Fouracre P. The New Cambridge Medieval History, Vol. 1: c. 500-c. 700

Подождите немного. Документ загружается.

Cambridge Histories Online © Cambridge University Press, 2008

The Celtic kingdoms 243

sea to Man and south-west Scotland and the battles for the Ulaid overkingship

became immensely complicated by the fact of D

al Riata interests (and power)

on the British mainland. Ainmere’s son Aed opportunely supported Aed

mac Gabr

al Riata at the so-called ‘convention’ of Druim Cett against

ain of D

the then Ulaid overking, probably B

an of the D

aet

al Fiatach;

the outcome

was an agreement that the Scottish D

al Riata should not owe tribute to the

Ulaid overking. Battles for the Ulaid overkingship continued on Man in the

570s and 580s, and in Ireland in the 620s, until 637 and Magh Roth; here Aed’s

son Domnall roundly defeated the D

al Riata, led by Domnall Brecc, both on

the Irish mainland, where he had allied with the D

al nAraide overking, and

also at sea, where he had allied with the Cen

ogain. From that time D

el nE

Riata interests focussed much more closely on Scotland, Ulaid overkingship

conﬂicts were less volatile, and the U

ıN

eill were much more consistently

successful. Domnall mac Aedo died peacefully in his bed (and – for his unusual

success – was the ﬁrst to be called ‘king of Ireland’ by the contemporary

annalist).

Though the U

ıN

eill were more consistently successful, in other words

constantly made themselves overkings and overoverkings, until the end of

the century they continued to battle with each other for supremacy, that is,

to establish overkingships. Although there continued to be conﬂicts within

northern and within southern U

ıN

eill branches, the most prominent conﬂicts

were between north and south and were increasingly focussed on the kingship

of the symbolic site of Tara: he who held Tara might be said to be overking

eill overkings. So, Diarmait mac Aedo Sl

of all U

ıN

aine of the south shared

the Tara kingship with his brother Blathmac and defeated Conall of Cen

Conaill in 654; his successor Finnechta, having killed his ﬁrst cousin the king,

ael D

el nE

defeated M

uin of Cen

ogain in 675; but Loingsech mac Oengusso of

the northern Cen

el Conaill, the second Irish king to be called ‘king of Ireland’

bya contemporary annalist, took the Tara kingship without opposition and

ruled till 703.

It is from the late seventh century that the ﬁrst list of Tara

kings survives, the Baile Chuind, and this in itself emphasises the now crucial

signiﬁcance of the Tara kingship in the politics of the north and midlands.

During the seventh century the geographical range of the prominent U

ıN

eill

overkings had increased dramatically: originally playing politics within the

narrow conﬁnes of a few tuatha,by 700 they moved over half of Ireland. In the

eighth they were to range even wider, as also to establish succession to the Tara

kingship, which from 734 alternated between Cen

ogain of the northern

el nE

Traditionally in 575, but perhaps as late as 590,Meckler (1997).

Annals of Ulster s.a. 565, 574, 575, 577, 578, 580, 581, 582, 586, 626, 627, 629, 637, 642.

Annals of Ulster s.a. 654, 675, 703.

Murphy (1952); Byrne (1973), pp. 91, 104–5, 276–7;Irwin (1998).

Cambridge Histories Online © Cambridge University Press, 2008

244 w endy davies

eill and Clann Cholm

eill. By 700 there were U

ıN

ain of the southern U

ıN

still other kings and other overkings – both Ulaid and Airgialla had their own

overkings until the ninth century – but it was absolutely clear that the U

ıN

eill

group provided the topmost overkings of the north and midlands.

By this time the Irish were also a signiﬁcant presence in Scotland. We do not

know, as indicated above, when the migration to western Scotland began nor

the eventual size of it, but its area seems to have been very conﬁned in the sixth

and seventh centuries. Despite this, the rulers of the D

al Riata ranged widely

across northern Britain in the late sixth and early seventh centuries, coming

at ﬁrst into conﬂict with the English as they sought to establish Bernicia

(Aed

ain was severely defeated by the English at Dawston) and

an mac Gabr

subsequently into conﬂict with the Picts.

The meeting at Druim Cett had

been something of a ﬁrst attempt at loosening the ties with the Irish mainland

and stopping the draining of resources to the Ulaid overking; the battles in

Man in the 580swere a further manifestation of this dispute. But it was the

Magh Roth defeat, in 637, which really concentrated D

al Riata attention on

the Scottish mainland. From their political focus at Dunadd – a site that has

been well excavated – they went north and east in the seventh century, although

there were other forts too, Dunaverty, Tarbert, Dunollie.

This threatened the

British kingdom of Strathclyde, with its focus at Dumbarton (near Glasgow),

and took the Irish over the mountains to the Pictish heartland of Fife. By 700

they were still in the early stages of a development that was to take them in the

eighth and ninth centuries to many conﬂicts in central Scotland and ultimately

to the kingship of Scotland.

Even in the seventh century the U

ıN

eill had sometimes come into contact

with the prominent dynasties of southern and western Ireland. The Laigin,

the people of the south-east (Leinster), also had claims on the kingship of Tara

and they sometimes raided northwards, as they did successfully at Dun Bolg

against Aed mac Ainmere in 598, and were themselves raided, as they were

by Domnall mac Aedo in 628.Within Leinster two leading dynasties vied for

ultimate overkingship, the U

ıD

unlainge in the north and the U

ı Cheinnselaig

in the south, and though U

ıD

unlainge were often overkings this was by

no means invariable. With family connections with the major monastery of

Kildare, they sometimes sought alliances with branches of the southern U

eill in the midlands (the daughters of Cellach, c.700, seem to have taken

ıN

eill husbands) and their conﬂicts extended to the western borders of

Leinster, where the unusually large tuath of the Osraige sported kings who

Bede, HE i.34; see also Thacker, chapter 17 below; Annals of Ulster s.a. 681, 683, 694;Koch (1997),

pp. xxxv–xli.

Lane (1984); Campbell and Lane (1992); Lane and Campbell (1993).

Cambridge Histories Online © Cambridge University Press, 2008

The Celtic kingdoms 245

were extremely resistant to the dominance of overkings, a pattern that was

sustained for many centuries after 700. The to-and-fro of battles between U

unlainge and U

ı Cheinnselaig continued past 700 and it was not until the

mid-eighth century that the former succeeded in establishing an overkingship

of Leinster that then continued until the eleventh century.

The U

ıN

eill also had close connections in the west, the area from which they

were thought to have originated, and one of the several prominent dynasties

uin, had fought with Cen

ogain and Cen

of the west, the U

ıBri

el nE

el Conaill

against Diarmait at Cuil Dreimne in 561. Later, in 703,it was Cellach of the

ıBri

uin who defeated Loingsech, king of Tara. Another dynasty, the U

Fiachrach, came up against the Cen

el Cairpri in 603 in Sligo, at the western

eill political inﬂuence, and against Diarmait mac Aedo Sl

extremity of U

ıN

aine

of the south in 649, when he defeated their long-lived king Guaire. Other

dynasties vied for overkingships, the U

ıMaine and the U

ı Ailello especially,

although the U

ıFiachrach were most notable through the seventh century and

ıFiachrach and U

ıBri

uin were tending to alternate in a principal overkingship

of Connacht by 700.But here again it was not until later, in the late eighth

century, that control of a consistent overkingship was established by the U

Bri

uin.

The south-west, Munster, is the most difﬁcult area to penetrate in the sixth

and seventh centuries, for there is little contemporary material in the annals

until the eighth. It seems that Thomond (modern Co. Clare) was a region

disputed between the overkings of the south-west and those of Connacht to

the north. When Guaire of the U

ıFiachrach, together with the U

ıMaine, was

defeated by Failbe Flann of Munster in 627, this may well have determined

that Thomond was to be dominated by Munster kings and detached from

Connacht, and the defeat contributed to the weakening of the southern U

Fiachrach. In Munster itself, Eoganachta kings created overkingships in east

and west, but it is too early to discern the pattern of their interrelationships. It is

very likely that Cashel, the later symbolic focus of southern overkingship, was

already utilised as a strongpoint, but it does not yet appear to have gained the

signiﬁcance that it had by the ninth century. Indeed, it would be straining the

evidence to suggest that a single overkingship of the south-west had crystallised

by 700. That development was yet to come.

By 700 we can see the shadow of the trend that was to clarify during the

eighth century, the establishment of four principal overkingships of roughly

equal territorial range, relating to north, west, south-east and south-west

(Ulster, Connacht, Leinster, Munster), although the contest for overkingship

Annals of Ulster s.a. 598, 628, 738.

Annals of Ulster s.a. 561, 703, 603, 649, 754, 792.

Annals of Ulster s.a. 627, 721, 737, 793; Annals of Inisfallen s.a. 721, 735.

Cambridge Histories Online © Cambridge University Press, 2008

246 w endy davies

of Munster had yet to start. In 700 the politics of the ‘middle’ were rather

closely associated with the north, although there were forays west and south.

In 700 the D

al Riata had a very solid political base in south-west Scotland and

an established inﬂuence in Scottish politics. These trends are clear, although

the framework of the political dynamic can still be puzzling and the observer

has to ask why there was so much ﬁghting, whether warfare was endemic and

whether early Ireland was in a state of perpetual anarchy. The answers are nei-

ther so simple nor so depressing. The record is dominated by ﬁghting because

this is what the annalists chose to record. It certainly seems to have been a

recurrent activity of a military elite, but this should not make us forget that

these were elite rather than common pastimes; ﬁghting did not mean total

warfare; production continued (indeed, as I have indicated earlier, everything

suggests that it increased in the seventh century); and learning ﬂourished: this

is the period when scholars learned to speak and write new languages, grappled

with complex computations, and began the high art of ornamental bookmak-

ing.

The basis of overkingly power was undoubtedly military; little about it

in this period was institutionalised and it had little (perhaps nothing) to do

with government; it was largely about the very crude, and unstable, politics of

learning how to extract – and keep – surplus. But simple kingship still very

much had a function at this time. It was still the king who represented and

protected his tuath, who dealt with emergencies, who guaranteed prosperity

and good fortune, whose wisdom could be evoked to make a judgement, who

negotiated with Christian clergy and – increasingly – acquired a foothold of

inﬂuence in the ﬂourishing religious communities. Even if the course of poli-

tics was largely made by overkings, ordinary kings still had local and regional

inﬂuence. The king of the Osraige, lying between Leinster and Munster, may

well have had an unusually large tuath but he provides a nice demonstration

of what kingship could do, without the complications of overkingship.

It is through the justice of the ruler that he secures peace, tranquillity, joy, ease, and

comfort.

It is through the justice of the ruler that he dispatches (great) battalions to the borders

of hostile neighbours . . .

It is through the justice of the ruler that there is abundance of every high, tall corn.

It is through the justice of the ruler that abundance of ﬁsh swim in the streams . . .

pictland and north britain

Although the Irish were already signiﬁcant in western Scotland in the sixth and

seventh centuries, the major population groups of north Britain were Pictish,

On Christian culture in Ireland, Stancliffe chapter 15 below.

Audacht Morainn,p. 7.

Cambridge Histories Online © Cambridge University Press, 2008

The Celtic kingdoms 247

British and English. The fact that the monastery of Iona was founded off the

west coast of Scotland in 563 and the monastery of Jarrow was founded about

a century later in the far north of England, and that both were major centres

of recording, writing and study, means that we know something about the

relationship between these groups at this early period – certainly far more than

we know in later centuries.

Though we can learn very little about them before this, by the late sixth

century there were a number of British kingdoms in the area which is now the

Scottish Lowlands and northern England.

British kings fell to the English in

the Newcastle–Bamburgh area and beyond in the decades running up to 600 –

Morgan, Urien, perhaps Gwallog – but a British king survived in southern

Yorkshire until about 617.

Despite that survival, this is essentially a story of

defeat – of military catastrophe, followed by the loss of political identity and

political independence – a tale and a mood encapsulated in the poem known

as The Gododdin (though more economically, and as powerfully, in the poems

of Canu Taliesin).

Warriors went to Catraeth, their host was swift,

Fresh mead was their feast and it was bitter,

Three hundred ﬁghting under command,

And after the cry of jubilation there was silence.

Though they went to churches to do penance,

The certain meeting with death came to them.

Warriors assembled from far and wide for the defence of Rheged, a large

kingdom that probably spanned the Pennines and may even have extended into

Galloway, and they met the English at a decisive battle at Catraeth, probably

Catterick, shortly after 600. Thereafter Rheged simply disappeared. Located

around the Firth of Forth, the kingdom of the Gododdin itself, whose warriors

gave their name to the poem, survived a further generation: its strongpoint at

Edinburgh was ﬁnally captured by the English in 638 (see Map 5).

The kingdom to the west, Strathclyde, had a far longer life and its kings, who

are recorded until 1034, interacted with both the Irish and the Picts over a long

period. It lay around the Firth of Clyde, with its political focus at Dumbarton

Rock. St Patrick had referred to its king, Ceredig, in the ﬁfth century and

Adomn

an, writing about 700, knew of a powerful Strathclyde king of the late

sixth century who had had dealings with the Irish saint Columba – Rhydderch

Hen, Rhydderch the Old.

We know very little about the kingdom, beyond

the fact that members of its royal family sometimes married into the Pictish

Though see below for ﬁfth-century Strathclyde; for all this see Koch (1997).

Historia Brittonum c.63.

Aneirin, YGododdin, lines 78–84; Canu Taliesin.

Patrick, ‘Epistola’ c.2; Vita Columbae i, c.15.

Cambridge Histories Online © Cambridge University Press, 2008

248

50 miles

50 100 km

Orkney

Isles

Rona

Burghead

Lismore

Dunollie

Dunadd

Applecross

Stirling

Dundurn

Antonine

Abercorn

Whithorn

Isle

of Man

RHEGED

Catterick

Jarrow

GODODDIN

Holy Island

Abernethy

Nechtansmere

Iona

Eigg

Mull

Skye

DÁL

Dumbarton

wendy davies

Tarbat

ATHOLL

Wall

Traprain Law

Tiree

RIATA

Map Northern Britain

Cambridge Histories Online © Cambridge University Press, 2008

The Celtic kingdoms 249

ruling families and that Dumbarton was powerful and symbolic enough in

the mid-eighth century to merit attack by Oengus of the Picts and in the late

ninth by the Vikings. Though not large, Strathclyde was clearly a consistently

signiﬁcant independent ‘enclave’, which its more powerful neighbours to west,

north and east could not ignore.

We are accustomed, however, to regard the Picts as a problem.

There are

problems of identity, settlement zone and political structure, to name but the

most serious, for the Picts themselves have left no written records beyond the

symbol stones, whose reading still poses problems;

our knowledge of them

derives almost entirely from external observations. Even the name Pict seems

to have been devised by writers outside Scotland and probably conceals a range

of different tribal identities, for Roman writers in the ﬁrst and second centuries

ad had used tribal names rather than the generic ‘Pict’. The Picts of the late and

post-Roman period were people who raided the provinces of Roman Britain

from their bases in the north, and the name may well have been applied to

all the inhabitants of central and northern Scotland beyond the Antonine

Wall. This need not imply any single social or political identity of the people

so-called, nor even a single language, and it is likely that there were several

groups in the ﬁfth and early sixth centuries. The establishment of the Irish in

the south-west and the English in the Scottish Lowlands obviously made for

change, as did the introduction of Christianity, and though the evidence is

far too fragmentary to be satisfactory it is not unreasonable to suggest that by

the seventh century – whatever the background and whatever the constituent

elements–a single political identity was developing.

The nature of Pictish kingship is a related problem. For reasons of size, space

and communications it is extraordinarily unlikely that there was no more than

a single Pictish kingship in the ﬁfth and sixth centuries, and as late as 739

we hear of a king of Atholl (in eastern Scotland) who was drowned by the

‘king of the Picts’; much later tradition has seven constituent and potentially

independent regions (though this could well be a deduction about origins

derived from existing administrative divisions).

The references are very few

but they suggest that until the mid-eighth century there were sometimes (if not

always) several kings in Pictland. This seems on the surface to be belied by the

Pictish kinglists. A list of Pictish kings was compiled on the assumption that

there was a single line of kings, including virtually every Pictish king who is

named in any contemporary source. It therefore appears to supply a record of a

very long-lived monarchy. It is clearly misleading in its earlier phases, although

in the end a single Pictish kingship undoubtedly was established. There is, of

Witness Wainwright (1955,repr. 1980); Small (1987).

Forsyth (1997).

Henderson (1967), pp. 15–18, 34–41;(1975).

Annals of Ulster 739; Annals of Tigernach s.a. 738 (Revue Celtique 17: 243).

250 w endy davies

course, the possibility that the kinglists record not a line of monarchs but a line

of Pictish overkings. Although it is true that Adomn

an, writing c.700, talks of

the great king Brude of the late sixth century and a regulus,lesser king, of the

Orkneys, an overkingship relationship is not made explicit in this reference

and there is little else to support it positively. Though possible, it remains to

be demonstrated that overkingship was central to Pictish political structures.

Whatever we might think about the background, for much of the seventh

century there was a principal king of the Picts, whose focus of royal power lay

in the Pictish heartland of eastern Scotland. Although many have disputed the

observation, Bede’s statement that Pictish royal succession was matrilineally

determined makes sense at this period.

A king tended to be succeeded by his

sister’s son, or by his brother, rather than by his own son. The fathers of Pictish

kings sometimes have British names, like Beli of Strathclyde, and sometimes

have English names, like Eanfrith (of Bernicia), and it is reasonable to sup-

pose that energetic males from neighbouring parts sometimes married eligible

Pictish royal females and thereby either acquired kingship for themselves or

fathered Pictish kings.

The English names are indicative of a sustained inter-

est in Pictland through much of the seventh century. By the late sixth century

the English kingdom of Bernicia was well established on the eastern coast of

the Scottish Lowlands and its kings campaigned widely; we have already noted

that Aed

ain was defeated by the English king Æthelfrith in 603.

an mac Gabr

By the second quarter of the seventh century, the English were campaigning

northwards, taking the British strongholds of Etin (Edinburgh) and Stirling in

638 and 642, and this brought them into the Pictish heartland.

The power of

Oswiu, king of Bernicia and Northumbria 643–671 and probably most pow-

erful of all the English kings, seems to have stretched by the 660sto include at

least southern Pictland; he and his son Ecgfrith seem to have operated effec-

tively as overlords, with Pictish subordinate kings.

Pictland, then, was not

only subject to English military campaigning during the seventh century; it

was also subject to English political control for about twenty-ﬁve years. The

control did not go unopposed: the subject king Drost was thrown out, there

were rebellions, and ultimately Ecgfrith was defeated and killed at the famous

battle of Nechtansmere, Dunnichen, in 685,bythe Pictish king Brude, son

of the British king of Strathclyde.

This was effectively the end of English

domination, although there were still some military encounters in 698 and 711

and, for a generation, English ofﬁcials attempted interference.

See Anderson (1973), pp. 77–102; Vita Columbae ii, c.42.

Bede, HE i.1.

Marjorie Anderson’s treatment of these problems (1973)remains extremely impressive.

Anglo-Saxon Chronicle s.a. 603; Annals of Ulster s.a. 638; Historia Brittonum c.65.

See Thacker chapter 17 below.

Bede, HE iv.26.

The Celtic kingdoms 251

We hear of much military activity in the later seventh and early eighth

century and of the sieges of military fortresses – Dunnottar on the east coast in

681,Dunadd on the west coast and Dundurn in the midlands in 683, Scone and

Moncreiffe a little later. There seem to be at least two elements to the conﬂict:

ﬁghting for supremacy among Picts, perhaps to establish the single kingship,

and ﬁghting for control of territory, in particular with respect to the Irish in the

west. In the late sixth century, the Picts had been defeated by raiders from D

Riata; by the mid-eighth they were raiding, and could take Dunadd itself, the

Dalriatan centre on the west coast.

The period between 685 and c.730 seems

to be one of clariﬁcation and consolidation of the Pictish kingship, of Pictish

political identity and of power. Although there were conﬂicts between Picts,

King Nechtan could enter into diplomatic relations with the Northumbrians in

the early eighth century and could expel the Columban (i.e. Irish) clergy from

the whole of Pictland,

while King Oengus in the mid-eighth century could

move about central and western Scotland with impressive military capacity.

Later material suggests that this Pictish kingship began to develop institutions

of government.

wales

Among the few fragments of contemporary evidence that we have for Wales

in the early sixth century, the work of Gildas is exceptionally important for

establishing the existence of two major kingdoms, Gwynedd in the north-west

and Dyfed in the south-west.

He makes it clear that they were not recent

creations and that their kings belonged to dynasties with sustained interests

in the kingdoms. It is quite possible that one or two of the other kings that

he names as his own contemporaries were also Welsh kings, but we do not

know where he was writing and we cannot reconstruct with any conﬁdence

the local contexts of each man. From the later sixth, and seventh, centuries

we have – unusually for Celtic areas – some charter material relating to the

south-east, most coming from the Book of Llandaff but with a few charters

being attached to the Life of Cadog from Llancarfan.

This material has been

copied and recopied many times, and ‘improved’ by a succession of editors, and

it cannot be taken at its face value. However, the bare elements that lie behind

the surviving charter texts make sense and suggest a coherent and credible

pattern of political development for the region.

Annals of Ulster s.a. 736.

SeeStancliffe, chapter 16 below.

Cf. Barrow (1973), pp. 7–68;Davies (1993).

Gildas, De Excidio cc.31, 33.

Book of Llan D

av, nos. 72a–77, 121–127b, 140–66; Vita Cadoci cc.55–68.See Davies (1979a) and

(1982b), p. 260, for critical assessment.

SeeDavies (1978).

252 w endy davies

The Llandaff material ﬁlls out the sixth-century political structure by show-

ing that in the late sixth century there were several very small kingdoms in the

south-east; these were Gwent, Ergyng (south-west Herefordshire), an unnamed

kingdom near Cardiff and one in Gower. We can also deduce with some cer-

tainty, from early seventh-century references in other texts, that another king-

dom had its focus in the north-east; whatever it was called then, it was the

ancestor of Powys. We therefore have reasonable knowledge that by 600 most

parts of Wales were part of some kingdom, although there were considerable

differences in the size of these units. (There may of course have been rather

more small units, which go unnoticed by our fragmentary source material. We

know least about mid-Wales, and about the central eastern borders in particu-

lar. It is possible to make a case, for example, for the existence of kingdoms of

Rhos (mid-north), Meirionydd and Ceredigion (centre west coast), Builth and

Brycheiniog (east, south of centre), and Shropshire/mid-Wales. None of these

is evidenced in the sixth century, but Ceredigion, Builth and Brycheiniog – at

least – have an eighth-century history and could well have had earlier origins

(see Map 6).)

We do not know how the kingdoms arose. Both Dyfed and Ergyng had some

relationship with the Roman past, the former taking its name (and basic shape)

from the civitas of the Demetae, the latter taking its name from the Roman

city of Ariconium (Weston-under-Penyard) and perhaps relating to the city’s

territory. They may, therefore, in some sense be continuities of units of Roman

local government, within a different political framework and with a new basis

of political power. The Cadelling, the dynasty that came to rule Powys, may

well have had a Chester origin. Others could have originated as a result of the

settlement of immigrant groups, like the Irish in the west. Later Welsh writers

recognised a powerful tradition that some of the Votadini (i.e. Gododdin)

had come from northern Britain to settle in the north-west, under a leader

called Cunedda.

Another tradition about the ﬁfth century had Vortigern’s

(Gwrtheyrn’s) family associated with Roman Gloucester, from which they were

displaced to establish the kingdoms of Builth and Gwerthrynion.

These are

ideas about origins; they suggest models of development; each is possible;

they may or may not relate to what actually happened and the real formative

processes.

The ﬁfth-century background is exceptionally hazy: we simply cannot per-

ceive the process of devolution from Roman imperial province (or part of one)

to complex of kingdoms, although we can see, at a more local (and a more

proprietary) level, that in the south-east many late Roman estates continued as

SeeDavies (1982a), pp. 85–102, for a longer discussion.

Historia Brittonum c.62.

Historia Brittonum cc.47, 48, 49.

Cf. Dark (1993), pp. 78–86.