According to his obituary in the Anglo-Saxon Chronicle, King William subdued both Wales and Scotland and would have conquered Ireland too had he lived a few more years. The comment reflected the aura of William’s power but not the reality of his policies, which were more defensive than aggressive. This was chiefly because of the limited value of Celtic Britain and William’s higher priorities elsewhere. It was also because of the nature of William’s kingship. Lanfranc, on becoming archbishop of Canterbury, investigated the ancient rights of his see and argued he was ‘primate of all Britain’. William, on the other hand, rarely toyed with the British imperial titles sometimes adopted by the Anglo-Saxon kings. Instead he proclaimed himself simply ‘king of the English’, a kingship which at most carried claims to tribute from the Welsh rulers and a loose overlordship over the king of Scots. William inherited no lands from the Confessor in Wales or Scotland and gained none from the forfeitures after the Conquest. He used Harold’s estates in Herefordshire to set up a frontier lordship for William fitz Osbern. It was not William but his Norman barons who were to transform the face of Wales, gripping it within a generation far more fundamentally than had the Anglo-Saxons in many centuries. No similar transformation in this early period took place in Scotland. Indeed in the north it was the king of Scots who was the aggressor, not the Normans.
* * *
In its entry for 1069 the Brut, the principal native chronicle for the whole period covered by this book, implied that Wales was divided into three political entities: Gwynedd, Powys and Deheubarth. Gwynedd was the whole of north-west Wales from the estuary of the Dee to the estuary of the Dyfi. Powys was north-east Wales from the Dee to around the upper valley of the Severn. Deheubarth meant broadly the whole of south Wales, Wales that is to the south of the Dyfi estuary and the Wye, although it came later to mean simply those parts of south-west Wales which had escaped Norman rule. Gwynedd and Powys were far more coherent politically than Deheubarth but all three were liable to division between rival rulers. This was facilitated by the way they were made up of a series of smaller administrative regions called cantrefs and commotes, the latter sometimes being subdivisions of the former. There were five commotes, for example, within Gwynedd’s western cantrefs of Llŷn, Ardudwy and Meirionydd. Cantrefs and commotes could themselves form parts of larger units. In Deheubarth, the latter included Ceredigion, Dyfed, Yystrad Tywi, and Glamorgan, the last with kings of its own. All these regions appear as separate entities again and again in the Brut. In Gwynedd the four cantrefs between the Conwy and the Dee gave the name ‘the Four Cantrefs’ to the whole area.
There has been much debate about the origins of the cantrefs and commotes. Had they grown up almost as separate kingdoms so that their holders enjoyed near kingly rights and status, or had they been created from above by kings who decided – at least in theory – the authority their lords enjoyed? Whichever was the case, cantrefs and commotes were both centres of lordship, with dues paid to a central court, and focuses of community; hence the ease with which Gwynedd, Powys and Deheubarth could be broken up, and the ‘hostile heart’ engendered by the round of plundering warfare between men of different regions.
Underlying many of these features were the basic facts of geography. Nearly all the divisions mentioned above had their own geographical logic, very often drawn by rivers. A single political entity embracing them all had none. The great mountains which dominated Wales prevented any easy journey across the country. In the north, Gwynedd itself was divided naturally by the river Conwy. Viewed from the eastern side of its estuary on the great rock of Deganwy, the menacing mountains of Snowdonia to the west, running sheer into the sea, still seem to guard an almost impenetrable land. In the south-west, Ceredigion centred on the narrow coastal plain between the rivers Dyfi and Teifi. In the south the major divisions all had their own characteristics: the lowlands of Dyfed with the great ecclesiastical centre of St Davids; Ystrad Tywi, the heart of the kingdom of Deheubarth, with its mountainous cantrefs of Mawr and Bychan severed by the deep-grooved Tywi; Glamorgan, cut through by rivers, with lowlands in the south, uplands in the north, and then to the east the rich pastures of Gwent. Acting as a hinge between north and south Wales, there was a mountainous region described as ‘between the Wye and the Severn’. The upper valleys of these two great rivers provided avenues eastwards into England and westwards through to south-west Wales. The strategic importance of this area led to constant battles for its control.
Given this geography it was not surprising that Wales was subject to a multiplicity of competing rulers. Before his defeat by Harold in 1063, Gruffudd ap Llywelyn had brought all Wales under his rule but it was a brief and unique achievement. A single language, a native law and a common descent from the ancient Britons and so, as legend had it, from Brutus and the Trojans made the Welsh think of themselves as a separate and distinctive people. But there were conflicting ideas about how or indeed whether this needed to be expressed politically. ‘They obstinately and proudly refuse to submit to one ruler,’ commented Gerald of Wales towards the end of the twelfth century. Welsh law books of that time and later might begin with a vision of a united Wales basking in the rule of the tenth-century King Hywel the Good, but they actually dealt with a Wales in which there was a plurality of kingdoms. (See below, p. 228.) Whether Welsh law also enforced division by stipulating that kingdoms should be apportioned among the sons of a ruler is more questionable. Such partition was certainly the law and custom with ordinary patrimonies, but with kingdoms the laws envisaged a single heir, the edling, designated by the ruler. Just who the edling should be, however, was less clear. Throne-worthiness was not confined to the sons, let alone the eldest son, of the previous king; therefore the potential existed for disputes over the succession, in the course of which kingdoms were divided up, so much so that, whatever the precise law, the practice became regarded as almost customary. Claims and feuds were also encouraged (as Gerald of Wales observed) by the practice of fostering out sons to different noble families and by the intense pride in lineage – Rhys ap Tewdwr ap Cadell ap Einon ap Owain ap Hywel Dda, ran one genealogy.
Conflict was also fostered by the nature of Welsh kingship. The Brut frequently used the word brenin which in Latin was translated as rex, that is ‘king’. Yet there was not one of Henry II’s knights (it was later said) who did not regard himself as worth a Welsh king. The latter were very different from those of England and not just because of their puny resources. They went through, as far as is known, no coronation or inauguration ceremony; and they had a much smaller role in the maintenance of law and order, which was largely a communal responsibility. To a far greater extent than their English counterparts they were simply warrior chiefs, their aim to secure ‘vast spoil and return home eminently worthy’ as Gruffudd ap Llywelyn did after his campaign of 1055.
The nature of this warfare and the politics which went with it was extremely violent, sometimes exultantly so. ‘Amidst that [battle] Trahaearn was stabbed in his bowels until he was on the ground breathing his last, chewing with his teeth the fresh herbs. Then Gwcharki the Irishman made bacon of him as of a pig.’ So the Life of Gruffudd ap Cynan celebrated the death of his rival, Trahaearn ap Caradog, at the battle of Mynydd Carn in 1081. Having won this victory Gruffudd marched to Powys where, according to theLife, ‘he straightaway displayed his cruelty in the manner of a victor’. ‘He destroyed and killed its people, burned its houses and took its women and maidens captive… He destroyed the land completely.’ Two things separated this kind of politics and warfare from that evolving in the Anglo-Norman world (discussed more fully below, pp. 126–7). One was the slaying of noble rivals. Between 1069 and 1081 no less than eleven Welsh princes fell in the violence. A minister of Henry I (1100–1135), in a later period of strife, arranged a truce ‘out of love of the land for he knew they were all killing one another’. The second difference was the seizure of women and children to be kept or traded as slaves, something encouraged by the absence of castles in which to keep prisoners and of money with which to ransom them. Thus the ruler Iorwerth ap Bleddyn was allowed to promise Henry I £300 of silver ‘in whatever form he could, horses, oxen and other things’.
At the heart of the violence between 1069 and 1081 were the conflicts within the kin of Gruffudd ap Llywelyn, king over all the Welsh, who was murdered by his men in 1063 following his defeat by Harold. ‘After innumerable victories… he was now laid in the waste valleys,’ lamented the Brut. At last in 1081, after the great victory at Mynydd Carn, Rhys ap Tewdwr established himself in Deheubarth, ‘the kingdom of the south’, while his ally Gruffudd ap Cynan, grandson of a king of Gwynedd, bid for supremacy in the north. These events had not been played out in a vacuum. The whole of the west coast of Wales was very much within the orbit of Irish and Scandinavian politics. Gruffudd ap Cynan’s father had fled to Ireland where he married the daughter of Olaf, the Danish king of Dublin, Gruffudd himself being the fruit of the union. Gwcharki who made bacon of Trahaearn was only one of many Irishmen who swelled the armies of the Welsh rulers. To the east, on the other hand, Wales was accessible to England. Before the Conquest there had been English settlement as far west as Rhuddlan in the north. In the south, Harold had established a hunting lodge at Portskewet in Gwent although this was soon burnt down by the Welsh. The Normans would not be so easily removed. Their hand was already apparent at the battle of Mynydd Carn for King Caradog of Glamorgan, killed there alongside Trahaearn, had been a client of King William. The purpose of William’s solitary expedition to Wales in 1081 was to secure a similar submission from Caradog’s supplanter, Rhys ap Tewdr. Meanwhile within a year of his victory at Mynydd Carn, Gruffudd ap Cynan found himself a prisoner of Earl Hugh of Chester, who was determined to establish is own supremacy in the north. ‘That was the first plague and fierce advent of the Normans to the land of Gwynedd,’ groaned Gruffudd’s Life.
The Normans had reached Wales very soon after their arrival in England. Within three years of Hastings William had established William fitz Osbern, Roger of Montgomery and Hugh of Avranches as earls with great power in the border counties of Herefordshire, Shropshire and Cheshire respectively. This was a strategy to protect the frontier from Welsh incursions. But it could also serve as the base for Norman advance. In the south William fitz Osbern established a castle at Chepstow at the mouth of the Wye and another castle higher up the river at Monmouth, and began the advance into Gwent. By 1086 there was a Norman castle at Caerleon on the Usk, and its lord held considerable lands west of the river. William also established Clifford castle four square in the upper valley of the Wye, and this could support the Norman advance (gathering pace in the 1090s) towards Brecon and Builth. Fitz Osbern’s death in 1071 and the rebellion of his son Roger in 1075 inevitably slowed the pace of advance, the Conqueror being content to hold the line through the sheriff of Herefordshire and receive tribute from the Welsh rulers. Domesday Book shows that paid by Rhys ap Tewdwr was an annual £40. Having secured Rhys’s submission William’s 1081 expedition metamorphosed into a pilgrimage to St Davids. Until his final demise in 1093, Rhys’s problems were less with the Normans than with his Welsh rivals whom he defeated and killed (after a period of exile in Ireland) in 1088 and 1091.
All this was to change in April 1093 when Rhys was killed by the Normans edging into Brecon. Cadwgan ap Bleddyn of Powys then immediately plundered Ceredigion and Dyfed and departed. Two months later the Normans overran the same areas and stayed. Arnulf, Roger of Montgomery’s son, now established the first castle at Pembroke, one never afterwards taken by the Welsh. South-west Wales, ‘which was not in their power before that’, as the Brut put it, thus came under Norman rule. Rhys’s death must likewise have facilitated Bernard of Neufmarché’s conquest of Brecon and further south Robert fitz Hamon’s of southern Glamorgan, for which Cardiff castle was the base. The Normans also moved into upper Gwent, the base here being the castle at Abergavenny.
Further north, after 1066, Roger of Montgomery’s position in Shropshire had actually been more dominant than that of fitz Osbern’s in Herefordshire because he was given all the non-ecclesiastical land in the shire. He parcelled much of it out, in solid strategic blocks, to his followers, who built their own mottes, like that of the Corbets at Caus and the Says at Clun. Roger himself established the first castle of Montgomery, named after his home in Normandy, just as Caus was named after the pays de Caux. Montgomery was a hinge on which Norman and subsequently English control of Wales turned. It stood just to the west of Offa’s Dyke, at the point where the Severn plunges into the mountains of central Wales, to connect via narrow passes with the Dyfi valley. Thence the way was opened to Aberystwyth and the whole of south-west Wales. Doubtless this was the route (protected by eight mottes built down the Severn valley) the Normans took when they advanced into Ceredigion and Dyfed.
It was north Wales which saw the most spectacular and ultimately the most illusory of these early Norman advances. In Cheshire Earl Hugh had received a concentration of lands much like Roger of Montgomery’s in Shropshire. From there his nephew, Robert, had advanced probably in the 1070s to establish a castle at Rhuddlan, by this means controlling the valley of the Clwyd. The seizure of Gruffudd ap Cynan after his victory at Mynydd Carn was clearly designed to eliminate any challenge from that quarter. By the time of Domesday Book the Normans had pushed on to the Conwy above which Robert had probably completed his castle of Deganwy. By sea as by land, that was the base for the conquest of the rich corn lands of Anglesey together with most of the rest of Gwynedd, the hold consolidated by castles which in their skilful siting (one was at Caernarfon) foreshadowed those of the ultimate Edwardian conquest in the thirteenth century.
With Anglesey, Gwynedd and the rest of the north held by the castles of Robert of Rhuddlan and Earl Hugh of Chester, and with Ceredigion, Pembroke, Glamorgan and Gwent being secured in the west and south, it looked as though all of Wales would soon fall to Norman rule. Yet it was not to be. On 3 July 1093, Robert of Rhuddlan was killed in a skirmish by his Welsh foes, led – at least according to Orderic – by Gruffudd ap Cynan. Next year, in the words of the Brut, the Welsh ‘being unable to bear the tyranny and injustice of the French, threw off their rule’. In the north, Gruffudd ap Cynan and Cadwgan ap Bleddyn of Powys destroyed the Norman castles in Gwynedd and slaughtered a relief expedition. In the south, helped by the death of Roger of Montgomery in July 1094, the new castles in Ceredigion and Dyfed, with the exception of Pembroke, were destroyed in risings. The Normans ultimately held on to most of the coastal lowlands of south Wales but their control of Gwynedd was never fully restored, the greatest reverse they suffered in all their conquests in Britain.
The changes wrought by the Normans remained profound. In the areas which came and remained under their sway, notably the southern lowlands, the native rulers were either eliminated or subjected and eclipsed. At a lower level, some Welshmen became peasants within the manors established by the Normans, being interspersed with English immigrants. In Dyfed in the early twelfth century substantial numbers of Flemings were also introduced (see above, p. 38). Alongside these new manors there were settlements which remained Welsh, some populated by freemen and some by bondmen, both groups now giving their services and renders to their new lords. There was a tendency for these ‘Welshries’, as they came to be called, to be pushed onto less fertile land, hence the profound difference which developed between south and north Glamorgan, lowlands and uplands, one Anglo-Norman and the other Welsh. Yet the transformation in Wales, dramatic though it was, proved less awesome than that in England. The English aristocracy was destroyed after 1066; the Welsh was not. Substantial parts of Wales, if shifting in size, remained under native rulers. In the thirteenth century, the descendants of Gruffudd ap Cynan, Cadwgan ap Bleddyn and Rhys ap Tewdwr still held sway in Gwynedd, Powys and regions of the south.
The Norman conquest of Wales was thus permanent yet incomplete. For the incompleteness there was, of course, one paramount reason, as Gerald of Wales pointed out. Wales was formidably defended by its mountains, woods and rivers. The political fragmentation caused by this geography in fact served Wales well. There could be no Welsh Hastings, no overthrow of the kingdom in a single battle. Wales would have to be conquered piecemeal, with armies, fleets and castle-building all co-ordinated. That, as Gerald noted, would take the ‘diligent and constant purpose’ of the king for at least a year. William, given his scale of priorities, could not give that amount of time. Nor could his successors in the 200 years after 1066. It followed, therefore, that the conquest of Wales remained a baronial, not a royal, enterprise. Yet the barons too had rival preoccupations. William fitz Osbern died fighting for William in Flanders; his son was disinherited after rebellion in England; the Montgomerys had wide lands in France. Moreover once the most fertile parts of Wales, the southern lowlands, had been absorbed, the incentive to ‘go on’ lessened. On a day-to-day basis, the consolidation and continuation of the conquest of Wales were left to knightly tenants with no more resources than their Welsh opponents.
The Welsh, moreover, were redoubtable opponents. They are ‘entirely bred up to the use of arms’, commented Gerald of Wales, almost echoing the Conqueror’s dying remarks about himself. A great chief (like Hywel ap Goronwy) slept with his sword above his head and his spear at his feet, and wished to die in battle, not in bed. Later evidence shows that Welsh rulers could raise armies of foot several thousand strong, exploiting the obligation on all freemen to perform military service as needed within the kingdom and for a period of six weeks a year outside it. (This at any rate was the obligation as stated in the Welsh law books of the late twelfth and thirteenth centuries.) Alongside such forces was the king’s permanent war band, his teulu, which was really the central institution of his kingship. Composed of young nobles maintained at court, the teulu might be used to eliminate the king’s opponents by mutilation or murder, and generally to act as the enforcer of his rule. It formed the core of royal armies, and was doubtless the chief beneficiary of the plunder. Confronted by such lightly armoured, sure-footed warriors, equipped with bows and spears, masters of the sudden ambush and the quick retreat, the armies of the Normans rarely achieved decisive victories. The Welsh, moreover, also learnt from the Normans. They had long bred horses and now learnt to fight on them. They donned mail and built castles. Militarily they thus had the best of both worlds.
The permanence of the Conquest, even if the conquered areas fluctuated, owed everything to the many-sided abilities of the Normans themselves not only in the realms of violence but also of accommodation. In Wales the Normans were the same explosively confident, brutally professional, free-wheeling warriors who had campaigned in Maine and Brittany and had conquered England. In Wales their war joy and greed could be unconfined. They now operated in the March, that is the frontier zone beyond the English kingdom. Instead of receiving their lands from the Conqueror they carved out their own marcher lordships. Instead of being hedged around by the structures of royal government, within their lordships they exercised almost sovereign power. When the powers of the marcher barons were defined in later centuries, they had near total control over justice (hence the king’s writ did not run) and enjoyed the right to wage war and build castles. Historians once thought that these powers were either conceded by the king or taken over with the cantrefs and commotes of native Wales. But it was far more fun than that. The Normans in Wales were conquerors, not constitutionalists. Sometimes they exploited existing territorial units, but equally (as in Glamorgan) they often ignored them. With the king’s acquiescence, the powers of the marcher lords were taken into their own hands. They were absolutely necessary to conquer and control.
Individual marcher baronies could expand and contract depending on the ebb and flow of conquest. Some remained small like Clun (held directly from the king after the forfeiture of the Montgomery family). Others like Pembroke and Glamorgan came to be considered as the equivalent of English shires. Central to the survival of all of them was the castle; no less than 300 pre-1215 sites have been identified in Wales. Castles were centres of aggressive lordship and bastions of safety in times of retreat. They were also intended to be psychologically crushing: Carew was built on top of ancient earthworks and overlooked a cross commemorating a former ruler of Deheubarth. Many began as simple motte-and-bailey structures topped with wood but usually at some point wood was replaced by stone. Sometimes stone was used from the start. Again and again in the narrative of the Brut, the Welsh risings swept over everything else only to break against the great stone keeps. Nothing marks the distinction between English and Norman methodsmore clearly than the contrast between Harold’s palace at Portskewet, so easily burnt down in 1065, and fitz Osbern’s massive keep at Chepstow, never afterwards taken by the Welsh. Of course, masonry is nothing without the men to man it, and here too the Normans had an answer. Many of the major castles were centres of what historians have called ‘castleries’ where the lord’s tenants held their land in return for providing a garrison for the castle’s defence. At Clun, for example, the building of the castle in the 1090s was quickly followed by the enfeoffment in the surrounding area of tenants who owed a total of seven knights for ‘castle guard’, the rest of the garrison probably coming from paid troops.
There was also a ‘spiritual’ side to the Norman conquest of Wales, as there was to that of England. The Welsh clas (‘community’) churches, with their hereditary clergy, unregulated in behaviour, seemed scandalous to these church-militant Normans. They used the lands of such institutions to found new houses (nineteen were established between 1070 and 1150) and endow existing monasteries in England and Normandy. In one evocative passage, Orderic Vitalis described the peaceful scene in the chapter house of his monastery, St Evroult, when Robert of Rhuddlan confirmed his gifts in England and Wales to the house. In another passage Orderic gave a graphic picture of Robert’s violent death under the towering cliffs of the Great Orme, just north of Conwy, as he rushed forward on foot to confront his foes; two scenes which encapsulate the gallantry, piety and ranging activity of these Norman conquerors.
The Normans were self-confident in their prowess and in their piety, but they were not blindly arrogant. They were quite ready to learn from the Welsh and adapt to conditions, even coming in the early days to see the possibilities of the slave trade. The Welsh long remembered with grudging respect how Earl Hugh of Chester paid off some Scandinavian mercenaries not with young men and women but with toothless hags. Given the Welsh terrain, the Normans soon saw the value of lightly armoured cavalry, and some tenants (for example in the lordship of Oswestry) held their lands in return for providing it. The Normans also recruited foot soldiers, sometimes in large numbers, here drawing on the Welsh who remained within their lordships. The Normans too got the best of both worlds. At a higher level, they were very ready to make alliances with the Welsh rulers, indeed there were Normans fighting for King Caradog at the battle of Mynydd Carn. There was nothing socially demeaning about such contacts. Just as they came to venerate Welsh saints like Dogmael and David, so the Normans also respected the status of the Welsh rulers. Thus Caradog rubbed shoulders with Norman barons at the consecration of St Mary’s church at Monmouth. Later in the 1120s Gruffudd, the son of Rhys ap Tewdwr, though now virtually landless, might still ride in company with two great marcher barons and be complimented on his ‘innate nobility’. By this time intermarriage between Welsh and Norman noble families was common-place. Once it was clear that the Welsh rulers were not going to be destroyed like their equivalents in England, the Normans, if they were to profit from their gains, had just as much interest in peace as they had in war. If violence brought the Normans their possessions in Wales, it was often accommodation which enabled them to keep and exploit those gains.
Of course, it takes two to accommodate and the Welsh proved willing partners. That was another important reason for their survival. King Caradog and Rhys ap Tewdwr were the first in a long line of Welsh rulers who tried by a policy of submission and alliance to limit Norman attacks, while gaining Norman support in their struggles for mastery over their native rivals. The Normans were far more, however, than simply new ingredients in old politics. Their advent was traumatic. ‘Why have the blind fates not let us die?… O [Wales] you are afflicted and dying,’ cried a despairing poet in the 1090s. The whole pattern of life had changed for ever. The Welsh within the marcher baronies faced new and exigent foreign lords. The Welsh rulers now played out their old politics as part of a much greater and more dangerous game, one in which, challenged by the marcher barons and the English king, they strove to retain what they held and recover what they had lost. The game only ended with the final conquest of Wales by Edward I.
* * *
In 1040 Macbeth (of Shakespeare’s play) killed Duncan, king of Scots, and seized his throne. It was not till 1058 that Duncan’s son Malcolm, after a period of exile in England, was effectively restored, thus inaugurating one of the most significant reigns in Scottish history. Two fundamental points need to be made about Malcolm’s kingship. The first is that he ruled an area far smaller than Scotland today. The second is that the Scots were only one of several peoples who inhabited the north of Britain. The history of the north during the period covered by this book is essentially that of Malcolm’s descendants expanding the area of their rule, both by conquest and accommodation, thus creating the boundaries of modern Scotland and a single people of the Scots.
The line of King Malcolm had been founded in the mid ninth century by a king of Scots, Kenneth MacAlpin, who had established himself east of the highlands between the firths of Forth and Moray. That this area was the original heart of the Scottish kingdom is suggested by the way the thanages, the basic units of royal administration in the localities (discussed later), were concentrated within it. It is also suggested by the use of the terms Scotland and the Scots. These are English forms of the Latin Scocia and Scottiwhich were themselves the equivalents of the Gaelic Alba and Albanaig. Now when we first have evidence of geographical meanings in the twelfth century, ‘Scotland’ could still stand for broadly this same area between the firths of Forth and Moray (see above, p. 11). The only difference, for reasons we will mention, was that the use excluded Moray itself, the area running westwards from the river Spey, although it possessed a significant number of thanages and had almost certainly been part of the original kingdom.
The limited extent of early Scotland is easily explained. There was a substantial Norse population in the far north in Caithness. Lothian, running south from the Forth to the Tweed, was largely inhabited by English. In the west, the Galwegians (the men of Galloway) and the Cumbrians (Cumbria extending from the Clyde to the southern edge of the Lake District) had their own identities, ones which owed much to the Norse Gaelic culture of Ireland. The rulers of Caithness, Argyll and Galloway were all striving to increase their power: Galloway itself, as a single political entity, was the creation of Fergus in the 1120s and 1130s. His style ‘king of the Galwegians’ well reflects his own view of his status. The rulers of Argyll and Galloway looked, not east to Scotland, but west to the Irish Sea, where they struggled for mastery over the Isle of Man and Western Isles with the dynasty – often highly factionalized – established as kings of Man by the great Irish warrior Godred Crovan around 1079. Here, as with Orkney and Shetland, the nominal overlord, formally recognized from 1098, was not the king of Scots at all, but the king of Norway.
The MacAlpins were certainly not content with this situation. Their claims to wider authority are reflected in the way ‘Scotland’ was also used in a second, larger, sense to mean the whole area north of the firths of Clyde and Forth. This ‘larger’ Scotland still excluded Lothian and Cumbria, but it was here that the kings made some of their early advances. Between around 960 (when Edinburgh was a royal base) and 1018 they had taken hold of Lothian, placing the south-east border on the Tweed. They had also gradually asserted dominion over Cumbria, ultimately after 1018 using it as an appanage for the heir to the throne. But there were also setbacks. In 1054 Cumbria south of the Solway seems to have fallen to Siward, Earl of Northumbria. Even worse, a division was opening within the original kingdom through conflict with Moray. Moray had been the power base of Macbeth (himself a member of the MacAlpin dynasty) and after his demise, as we will see, a kinsman continued to rule there. The growing fissure explains why in the next century Moray was not part of Scotland at all in the narrow use of the term referred to earlier.
In the assertion of royal power, one important factor made for continuity and stability. Although there might be dispute over who should be king of Scots, the kingship itself was not usually divided. The severance with Moray was the exception which proved the rule. Already therefore there was a fundamental contrast with the multiple kingship prevalent in Wales. The MacAlpin kings, while they were not anointed, enjoyed a much higher status than their Welsh counterparts, being almost certainly inaugurated with ceremony at Scone. The king had also, as Alexander Grant has shown, a local organization which enabled him to exploit his resources. The latter were essentially derived from the royal lands, which were organized into shires or what were later called ‘thanages’ after the official, the thane, who ran them. A thanage, of which seventy-one are known, normally embraced one or two parishes (or what were to become parishes), with the thane collecting renders in kind from the dependent settlements at the estate centre. Although their existence has always to be deduced from later evidence, the great majority of the thanages probably existed at the time of Malcolm’s accession, being largely situated between the Firths of Forth and Moray. Within the area of the thanages the king did not hold all the land, for there were many lords with their own estates. There were also superior officials, who took the proceeds of land attached to their offices, and held the title ‘mormaer’, or ‘earl’. (The titles were interchangeable, with the latter eventually becoming general.) However, the earldoms within the core of the kingdom, Fife, Gowrie, Angus and Mearns, were hedged around and honeycombed by the thanages. Gowrie, indeed, was in the king’s hands. The thanages thus formed the heart of royal Scotland. Round them was a ring of provinces: Menteith, Strathearn, Athol, Mar and Buchan, in which, unlike the inner-core earldoms, there were no thanages at all. These too were under mormaers or earls who were certainly subject to the king in a way the rulers of Galloway, Argyll and Caithness were not. But with their offices becoming hereditary and without thanages to contend with, they also enjoyed a large measure of independence. This distinction between the inner core of the kingdom formed by the thanages and the outer ring by earldoms was to be fundamental to the history of Scotland throughout our period (see Map 3).
Within the kingdom serious crime was settled by compensation, as in England, but without being an offence against the king: there was no king’s peace covering the whole realm. The hereditary legal official, the ‘brithem’, played an important part in both arranging compensation and settling disputes over land. The king, as far as the evidence goes, made no use of writing and minted no coins. Yet in context the MacAlpin monarchs were powerful and successful rulers. It is highly likely that they had more wealth than all the mormaers put together, partly because the lands of the latter were generally in the uplands. Even within the earldoms a general system of military service obtained, similar to the five-hide system in England, with (at least in theory) a specified number of men coming from each ploughgate or its Gaelic equivalent, the ‘davoch’. An important task of the mormaers was to lead out this ‘common army’ from their provinces, an army which fought on foot and could be large and ferocious. The kingdom as a whole was liable to tax, namely ‘cain’, comparable to the geld, though paid in kind and raising unknown amounts. The royal estates themselves seem impressively organized. They provided ‘conveth’, a food render designed to support the royal household. This was organized so as to allow the kings literally to eat their way round the kingdom, in the process doing much to impress royal authority within the core of this small but manageable realm. All this reflected another fundamental difference from Wales, for the Scottish lowlands were sufficiently large, coherent and central, as the lowlands of Wales were not, to encourage the development of a powerful single kingship which would eventually greatly expand its territorial authority.
After ascending the throne in 1058 King Malcolm was indeed eager to expand his power, but should he attempt to do so to the north or to the south? His first marriage to the widow of Thorfinn, earl of Orkney might help him reassert authority in Moray and lay claim to over-lordship further north. In 1078 he crushed Macbeth’s step-grandson, Malsnechtai, who ruled in Moray, seizing his treasures and cattle. But the victory seems short-lived for in 1085 the Ulster annals recorded the ‘happy death’ of Malsnechtai ‘king of Moray’, and the history of the area is then obscure for over forty years. Essentially Malcolm looked south. The recovery of Cumbria south of the Solway was a major ambition and there were also rich pickings to be had across the Tweed border. An English king based in the south before 1066, and thereafter a Norman one who was often overseas, together with the general chaos of the Conquest, all encouraged Malcolm to chance his arm. He was to lead no fewer than five southern expeditions, being remembered at Durham as ‘a man of the greatest ferocity and bestial character, who ravaged Northumbria miserably with frequent invasions’. In these invasions he was to set a pattern followed by Scottish kings for over 150 years.
The first incursion was in 1061. While Tostig, Siward’s successor, was absent in Rome, Malcolm marched into Northumbria and then probably crossed to the west to recover control of Cumbria south of the Solway. This southern orientation was confirmed by Malcolm’s second marriage. In 1068 he welcomed to his court Edgar Atheling and other English exiles. Some two years later he married Margaret, Edgar’s sister and granddaughter of Edmund Ironside, thus uniting himself with the ancient line of Wessex kings. Margaret was significantly to influence her husband’s policies and to found a line of Scottish kings.
In 1070, in the aftermath of Edgar Atheling’s abortive invasion of England, but after King William himself had disbanded his forces, Malcolm went south again. He consolidated his hold over southern Cumbria and then ravaged Teesdale. It was this invasion (and perhaps also the marriage to Margaret) which provoked William’s one intervention in Scotland when in 1072 he marched to Abernethy. Malcolm ‘gave hostages and was his man’, as the Anglo-Saxon chronicler put it, but he did not remain quiet for long. In 1079, having triumphed in Moray, and with William back in Normandy, he harried Northumbria as far south as the Tyne. William responded by sending his son Robert north. He penetrated as far as Falkirk, secured Malcolm’s renewed submission, and on his way back began the building of a new castle on the Tyne, the beginnings of that great northern city.
The castle was formidable yet it was also a recognition of Malcolm’s power, for it stood not on the Tweed but on the Tyne, some seventy miles further south. As the northern chronicler Simeon of Durham complained, the way remained open for Scottish incursions, hence the need for the castle built at Norham on Tweed in the 1120s. Equally important, indeed related, was the fact that by 1086 the enfeoffments of Norman barons had gone no further north than Mitford, some fourteen miles beyond Newcastle. This highlights the greatest of all the contrasts between Wales and Scotland. There was never any equivalent in the north of the Welsh marcher baronies; there were no Norman lords aggressively pushing into the frontier zones and planting out their castles beyond the Tweed or the Solway. On the contrary the pressure was often from the other side.
Malcolm’s fourth invasion came after the Conqueror’s death in 1091. With Edgar Atheling once more at his court and William Rufus away in Normandy, he penetrated almost to Durham. Later in the year both Rufus and Robert marched north and Malcolm yet again submitted. But the real response came in the following year, when Rufus assembled a large army, expelled Malcolm’s client ruler from Cumbria south of the Solway, and established a castle and town at Carlisle. For Malcolm this was a devastating blow. He was deprived of the great territorial gain of his career and, with Carlisle stopping one end, his access to the primary east–west route through the valleys of the Irthing and the south Tyne. In 1092 Malcolm went south to see Rufus at Gloucester, and was told his status was simply that of an English baron. Not surprisingly, the next year he launched his fifth and last invasion. He was trapped by Earl Robert de Mowbray in Northumbria and killed, together with Edward, his eldest son. Queen Margaret, already ailing, died when she received the news.
Malcolm’s invasions of England, until the last, were superbly timed. Their motives were mixed. Doubtless he hoped to extend his authority over an area south of the Tweed on a permanent footing, just as he tried to do south of the Solway, but he lacked the technology either to acquire the one or retain the other. No castles were built by Malcolm. The invasions were still immensely worthwhile if only for booty, partly in the form of slaves. The Durham source quoted earlier tells of ‘the very many men and women led away as captives’. Queen Margaret herself tried to purchase the freedom of English slaves she found in Scotland. Yet paradoxically she was also one cause of the invasions.
Margaret, thanks to her Wessex descent, was a queen of high status, something Malcolm acknowledged by naming their first four sons after Anglo-Saxon kings. The children themselves enhanced Margaret’s role because she controlled their education and was inevitably involved in what were, as we shall see, the complex politics of the succession. A Life of Margaret, written within a few years of her death almost certainly by her former confessor, Turgot, the prior of Durham (1087–1107), affords a unique glimpse of the personal chemistry between husband and wife, on which so much queenly power depended. Malcolm honoured his wife as more educated and in religious matters more fervent and informed than himself. He gazed at her books which he could not read and jestingly threatened her with punishment when she took his money to give to the poor. Margaret’s fabled piety was perhaps the product of her early upbringing in the only recently converted Hungary, whence she had arrived at the Confessor’s court in 1057. She heard several Masses a day, read and re-read the Psalter, and filled the palace with paupers. She was also concerned with the more general reform of the church.
Margaret venerated some of the Scottish hermits but it seems certain that the church in general needed reform. There had been no wave of monastic revival in the tenth century or later, as there had been in England and Normandy, and apart from a few earlier survivals the nearest approach to monasteries were communities of Culdees (‘vassals of God’), established from Ireland in the ninth century and now largely regarded as the private domains of noble families. As for local churches, these were often grouped together, in a way comparable to the Anglo-Saxon minsters, under Culdee communities and communities of other clerics. But the ministry provided, judging from Margaret’s legislation, was often lax. Bishoprics certainly existed, but long vacancies were apparently customary and neither the numbers of bishoprics nor the diocesan boundaries seem very fixed.
The scope of Margaret’s reforms was limited. The abuses condemned by the councils over which she and her husband presided (neglect of the Sabbath and improper celebration of the Mass, for example) were to be condemned again when Turgot became bishop of St Andrews. Margaret certainly built a ‘noble church at Dunfermline’ to house a Benedictine priory – the first regular Benedictine monastery in Scotland – but this was not followed by a monastic revival. Margaret’s daughters had to be sent for their education to Romsey abbey in England; and there was no prospect of Margaret and Malcolm working hand-in-hand with a great bishop as William did with Lanfranc in England. St Andrews was probably recognized as the senior Scottish bishopric, but it had no metropolitan authority, and in any case Fothad, who held the see between 1059 and 1093, was no reformer.
There was a political dimension to such reforms as there were because they fitted well with the agenda of southern conquest. In 1072 Lanfranc adopted the title of ‘primate of all Britain’ but accepted the authority of the archbishop of York over Durham and northwards ‘to the ultimate limits of Scotland’. Later Scottish kings were determined to maintain the independence of the Scottish church from the pretensions of York and Canterbury, yet there is no sign that this was the agenda of Malcolm and Margaret. Indeed, Margaret sought Lanfranc’s help in founding Dunfermline. Malcolm expected to expand southwards, not to defend himself from any southern threat. Both his violence and Margaret’s reforms might, in different ways, encourage the English to accept his rule. So might their adoration of St Cuthbert at Durham. Malcolm laid the foundation stone of the new cathedral there in 1093 and secured the convent’s prayers in perpetuity for himself and his wife.
Both reform and invasion were related to another consideration, one which dominated the later stages of Malcolm’s reign: the question of the succession. Although the Scots of the MacAlpin realm had long had a single king, and probably sometimes a successor designated in advance (the ‘tanaise’), succession was not by primogeniture. Rather (until 1005) the throne had gone to the member of the royal house who had seemed most suitable or been most powerful. Malcolm II’s succession by his grandson Duncan in 1034 was the first example of the throne’s passage in the direct male line since the mid ninth century. And then Duncan had soon been overthrown by Macbeth. There were now several potential challengers for the succession: Malcolm’s brother Donald Bàn, for example, or an even more likely candidate, Duncan, the son of his first marriage, who had been dispatched to King William as a hostage in 1072 but was still inconveniently alive at the Norman court. Malcolm himself was absolutely determined to be followed by Edward, Margaret’s eldest son, thus grafting the MacAlpin line on to that of the Wessex kings. But Margaret was unpopular. At the councils she had to argue down opposition and this was not helped by her lack of Gaelic; Malcolm himself had to act as her translator. It was absolutely vital therefore to enhance the prestige and power of the new dynasty. The reform of the church, it was hoped, would win the favour of God, if not of man. The southern wars might also contribute to this end – Margaret may have regretted the slaves, but she needed the wealth and the prestige which the expeditions brought. In some remarkable passages Turgot’s Life makes it clear that at court Margaret was the mistress of ceremonial, striving to make ‘the magnificence of royal honour much more magnificent for the king’. Had she seen Queen Edith (after whom she named her eldest daughter) do much the same at the court of the Confessor? Margaret thus adorned herself ‘in costly elegance as befitted a queen’. She decorated the palace with silken cloths, made native courtiers dress in coloured robes, introduced gold and silver vessels to the table, and ‘instituted more ceremonious service of the king’ so that henceforth he rode and walked surrounded by a large retinue. Malcolm certainly had a strong sense of his kingly authority and insisted his subjects swear an oath of allegiance to him. When Turgot refused to do so he was prevented from setting up a monastery at Melrose.
In the end it was all worth it, for the line of Malcolm and Margaret did indeed survive in Scotland. Through the marriage of their daughter to Henry I it became established in England too.