  of the ninth century!).

Indeed western Britain has attracted many archaeologists who wish to place King Arthur as a historical figure. Though there is little contemporary written evidence for this, archaeological evidence does suggest that a Romano-British ruler or leader (like Ambrosius Aurelianus) might have wielded considerable power during this initial sub-Roman period, as demonstrated by the creation of sites such as Tintagel and earthworks such as the Wansdyke. Such interpretations continue to attract the popular imagination and the researches of academics. 

Map of Sub-roman Britain in 500 AD showing the Romanized Britons, with the name "Welsh", in the area in pink

Indeed it is demonstrated that this British leader Ambrosius Aurelianus (after the withdrawal in 410 AD of the Roman legions from 'Britannia') fought against the Anglo-Saxon barbarians in a number of battles apparently over a long period. Towards the end of this period was the Battle of Mons Badonicus (read also:http://darkagehistory.blogspot.com/2014/12/the-campaigns-of-arthur-dux-bellorum-in.html), around 495 AD, which later sources claimed was won by King Arthur, though the historian Gildas (the only contemporary source we have) does not identify him. After this there was a long period of peace. The Romanized Britons seem to have been in full control of what is now England and Wales roughly west of a line from the area of York to the shores east of Bournemouth. The Saxons had control of eastern areas in an arc from East Yorkshire through Lincolnshire and perhaps Nottinghamshire, to East Anglia and South East England. 

What ended this period, when the Romanized Britons were prevailing in former Roman Britain over the invaders from German-Danish shores? Probably it was the "Constantine plague" that decimated the Romanized Britons, who were in huge commercial contact with the Mediterranean region devastated by the plague (while this plague seems to have spared the Anglo-Saxons in their part of the British islands, because they were isolated).  

Writing in Latin, perhaps about 540 AD, Gildas gives an account of the history of Britain, but the earlier part (for which other sources are available) is severely muddled. He castigated five rulers in western Britain – Constantine of Dumnonia, Aurelius Caninus, Vortipor of the Demetae, Cuneglasus and Maglocunus  – for their sins. He also attacked the British clergy. He gives information on the British diet, dress and entertainment. He writes that Britons were killed, emigrated or were enslaved but gives no idea of numbers: but his work named "De excidio Britanniae" clearly indicated that Roman Britain was "murdered" with a terrible bloodbath ('excidio' means in latin: 'kill').

In the late 6th century there was another period of Saxon expansion (after the years of the Constantine plague), starting with the capture of Searoburh in 552 AD by the dynasty that later ruled Wessex, and including entry into the Cotswolds area after the Battle of Deorham (577 AD), though the accuracy of the entries in the Anglo-Saxon Chronicle for this period has been questioned. These conquests are often said by modern writers, on no clear evidence, to have separated the Britons of South West England (known later as the West Welsh) from those of Wales. (Just after the period being discussed, the Battle of Chester in 611 AD might have separated the latter from those of the north of England.)

Until the 570s, the Britons (more or less romanized or partially romanized) were still in control of about half of England and Wales. But at the beginning of the seventh century the Romano-British word started to disappear completely.

Additionally we must remember that the Plague of Justinian (that killed as many as 100 million people across the world: as a result, Europe's population fell by around 50% between 540 and 600 AD !) entered the Mediterranean world in the 6th century and first arrived in the British Isles in 544 or 545 AD. Just before the battle of Dyrham in 577 AD, that was the beginning of the final conquest of Subroman Britannia by the Anglosaxons: the important Romano-britons city of Calleva was abandoned in those years, because hard hit by this terrible plague.

Richard Lehman wrote that "....in 550 AD, the island of Britain was predominantly Romano-British: they were unable to maintain a full urban civilisation after the departure of the Romans in 410 AD, but were successful at keeping the Angles and Saxons confined to Anglia and Kent (after the battle of Badon Hill). There was no trade or social exchange between the Christian British and the pagan Angles and Saxons, once they had had fought each other to a standstill under King Arthur. The British carried on some trade with the Mediterranean, whereas the English lived on what they could grow. So when the plague reached Britain in boats from mainland Europe, it killed up to half of the native Romano-British population but left the English colonists largely unscathed. Not long afterwards, the English began to mount probing raids into British territory and found that there was little opposition. They sent word back to their relatives in Schleswig-Holstein and the Danish peninsula that the whole island was up for grabs. The king of the Angles was so impressed that he put his entire population into boats and left the area west of Hamburg deserted for several centuries. And so, 150 years after Hengest and Horsa first brought in Saxon warriors to police the borders of crumbling Roman Britain, the English decisively colonised plague-ravaged Britain from the borders of Wales to the middle of Scotland...…"

Map of southern Sub Roman Britain in 575 AD, just before the "Battle of Dyrham", showing the areas of Romano-Britons, Saxon & Jute settlements according to the historical sources (Bede)

Furthermore, scholars such as Christopher Snyder (read http://www.the-orb.arlima.net/encyclop/early/origins/rom_celt/romessay.html) believe that during the 5th and 6th centuries – approximately from 410 AD when Roman legions withdrew, to 597 AD when St. Augustine of Canterbury arrived – southern Britain preserved a sub-Roman society that was able to survive -for a while- the attacks from the Anglo-Saxons and even use a vernacular Latin (called "British Latin" or "Insular Romance") for an active culture. There is even the probability that this vernacular Latin lasted to the late 7th century (or possibly later) in the area of Chester and Gloucester, where amphorae and archaeological remnants of a local Romano-British culture (mainly in the locality called 'Deva Victrix') have been found (read http://www.british-history.ac.uk/report.aspx?compid=19184#n4 about the discovery that Amphorae of 616 AD were found in Sub-Roman Chester). This Roman city is thought to have lasted until 650 AD, and probably some Roman cultural presence in a small romanized population remained until the beginning of the 8th century (and probably also later) (http://www.chesterwiki.com/Dark_Age_Chester). Indeed in the early 8th century [[Bede]] called Chester a city ("civitas") and clearly knew of it as a Roman place (in 'Early medieval Chester 400-1230', A History of the County of Chester: Volume 5 part 1: The City of Chester: General History and Topography  (2003), pp. 16-33).

Indeed -according to H. R. Loyn (in his "Anglo-Saxon England and the Norman Conquest". Harlow: Longman; p. iii; 1962)- as late as the eighth century the Saxon inhabitants of St. Albans (an important city nearly 70 km west of Camulodunum) were aware of their ancient neighbors of the Roman city called 'Verulamium', which they knew as "Verulamacæstir" (the fortress of "Verulama"), possibly a pocket of Romano-British speakers remaining separate in an increasingly Saxonised area.

Scholars have seen signs of continuity between many "late" Roman towns and their medieval successors. Urban continuity has been confirmed for Bath, Canterbury, Chester, Chichester, Cirencester, Exeter, Gloucester, Lincoln, London, Winchester, Worcester, and York. At Verulamium (St. Albans), where the medieval town grew up around the Saxon abbey outside of the Roman walls, archaeologists found several late fifth-century structures and a newly-laid waterpipe indicating that a nearby Roman aqueduct was still providing for the town's sub-Roman inhabitants in the sixth century. And at Silchester, which did not become a medieval town, excavations revealed that economic activity at the forum continued into the fifth century (dated by coins and imported pottery and glass), while jewelry and an ogam inscribed stone hint to late sixth century contacts with Irish settlers (read for further information: http://www.vortigernstudies.org.uk/artgue/snyder.htm).

Indeed the British economy did not collapse during the early Sub Roman period. Pottery - "roman-style"- was for sure produced in the fifth and sixth centuries (read https://quod.lib.umich.edu/f/frag/9772151.0005.001/--perils-of-periodization-roman-ceramics-in-britain-after-400?rgn=main;view=fulltext;q1=robin+fleming). 

Although no new coinage was issued in Britain, coins stayed in circulation for at least a century (though they were ultimately debased); at the same time, barter became more common, and a mixture of the two characterized 5th (and early 6th) century trade. Tin mining appears to have continued through the post-Roman era, possibly with little or no interruption. Salt production also continued for some time, as did metal-working, leather-working, weaving, and the production of jewelry. Luxury goods were even imported from the continent -- an activity that actually increased in the late fifth century. 

Furthermore we must remember that after the collapse of Roman rule, coin minting and the use of money appears to have ceased in Britannia for some two centuries. Britain reverted to barter and a largely rural, moneyless economy. Glyn Davies noted in 1996 that "after the fall of Rome Britain showed the unique spectacle of being the only former Roman province to withdraw completely from using coined money for nearly 200 years...the absence of money reflected and intensified the breakdown of civilized living and trading." 

Southern Britannia map showing the possible place of mount Badon, north-west of the isle of Wight (Vectis in latin). Note that there are indicated the places of battles in the the fifth & sixth century.

However, in 1997, a hoard of 22 gold "Solidi", 25 silver coins or fragments of silver coins, 2 heavy gold rings and 50 small pieces of silver bullion dating from 333 AD until 470 AD was found at Patching, near Worthing, Sussex. The coins included two imperial coins from Ravenna (reign of Valentinian III) dated c.  440 AD and Visigothic coins from the reigns of Majorian (c. 460 AD) and Libius Severus (c. 461 AD). The hoard was buried, possibly in advance of Saxon incursions, around 475 AD. This is evidence to show that Roman coins were still reaching Britain well into the fifth century. The older coins were heavily clipped, suggesting their use in circulation whereas the later coins were almost pristine. The find also implies that Romano-British institutions of a sort were still in existence in what is now Southern England late in the fifth century (and probably in the early sixth century).

Only after the mid sixth century started a deep crisis for the Roman Britons: this fact coincided with the decades of the terrible "Constantine plague" (that reached Britain around 545 AD).

The R1b-U152 presence in Roman Britannia

Furthermore, we have to remember that the Romans established over 70 cities or towns in Britain, including important cities like London, St Albans, Colchester, Winchester, Gloucester, Exeter, Leicester, Lincoln, Manchester and York. But they "genetically" left only a minimal presence in the actual population of these cities. Only the area around Lincoln and York has maintained a relatively huge amount of people with the R1b-U152, the so called "Haplogroup of ancient Romans" (see the following related map). 

Indeed Maciamo Hay wrote in this interesting essay (read the complete article here:https://www.eupedia.com/genetics/britain_ireland_dna.shtml) that "genetically and cromosomically"....

"it is very difficult to assess the genetic impact of Romanization on the British population as the Roman citizens, soldiers and slaves who settled in Britannia were not merely people from the Latium or Italy, but could have come from anywhere in the empire. 

Even if we assume that Britain was fully Celtic before the Roman conquest, similar to Ireland or the Scottish Highland once Germanic DNA has been removed, it is still very hard at present to clearly differentiate Brythonic Celts from other Celts from the continent, notably Gaul, who might have settled in Roman Britain. Even the Romans from Italy appear to have belonged predominantly to the same R1b-U152 as Hallstatt and La Tène Celts, also accompanied by significant minorities of G2a-U1 and J2b. Deeper subclade analysis may soon allow population geneticists to distinguish between Roman/Italic and Celtic subclades within these haplogroups. At present it seems that the L2 and Z36 subclades of R1b-U152 are more Celtic/Gaulish, while Z56 and Z192 are more Italic/Roman.

Map of the R1b-U152 in western Europe 

So, experts think that the Romans are in a maximum roughly 15% of male lineages of "Roman" origin for England, 10% for Wales, and 7% for Scotland. However, it would be reasonable to assume that at least half of these come from Alpine Celts and Normans, and probably more in Scotland's case. It is hard to explain the discrepancy with the 30-35% of autosomal genes of Mediterranean origin that are now present in the modern population of England. One explanation is that a substantial share of Romano-British male population was killed by the invading Anglo-Saxons, and that autosomal genes were passed on through Roman-British mothers who bore the children of the Germanic invaders....."

Some researchers (like D'Ambrosio of the University of Genova) wonder why the farm area around Lincoln (read also https://www.reading.ac.uk/web/files/archaeology/Reading_April_14_talk_AS.pdf )  is the only one in modern Great Britain with actual population showing genetically a 15/20% of R1b-U152, the so called "haplogroup of Ancient Romans" (a percentage that is similar to the one in Mediterranean Spain around Valencia!). One possible explanation is that the Roman villas population of the area survived the onslaught remembered by Gildas in his "De excidio Britanniae" getting refuge in the nearby southern Pennines  mountains and later-when better times arrived during the last Sub-Roman Britain decades- these Romanized populations came back to where they used to live.

Probably this is what happened near Lincoln in the small city of Southwell, that was called "Tiovulginacester" in Sub-roman times, where there also are huge percentages of the Haplogroup J2 (typically originated in Anatolia & southern Italy).

Additionally we must pinpoint that in the city of Southwell (near Lincoln) the remains of an opulent Roman villa were excavated in 1959. Part of a mural from the excavation is displayed in the Minster. It is one of three of its type found in the territories of the Corieltauvi (or Coritani) tribes – along with Scampton in Lincolnshire and Norfolk Street in Leicestershire. A stretch of the Fosse Way runs on the far bank of the River Trent, with evidence of Roman settlement at Ad Pontem ("to the bridge" or "at the bridge"), north-west of the village of East Stoke. Other evidence of Roman settlement includes the use of Roman bricks in prebendary buildings around the Minster, remains of a ditch or fosse discovered at Burgage Hill in the 19th century, and possible Roman remains beneath the Church Street site of the recently vacated Minster School...The Venerable Bede records a multiple baptism in the "flood of the Trent" near "Tiovulginacester" by Paulinus in the presence of Edwin of Northumbria, whom he had converted to Christianity in 627. There is disagreement on the location of Tiovulginacester, but Paulinus certainly visited it and may have founded the first church in Southwell: the existence of this place called Tiovulginacester clearly shows that a partially romanised population was still living there in the early seventh century...BD

Cultural division in "sub-Roman" Britain

In late Roman Britain there is an oft-noted, though somewhat simplistic dichotomy drawn between a "civil", "lowland" zone and a "military", "highland" zone, the latterly largely in the North and West and the former in the South and East. Its continuity was seen into the post-Roman period after 410 AD (read also: https://www.gutenberg.org/files/33059/33059-h/33059-h.htm ).

This "highland" and "lowland" zone division is, however, recognisable in the post-Roman period, even within the British controlled parts of the island. About two hundred and forty-two inscribed stones survive from western Britain which have been dated, on palaeographic grounds, to between AD 400 and 700 (or, less cautiously, between 450-650). Twenty-eight of these are bilingual inscriptions in Primitive Irish and Latin and the rest simply in Latin. These monuments are, perhaps, the most characteristic landscape feature surviving from Late Antiquity in this region. 

The period during which these inscribed stones were being erected coincides fairly exactly with the period in which so called "pagan" AngloSaxon burials of various sorts were taking place in the east of the country.  If one plots the distribution of inscribed stones against the distribution of "pagan" Anglo-Saxon burials one is left with a broad band of territory in which neither class of site is found, running north from the stretch of coast between Dartmoor and the Hampshire Avon, comprising east and north Devon, most of Dorset and Somerset, most of Gwent and a large part of Glamorgan, western Gloucestershire, Herefordshire, much of Powys, Shropshire, most of Worcestershire, Cheshire and most of Staffordshire and Derbyshire as well as most of the north of England outside of East Yorkshire. 

In addition to this broad band one might, tentatively, add the area immediately north and north-east of London. The extreme south of Scotland contains a few inscribed stones with one or two outliers in the far north of England. There are of course a few stones and burials, such as the Maiden Castle warrior burial, which lie outside the main distribution.

While a large part of what had been the "lowland" or "civil" zone seems to have fallen under the political control of Germanic-speaking groups with their distinctive burial practices, the boundary between those parts of this zone which remained in British hands and the "highland" or "military" zone seems to have been marked by the transition between the regions in which the inscribed stone tradition took root and that in which it did not.  Interestingly, this division seems to be reflected in our one literary source for the period, the "De Excidio Britanniae" of Gildas. De Excidio takes the form of a letter addressed by Gildas to the leaders of British society. This letter attributes the decline of British national fortunes to its recipients poor leadership and sinful personal lives. It almost certainly dates from the mid-sixth century.

Gildas’ lament is addressed to the secular and ecclesiastical leaders of the Britons, but the only individuals amongst his audience whom he names are five kings. The kings named are, in order, Constantinus, Aurelius Caninus, Vortiporix, Cuneglasus and Maglocunus

The identification of Gildas’ kings has been to make the point that the rulers he addresses occupy those territories which are characterised archaeologically by the presence of Class I inscribed stones. Within southern Britain the only region with a concentration of stones whose ruler he does not appear to address is the tiny land-locked realm of Brycheiniog. 

At the same time Gildas addresses by name no rulers of any of the other regions which seem to have remained under British control during the sixth century but which do not have significant concentrations of Class I stones; that is to say, central and south-eastern Wales and western England. Apart from concentrations of stones around Whithorn and Kirkmadrine, which may bear witness to direct contacts with Gaul and stand somewhat outside the Class I tradition, there are very few Class I stones in the North. 

The most northerly Class I stone is the Catstane at Edinburgh Airport and the most southerly at Chesterholm (Vindolanda) on the Stanegate, but all the others lie between the Lammermuir/Lowther watershed and the line of Hadrian’s Wall. Since Vindolanda was part of the Wall complex it is probably safe to conclude that here too the distribution of Class I stones coincides with the "military" zone. One point worth noting is the absence of any Class I stones from Lancashire, the Pennines or the Lake District, probably all regions controlled by Britons into the seventh century and traditionally viewed by Romanists as part of the "military" or "highland" zone. 

The lowland British have proved notoriously difficult to identify in the archaeological record. The main problem in identifying and quantifying their material remains is that while the bulk of the population seems, in terms of agrarian and domestic practice, to have continued to enjoy fourth-century life-styles, coinage and mass-produced ceramics, the cornerstones of Romano-British chronology, ceased to circulate in Britain, or at least to be produced. The lowland British are best known from cemetery evidence. The dominant mortuary rite continued late Roman practice and comprised extended inhumation accompanied by few or no grave goods. There is little evidence for social stratification within these cemeteries even though some of them are very large; Cannington, in Somerset, for example is thought to contain some two thousand burials, only a quarter of which have been excavated, dating from the second to the eighth century. Further east similar cemeteries are also known, though they rarely seem to have stayed in use so late. At Queensford Farm, near Dorchester-on-Thames, a cemetery, also containing about two thousand graves, was in use from the late fourth to, at least, the mid-sixth centuries, and perhaps even into the late seventh. 

Roman coins findings clearly indicate the areas of biggest "romanization" and presence in Roman Britannia

The other major archaeological phenomenon associated with the lowland British is the re-occupation of Iron Age hill-forts; most famously South Cadbury and Cadbury Congresbury. Many of these sites had remained in some kind of use through the Roman period, but this was mainly of a ritual nature. In the fifth and sixth centuries refortification took place and domestic occupation was re-established. At the larger sites, including the two just mentioned, imported Mediterranean pottery has been recovered. 

The re-emergence of hill-fort settlement coincides with the disappearance of evidence for occupation on villa sites and it is fairly safe to assume that the former replaced the latter as centres of elite residence, albeit for a more restricted elite. It is not clear, however, that the adoption of such sites should be seen as a conscious militarization of the elite rather than as part of a simple desire for greater personal security and the appropriation of dominant places within the landscape. One should certainly be cautious of crying "continuity" from the Iron Age; reoccupation after several hundred years of abandonment as residence sites may reflect conscious archaism but this is not the same as continuity. Whilst the villa, an Italianate country house, symbolised its owners’ links to the affluent, yet remote, society in which the god-like emperor and the imperial court existed, the hill-fort served as a cruder reminder of exactly what the sources of social power were.  

What can be said with some conviction is that urban life certainly came to an end at some point in the fifth century: although some occupation of town sites may have continued this occupation was not urban in character. At Wroxeter and Verulamium large timber structures do seem to have been built in the mid- to later-fifth century, but these seem to have been the dwellings of high status individuals, perhaps using the enclosed urban area much as some of their contemporaries used the ancient hill-fort ramparts. With the disappearance of towns and coinage, craft specialisation and mass production also came to an end. 

On the other hand, the "highland" zone was not entirely devoid of Romanizing, or Romancing, traits. The inscribed stones do, after all, almost all bear inscriptions in Latin. Some also bear inscriptions in Irish but only one, at Tywyn, and that very late in the sequence, bears an inscription in British (or rather Old Welsh by this stage). “Moreover,” as Thomas Charles-Edwards writes, “the character of the Latin used in the inscriptions shows that it was a spoken language, not merely a language of the quill and chisel.”Charles Edwards goes on to argue that the epigraphic evidence can be used to show how long Latin, or rather Romance, was “used in a wide variety of styles and registers”.

A further example of Romance interference in the epigraphers’ Latin is the indiscriminate use of second declension genitive singular ending, without regard to syntax. Professor Charles-Edwards goes on to contrast these kinds of flaws in the Latin of the Class I stones with the types of mistakes made by medieval authors whose native tongue was not Romance and who learned their Latin from grammar-books, pointing out that while they may make mistakes in their use of Latin cases, “many of those responsible for the texts of the inscriptions were unaware of any case system at all. They had not learnt their Latin from grammar” he goes on, and “Latin -or better: "British Latin", called also "Insular Romance"- was, therefore, in the time of Voteporix [the mid-sixth century], a spoken language alongside Welsh and most of the population of western Sub-Britannia”.

A much more widely discussed area of evidence for the existence of Insular Romance lies in the relationship observable between Latin, on the one hand, and Irish and Welsh on the other. The peculiar orthography adopted to write these two languages, probably in the sixth or seventh centuries, almost certainly reflects the way written Latin was pronounced in Britain in this period.

Formal educated Latin, at least as a written standard, existed alongside Insular Romance and this implies that the Romance speaking community was relatively large, even in the West. We are not to imagine that Latin usage was an affectation of a tiny number of kings and clerics but that Insular Romance was the normal language of intercourse for a significant proportion of the population. Perhaps we should imagine a linguistic divide similar to that apparent in twelfth-century England with Insular Romance playing the part of Old French. 

Welsh Ethnogenesis & disappearance of Insular Romance
 

Alex Woolf wrote that when considering the establishment of barbarian kingdoms on the Continent one finds oneself musing on the survival of Germanic language and the rate at which it was replaced by Romance as the medium for elite discourse in the Western provinces. 

In Britain, paradoxically, the decline of Romance and the adoption of British Celtic by the elites is the parallel phenomena. In the present context we might also consider that this transition is in some way related to the establishment of a recognisably British gens separate from Gens Romanorum. Thomas Charles-Edwards evades this thorny problem, pointing out merely that by the ninth century the neo-Brittonic languages (principally Welsh) were certainly the languages of the elites in the British World.

image of the Deva castrum (actual Chester), showing some houses of the civilian settlement outside the walls that survived during Sub-roman Britain era (and where was spoken the "insular romance")

Evidence for the exact timing of the death of Insular Romance has not, so far, been identified, but this need not mean that this evidence does not exist. One place in which we might look for this evidence is in the literary and linguistic forms that supplanted the Latin and Insular Romance traditions, and the most likely location of those traditions is in early Welsh verse. 

the switch from Insular Romance as the preferred language of a Romanizing elite to the language of the country, Cymraeg, will have taken place in the period between the emergence of the earliest vernacular praise poetry and its adoption throughout the British-speaking world. In absolute terms we should probably think of a   transition starting in the North c. 550, taking root in Wales before the middle of the seventh century and reaching Dumnonia by the early eighth at the very latest.

Seen in this chronological context, the impetus for the abandonment of Insular Romance in the West can be seen to coincide with the conquest of the majority of the eastern, lowland, Britons by the Anglo-Saxons in the same time-period (c. 550-700). In these terms the long survival of Insular Romance in the West, for more than two hundred years after the disappearance of effective Roman power and centralised government, can be re-assessed. 

The western and northern kingdoms maintained their Romanizing character so long as they lay adjacent to a large contiguous zone in which Romance culture throve. To some extent, the kings and their courts must have seen themselves as peripheral to and dependent upon, in cultural terms at least, the lowlands. This should not surprise us. Despite the absence, as yet, of a rich material culture emanating from amongst the lowland British, their land was far more productive and the density of affluent, educated aristocrats must have been much higher. Watching this world collapse before them, and finding the new rulers of the lowlands uninspiring, the kings of the West must have turned towards their own countrymen and to the language of local people for cultural comfort. This reorganisation of identity will not have happened overnight. At the time it may well have seemed almost imperceptible. 

For the most part people capable of speaking both languages well will have simply begun to switch their preference from the one to the other. After a generation or two children would have grown up without sufficient exposure to Insular Romance to perpetuate it. An interesting question to consider is whether this switch had any effect on the character of the Welsh language. 

Kenneth Jackson, in 1953, thought that most of the borrowings from Latin into British (between 700 and 1200 items) had occurred prior to AD 400. Subsequent scholarship in his own and adjacent fields has, however, shown that he was imposing needless constraints upon the data. Recognition that British Latin (Insular Romance) developed, at least in part, in tandem with British Celtic, has at once freed us from such constraints and left us somewhat at a loss for chronological markers in the development of either British or Insular Romance historical phonology. Going beyond lexical borrowings, David Greene noted “a number of vague parallels between Vulgar Latin [Romance] and British (but not, significantly, Irish) [...] and there may be some connection [...]. If so, it would be an 'areal' development, not a matter of Vulgar Latin loan words”. Unfortunately no systematic research has been carried out in this area.

Turning to the archaeological record it is interesting to note that the switch from Insular Romance to Welsh as the language of the elite coincides with the falling off of imported ceramics from the Mediterranean and western Gaul. Ewen Campbell has suggested that it is the ending of this exchange which causes various transformations in the archaeological record of western Britain, such as the abandonment of central fortified sites in Cornwall, and perhaps Wales, and widespread changes in ecclesiastical art and architecture.An alternative scenario might locate agency with the Britons, since in the Gaelic World these overseas contacts continued and even expanded, suggesting that there was no problem with supply. If supply was not in issue then it is perhaps demand that we should consider. The adoption of a vernacular identity may also, as noted when discussing the Llandaff material, represent a shift in ideology. Distinctive sociolects, and even more so language differentiation on the basis of class, bespeaks a highly stratified society in which the aristocracy are primarily exploitative. 

This is probably one reason why the bulk of the lowland British population are so hard to identify in the archaeological record; their landlords did not leave them with much to leave us. The social transformation in the West, c. 550-650 AD, can be seen as the replacement of an exploitative consuming aristocracy, which had developed under the aegis of the Empire, by a series of localised, kin-based, redistributive chieftaincies, much like those which were simultaneously transforming themselves into stratified societies in Anglo-Saxon England. 

Indeed Bede in ''Historia ecclesiastica gentis Anglorum'' (completed in 731 AD) wrote that "currently, [there are in Britain] the languages of five peoples, namely that of the Angles (English), the Britons (Welsh), the Scots (Gaelic), the Picts (Celtic) and the "Latins" (or Romano-Britons; in ''Historia Ecclesiastica'' 1.1: ''in praesenti ... quinque gentium linguis, ... Anglorum uidelicet, Brettonum, Scottorum, Pictorum et Latinorum''). This historical declaration by Bede is considered a proof -according to Di Martino and other historians- that at the beginning of the eight century there was a group of Romano-Britons who were still speaking their "insular romance" (https://books.google.com/books?id=zHZ4BwAAQBAJ&pg=PT552&lpg=PT552&dq=quinque+gentium+linguis,Bede+...+Anglorum+uidelicet,+Brettonum,+Scottorum,+Pictorum+et+Latinorum&source=bl&ots=jAH-vAamaA&sig=ACfU3U16kOkM9pX1dDbKAfkqH49ReK_zWQ&hl=en&sa=X&ved=2ahUKEwiwocDkqcPqAhXBl-AKHQ2IARQQ6AEwAHoECAcQAQ#v=onepage&q=quinque%20gentium%20linguis%2CBede%20...%20Anglorum%20uidelicet%2C%20Brettonum%2C%20Scottorum%2C%20Pictorum%20et%20Latinorum&f=false)

The last "testimonies" of Sub-Roman Britain


Last but not least, I want to remember that sixty miles west of the Wight island -on the coast surroundings of modern Dorchester- there was the sub-roman settlement of "Durnovaria". This area remained in Romano-British hands until the end of the 7th century and there was continuity of use of the Roman cemetery at nearby Poundbury until the end of the next century. Dorchester has been suggested as the centre of the sub-kingdom of "Dumnonia" or other regional power base, that had some commerce with continental Europe.

Tintagel castle in Dumnonia is worldwide known as a possible link to the famous "King Arthur": in 1998, the "Artognou stone" was discovered on the island, demonstrating that Latin literacy survived in this region after the collapse of Roman Britain.

In 1998, this "Artognou stone", a slate stone bearing an incised inscription in a "modified" Latin, was discovered on the Tintagel island, demonstrating that Latin literacy survived in this western region during the Sub-Roman years and that probably the Romano-Britons of the region used a romance language (the "British Latin": read for further information https://web.archive.org/web/20140821232929/http://www.mun.ca/mst/heroicage/issues/1/hati.htm) for some centuries after the Roman legions departure.

The Vortiporius dedication, a late 6th-century stone inscription in "Insular Latin" found in West Wales in 1895. According to Thomas Charles-Edwards, the inscription provides "decisive evidence" of how long Vulgar Latin was spoken in this part of the western Britannia (note that in Latin the word "protictoris" is written "protectoris", with an "e")

Furthermore, in a village near Durnovaria archaeologists have found evidences of a limited Romano-Britons presence until the second half of the eight century: the oldest "testimony" of Sub-Roman Britain!. 

Indeed the last eighty years of excavations has shed new light on both the sub-Roman phases of many Roman towns and on rural fortified settlements--"hillforts"--and monasteries. If not spectacular in their artifactual assemblage, these excavations have revealed impressive structures and the insight that sub-Roman Britain was anything but isolated and culturally impoverished.

The first, and perhaps most important, discovery came in the 1930s with Ralegh Radford's excavations at Tintagel, Cornwall (Radford 1939). Beyond the inner ward of Tintagel's Norman castle, Radford uncovered the remains of several small rectangular structures made of stone and slate as well as thousands of sherds of imported Mediterranean pottery, then termed "Tintagel ware." Much of the pottery came from wine and oil containers datable to the fifth to seventh centuries, leading Radford to interpret this settlement as a Celtic monastery. Subsequent excavations at Tintagel have revealed more structures and pottery, though alternative interpretations--a princely stronghold, an active trading post--have recently overshadowed the monastic model.

Tintagel's impressive commercial activity showed that Britain was not isolated in the sub- Roman period. On the contrary, Britain seems to have opened up new trade relations with Gaul, North Africa, and the eastern Mediterranean in which British commodities (most likely tin and slaves) were exchanged for luxury goods. "Tintagel ware" soon began to be identified from pottery finds at other sites, and new excavations turned up more examples along with imported glass and jewelry. The defended hilltop settlement at Dinas Powys, near Cardiff in Glamorganshire, yielded an abundance of these imports, even though its occupation area is quite small compared to Tintagel and the sub-Roman hillforts. Leslie Alcock's excavation's at Dinas Powys in the 1950s also revealed evidence of a thriving native metalworking industry, perhaps controled by local rulers who exchanged goods for military services.

Native hillforts were not the only form of defense for the sub-Roman Britons. Recent excavation along Hadrian's Wall has revealed much fifth- and sixth-century evidence, including new timber structures built at the Wall forts. At Birdoswald, for example, timber "halls" replaced two Roman granaries, while at South Shields a new gateway was constructed. Once thought to have been abandoned after 410, it now appears that Hadrian's Wall continued to be used to defend local civilian populations in the sub-Roman period. This may have been the case as well for the Saxon Shore forts in the southeast. Both Portchester and Richborough show signs of lingering occupation until the eight century, while the walls of the latter may have sheltered a Christian church.

Finally I want to pinpoint the existence of some isolated villages called "vicus" -in central and north Britannia- where Romano-Britons (because isolated) maintained their identity for some centuries (Sub-Roman Britain lasted nearly 4 centuries, from 410 AD to approximately the second half of the 700s), even if totally surrounded by the Anglo-Saxons and the celtic populations: for example, just south of the Hadrian Wall there were a few "vicus" near Piercebridge Roman fort that possibly lasted until the 700/750 AD (read http://www.yorkshireguides.com/piercebridge_roman_fort.html) with a Roman bathhouse, that was still in use around the year 800 AD.

Another was the civilian settlement of the "Deva Victrix" castrum (in actual "Chester"). 

The buildings of the canabae legionis of Deva (called also "Castra Legionis") were originally timber, but during the early 2nd century began to be rebuilt in stone. The settlement expanded throughout the 2nd and 3rd centuries as the population increased. Settlement extended around the fortress to the east, south, and west. Once the legion had left, the civilian settlement continued, eventually becoming part of the town of Chester during the early Middle Ages. 

Many buildings would have fallen into disrepair, although some of the larger structures are known to have survived for many decades after 410 AD. The town nevertheless probably remained the military and administrative centre -partially Romanized- of the region in Sub-roman Britain. Some amphorae from the Mediterranean region were discovered in this civilian settlement, showing the existence of commerce with the Roman world until the sixth century at least.  King Arthur -according to Bede- is said to have fought his ninth battle against the Saxon invasion at the "City of the legions" and later St Augustine came to the city to try and subjugate the Romano-british bishops to his mission. 

In 616 AD, Æthelfrith of Northumbria defeated a Romano-British army at the Battle of Chester and probably established the Anglo-Saxon position in the area from then on. Archaeological excavations at Heronbridge, just south of modern Chester, in 2004 uncovered post-Roman graves buried beneath a defensive earthwork over an old Roman settlement. There is evidence that they contain the bodies of casualties from the Battle of Chester (read for further information:  https://www.webcitation.org/67hhMeYza?url=http://www.chesterarchaeolsoc.org.uk/heronbridge.html).

After the arrival of the Anglo-Saxons in the early seventh century, the settlement became known as Legacaestir, meaning "City of the Legions" in Old English. Some archaeological evidences suggest that the "insular romance" (or what remained of this neo-latin language) was probably spoken in these civilian settlements until the early ninth century.

ADDENDUM

Roman mosaics discovered in 2020 near Bath at the actual Chedworth clearly pinpoint that Subroman Britain civilization survived into the fifth and sixth century (read https://www.ancient-origins.net/news-history-archaeology/chedworth-roman-villa-0014654)