The Makers of Scotland E-Book

Tim Clarkson

• Herausgeber: Birlinn
• Kategorie: Geisteswissenschaft
• Sprache: Englisch

Seitenzahl: 443

Veröffentlichungsjahr: 2012

Beschreibung

During the first millennium AD the most northerly part of Britain evolved into the country known today as Scotland. The transition was a long process of social and political change driven by the ambitions of powerful warlords. At first these men were tribal chiefs, Roman generals or rulers of small kingdoms. Later, after the Romans departed, the initiative was seized by dynamic warrior-kings who campaigned far beyond their own borders. Armies of Picts, Scots, Vikings, Britons and Anglo-Saxons fought each other for supremacy. From Lothian to Orkney, from Fife to the Isle of Skye, fierce battles were won and lost. By AD 1000 the political situation had changed for ever. Led by a dynasty of Gaelic-speaking kings the Picts and Scots began to forge a single, unified nation which transcended past enmities. In this book the remarkable story of how ancient North Britain became the medieval kingdom of Scotland is told.

Leseprobe

This eBook edition published in 2012 by Birlinn Limited West Newington House Newington Road Edinburgh EH9 1QSwww.birlinn.co.uk

Copyright © Tim Clarkson 2011

First published in 2011 by John Donald, an imprint of Birlinn Ltd

The moral right of Tim Clarkson to be identified as the author of this work has been asserted by him in accordance with the Copyright, Designs and Patents Act 1988

All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form without the express written permission of the publisher.

ISBN: 978-1-906566-29-6 eBook ISBN: 978-1-907909-01-6

British Library Cataloguing-in-Publication Data

Contents

List of Maps

Introduction

1 BC to AD

2 The Later Roman Period

3 Britons, Picts and Scots

4 Christian Beginnings

5 Celt and Saxon

6 The Struggle for Power

7 The Northern Churches

8 The Vikings

9 Alba

10 Kings and Bishops

11 Overview: The Birth of Medieval Scotland

Appendix A: Genealogies

List of Maps

1 Scotland: mountain, river and sea

2 Roman Scotland: peoples and frontiers

3 Scotland: old and new territorial divisions

4 Kingdoms and kindreds, sixth to eighth centuries AD

5 Major centres of Christianity

6 Centres of power, sixth to eighth centuries AD

7 Scotland in the Viking Age

Scotland: mountain, river and sea

Introduction

Objectives

This book aims to provide a narrative history of Scotland from the eve of the Roman invasion to the final phase of the Viking Age. It encompasses one thousand years of social, political and religious development in which various groups and individuals each played their part. The book’s subtitle names four of these groups – Picts, Romans, Gaels and Vikings – all of whom contributed to the shaping of Scotland’s medieval identity. Two others – the English and the Britons – are intentionally absent from the subtitle. Their omission does not diminish their role in the narrative. Indeed, both peoples were major players in early Scottish history and are accorded due prominence in the following chapters. Omitting them from the subtitle is rather an acknowledgement that their names convey no specific sense of time and place to the reader. By contrast, the mention of ‘Romans’ and ‘Vikings’ indicates the broad chronological context, while the naming of ‘Picts’ and ‘Gaels’ emphasises that this is a book about Scotland’s ancient inhabitants. The narrative begins at the end of the first century BC. in the years when Rome began to take a keen interest in the British Isles. It then follows a chronological path, tracing Scotland’s origins through the ensuing centuries to the battle of Carham-on-Tweed in 1018.

This is not a study of topics or themes. Broad subjects such as warfare and economics are incorporated within the narrative, but the main purpose of the book is to present a linear history. The sole exception to this rule is Christianity, a topic forming the main focus of three chapters which together chart the decline of paganism, the establishment of churches and the increasing role of the clergy in political affairs. Chapter narratives are not punctuated by citations of primary and secondary sources, or by footnotes and endnotes, but the closing pages of the book include suggestions for ‘further reading’ on particular topics. Throughout the book a number of maps indicate the locations of kingdoms, settlements, religious sites and other significant places. Genealogical tables in Appendix A show the kinship between important individuals, while Appendix B provides a chronological summary or ‘timeline’.

Sources

The period covered by this book includes the three and a half centuries of Roman rule in Britain, from AD 43 to c.400, plus a further six hundred years which eventually saw the northern part of the island evolve into the Gaelic-speaking kingdom of Alba. Only a segment of what is now Scotland was brought within the Roman Empire, but the unconquered portion remained a source of trouble for the imperial authorities despite many attempts to subdue it. Military campaigns launched against the northern peoples were documented by Roman writers whose accounts represent a rich vein of data on early Scottish history. After the Empire abandoned Britain in the early fifth century, Roman texts effectively ceased to mention Scotland, their disappearance depriving modern historians of a contemporary record of events. Sparse information relating to the fifth and sixth centuries does exist, but much of it was created retrospectively by writers of a much later era who were less interested in historical accuracy than in reconfiguring the past in ways appropriate to their own purposes. This means that today’s historian must sift each textual source very carefully, in the hope of unearthing fragments of genuine early information among passages written in the twelfth or thirteenth century or even later.

Reliable sources from the first millennium AD are scarce and are usually biased in favour of political or religious interests. A typical example is Bede, a northern English monk whose lifetime spanned the seventh and eighth centuries. Not long before his death in 735 he completed his greatest work, The Ecclesiastical History Of The English People, in which he frequently made reference to the peoples of early Scotland. His book is a valuable contemporary source and a repository of fascinating detail, but, notwithstanding his reputation as a meticulous scholar, the Ecclesiastical History needs to be treated with caution. Bede was not a historian in the modern sense, but a monastic writer who regarded history as the unfolding revelation of events predetermined by God. This meant that he constructed his narrative in ways that allowed him to demonstrate how the English – whom he believed to be a chosen people like the Israelites of the Old Testament – enacted God’s will among themselves and their neighbours. In so far as the peoples of early Scotland were concerned, Bede saw some – such as the Picts – as playing a positive role in the Divine plan, while others – such as the Britons, whom he regarded with contempt – were earmarked by God to suffer well-deserved woes at the hands of English enemies. The Ecclesiastical History is therefore coloured by its author’s biases and preferences and is not the sober textbook it might seem at first glance.

Contemporary sources written in Scotland are few in number. There is no Scottish equivalent of Bede and only a handful of eyewitness or near-contemporary accounts relating to this period have survived from northern Britain. This does not mean that reliable information cannot be found, but the task of identifying it is made more difficult. Our most detailed source is a group of texts known collectively as the Irish annals. Originating as year-by-year entries in records maintained by monasteries, these texts refer not only to ecclesiastical matters but also to events in the secular world. Although the various manuscripts in which the annals survive were written no earlier than the twelfth century, their information was copied from older documents and can be traced back to lost texts compiled many centuries earlier. One of the lost compilations was a set of annals maintained at the Hebridean monastery of Iona, a religious settlement established by the Irish priest Columba in the sixth century. The annals were destroyed when Iona fell prey to Viking raids, but a number of entries were incorporated into the two main Irish compilations – the Annals of Ulster and the Annals of Tigernach – to preserve a valuable account of Scotland’s early medieval history. Iona’s annalists had a keen interest in political affairs and noted important events such as battles, alliances and the deaths of kings. Like Bede, however, they were not immune from bias and crafted their words carefully, selecting events they wished to write about and ignoring others. After the middle of the eighth century, the Iona entries in the Irish annals cease, but Scottish information continued to be recorded in Ireland throughout the Viking period and beyond. Some of this data was spurious or inaccurate, or was added to the annals retrospectively, or was amended much later by scribes when manuscripts were being copied. Rigorous modern analysis of the surviving texts has revealed numerous problems and pitfalls but has enabled the older, more reliable entries to be identified, thereby allowing the Irish annals to be used as a valid source for early Scottish history.

Other types of literature are generally held in less high regard, mainly because of the lateness of their manuscripts, even if they seem to contain material of older origin. Into this category fall the Pictish and Scottish regnal lists – schedules of the reigns of kings – together with a plethora of poems and stories. All of these texts tend to be incomplete, or ambiguous, or found in different versions which contradict one another. Some include fictional characters drawn from legend. Genuine historical information or ‘real’ history is frequently present, awaiting identification and extraction, but it is all too often embedded in impenetrable layers of later, less reliable material. The king-lists of the Picts, for example, are a rich source of information and are often widely used as chronological guides despite posing many questions of their own. Many of the figures named in the early or upper portions of the lists are clearly fictitious. Some historical kings are listed in one or more groups of manuscripts, but are then ignored, perhaps deliberately, in others.

Equally controversial are the vitae or ‘lives’ of saints generated by writers attached to major Christian cult-centres. This type of literature is called hagiography and is encountered in several chapters of this book. Here it will be sufficient to mention that the genre has little in common with modern biography. Even the renowned Abbot Adomnán of Iona, who wrote a vita or ‘Life’ of Saint Columba at the end of the seventh century, cannot be called a biographer, regardless of the fact that he was born only thirty years after Columba’s death. Adomnán was an author of great eloquence and knowledge, but his motives in writing the Life were not those of a modern historian. Like Bede, he viewed past events through the lens of his own biases and prejudices. As Bede’s contemporary he nevertheless provides a useful Gaelic perspective to balance the strongly Anglocentric focus of the Ecclesiastical History.

Scotland’s early history is full of gaps: it is not a complete or accurate record. Basic detail is lacking for many important points, such as the ethnic and geographical origins of certain kings or the locations of ‘lost’ kingdoms. To compound these problems, the information given by the various sources is often contradictory, allowing different interpretations to be drawn. When uncertainty arises, it is not always confined to obscure events or to little-known individuals, as the case of Cináed mac Ailpín (‘Kenneth MacAlpin’) illustrates all too clearly. This enigmatic ninth-century king is examined more closely in Chapter 8, but here, in this brief survey of the sources, he becomes a useful example of just how incomplete the picture can become. Cináed is one of the most famous figures in Scottish history, the king traditionally credited with unifying the Picts and Scots to form a single nation, but he is also one of the most controversial. The problem lies with the sources, which seem to be so unclear about his origins that they fail to answer the most basic questions about him: Who was he? Where did he come from? Was he a Scot or a Pict? Faced with such ambiguity we feel tempted to seek answers in places we might normally avoid, even turning to the bogus ‘histories’ written by John of Fordoun and Walter Bower in, respectively, the fourteenth and fifteenth century. Today, the temptation to consult such works is less compelling than it appeared one hundred years ago, chiefly because Fordoun and Bower are better understood today. Their status as custodians of supposedly genuine ‘tradition’ is now being vigorously challenged.

‘Tradition’ is itself a rather vague term, but it appears from time to time in this book, usually in relation to information of dubious reliability. Folklore and legends, often of localised origin, fall into this category. Most historians are rightly sceptical about the use of such data, even if it appears to be old, but some acknowledge its value as a starting-point for discussion where reliable information is otherwise lacking. A few traditions relating to early Scotland are of great antiquity and seem to preserve nuggets of history concealed among works of poetry or saga. The enigmatic Berchan’s Prophecy, for example, appears to contain many such nuggets and is therefore regarded by historians as a potentially useful source. Although preserved only in a manuscript of 1722, it appears to have been originally composed in the twelfth century from older ‘traditions’ circulating at that time. It is a difficult and controversial source, but among its cryptic verses is a group dealing with Scottish kings of the ninth to eleventh centuries. With careful handling these verses can sometimes be used to add flesh to a sparse entry in the Irish annals, or to an allusion in a twelfth-century chronicle.

Terminology

This book deals with a period before the country now called Scotland came into being. Thus, although the narrative is chiefly concerned with Britain north of the River Tweed and Solway Firth, only a portion of this area was regarded as ethnically or politically ‘Scottish’ during the period studied here. In chronological terms the book traces the history of northern Britain from the end of the Iron Age to the birth of medieval Scotland. This was an era when indigenous peoples such as Picts, Scots and Britons became more or less distinct from one another in the eyes of contemporary writers. Only in the final chapters of the book, during the time of Cináed mac Ailpín and his heirs, does an embryonic ‘Scottish’ identity begin to appear across a broad swathe of northern Britain. Whenever the term ‘Scotland’ is used in the following narrative, the meaning is usually geographical, relating to the physical landscape, rather than political or territorial. The adjective ‘Scottish’ is also used, either in the specific sense of ‘pertaining to the early Scots’ or in reference to abstract concepts such as ‘Scottish history’ or ‘Scottish landscape’. Another adjective is ‘British’ which, like ‘Scottish’, can be used in a specific as well as a general way in early medieval contexts. Historians sometimes use it to describe a particular ethnic group, the Britons, in studies of the period before c.1100. It is used here as a broad geographical term in phrases such as ‘the British Isles’ or in the narrower sense of ‘pertaining to the Britons’. The latter were regarded as an identifiable ethnic group by Roman writers in the first century AD and continued to play a separate political role throughout the period covered by this book.

Most personal names appear here in native, non-Anglicised forms. Domnall and Dyfnwal are therefore preferred to ‘Donald’, while Cináed is preferred to ‘Kenneth’. The exception is the Pictish name Constantin or Causantin which appears in this book in its more familiar form ‘Constantine’. Other Pictish names are Gaelicised, with Óengus, Nechtan and Brude representing the likely Pictish forms Unuist, Naiton and Bridei respectively. This choice of nomenclature might seem obstinately old-fashioned to supporters of the current trend for presenting personal names in the original languages of their bearers. In fact, the choice reflects nothing more than the present author’s own preference. The same can also be said of names borne by Britons, a people whose native language was similar to the ancestor of modern Welsh. Nowadays, these names are increasingly appearing in published works in archaic forms rather than in those found in medieval Welsh literature. Here, the later forms are retained, with names such as Dyfnwal and Owain being preferred to archaic equivalents such as Dumngual and Eugein.

Structure

The following chapters form a continuous chronological narrative spanning the first millennium AD. Each chapter deals with a segment of this chronology. Viewing Scotland’s early history in this way will not be to everyone’s taste, but might suit those who prefer to read history as an unfolding ‘tale of years’. The alternative is the looser chronology offered by chapters devoted to broad themes. As previously stated, this book’s only real concessions to a thematic approach are the three chapters dealing with Christianity, the first of which includes a study of pre-Christian pagan beliefs which the new religion supplanted in the post-Roman period. The book’s final chapter provides an overview of the ten centuries to AD 1000 and looks at how particular aspects of the millennium were perceived in later times. It also considers other issues, such as modern attitudes to Scotland’s archaeological heritage.

Roman Scotland: peoples and frontiers

CHAPTER 1

BC to AD

At the dawn of the Christian era the British Isles were inhabited by people whose society was essentially barbaric and prehistoric. Neither of these labels implies primitiveness or backwardness, despite attempts by Roman writers to portray tribes living outside the Empire as untamed, unsophisticated and – in contrast to Roman decadence – admirably uncorrupted. In describing the ancient inhabitants of Britain and Ireland as ‘barbarians’ we are simply distinguishing them from native communities in mainland Western Europe who had already fallen under the heel of Rome. Barbarian societies were typically those in which large urban settlements and a coin-based economy were absent.

The first millennium AD began with much of Western Europe already in Roman hands. Within the Empire’s borders native cultures were being steadily eroded by a deliberate process of Romanisation. This meant that in areas such as Gaul (roughly coextensive with modern France) the distinctively Celtic character of indigenous society was giving way to the Latin culture of the conquerors. One major casualty was the ancient Gaulish language – a branch of the same linguistic group to which the tongue of the Britons belonged – which faced extinction after being replaced by Latin in all important forms of communication. Thus, at the beginning of the first millennium, the still-unconquered British Isles represented the last bastion of Celtic language and culture.

In those days the northern part of Britain was not yet called Scotland, nor was the southern part called England. The whole island was regarded by Roman travellers and other contemporary observers as a single geographical entity called Britannia. All of its inhabitants, regardless of whether they lived on the southern coast or in the far northern isles, were known collectively as ‘Britons’. Their language – usually referred to by modern scholars as Brittonic or Brythonic – was the ancestor of Welsh, Cornish and Breton. Like Gaulish, it formed part of the ‘P-Celtic’ linguistic group which had already started to diverge from the Goidelic or ‘Q-Celtic’ group when the first millennium commenced. Q-Celtic includes the Gaelic languages of Ireland, Scotland and the Isle of Man, all of which derive from a single ancestral tongue once spoken in many shorelands around the northern waters of the Irish Sea.

Giving any group of ancient peoples the label ‘Celtic’ is not an exercise in precision but rather a convenient way to identify them as non-Roman, non-Germanic inhabitants of north-western Europe. The Celts were not in fact a homogeneous ‘race’, but a myriad collection of communities linked by similarities in culture and language. Most of the ancient people now described as ‘Celts’ would have been puzzled to find themselves lumped together in modern history books as if they were all members of a single ethnic group. The idea of a common Celtic identity is actually a fairly modern concept promoted by nineteenth-century historians searching for an umbrella term to encompass large areas of Europe – including the British Isles – which appeared to share features of a common material culture in the first millennium BC. The name ‘Celts’ was borrowed from the Celtae or Keltoi, a people inhabiting parts of south-western Europe when the region was conquered by Rome in the first century BC. These folk were neither the creators of Celtic culture nor, in ethnological terms, the ancestors of other Celtic nations. Their name was merely chosen as a convenient label for a prehistoric North European culture first identified in the nineteenth century at archaeological sites in Switzerland and Austria. A more accurate use of the label ‘Celtic’ restricts its application to a distinct sub-group of the Indo-European family of languages. It is less appropriate to give it a cultural or ethnic dimension. In this book, it is therefore used as an all-embracing term for speakers of a Celtic language, regardless of whether they lived in Gaul, the Iberian Peninsula or the British Isles. In the narrower context of early Scottish history, ‘Celtic’ is often used by historians to distinguish the natives of northern Britain from the Germanic immigrants with whom they came into contact during the first millennium AD.

Settlements

Celtic language and culture were well established in the British Isles and in parts of mainland Europe long before Rome grew powerful enough to build her empire. Historians and archaeologists formerly believed that a Europe-wide process of Celticisation occurred during the middle of the first millennium BC as part of the transition between the Bronze and Iron Ages. Iron was certainly replacing bronze as the main material for tools after 600 BC and its use in everyday life is rightly seen as an indicator of Celtic influence, primarily because the ‘Celts’ were the first ironworkers in north-western Europe. Changes in the design and location of dwellings in Britain also occurred around this time and have often been attributed to the spread of Celtic fashions from Gaul. In particular, the fortification of isolated hilltops is strongly associated with the Celts and many examples are found on the European continent. The presence of similar hillforts in Britain and Ireland is thus seen as further evidence of Celticisation crossing from Gaul to permeate Insular (i.e. British and Irish) society. Conventional wisdom rooted in past scholarship has tended to explain these social and cultural changes in terms of a mass migration by Celtic groups from mainland Europe into the British Isles after 700 BC. This has recently been challenged by an alternative explanation which, rather than imagining waves of seaborne immigrants, suggests instead that the Celtic languages and cultures of Britain and Ireland were largely home-grown. In this scenario the idea of new linguistic and cultural influences arriving from outside gives way to a simpler theory which sees the peoples of the British Isles and Continental Europe developing a shared ‘Celticness’ as a natural evolution of their common Indo-European origins. Any migration from the European mainland in the first millennium BC would probably have involved small numbers of people – such as groups of ambitious Gaulish warriors – seizing power in certain areas of Britain and Ireland, from where they perhaps disseminated ‘foreign’ influences among the natives. This would mean that the ancient Celtic peoples of the British Isles were not newcomers from elsewhere but the descendants of an indigenous population whose ancestry lay in the pre-Celtic past and whose ultimate origins reached back into the Stone Age.

If Britain and Ireland became ‘Celtic’ without the need for a mass immigration of European Celts in the first millennium BC, then other factors must account for the changes that make their societies recognisably ‘Celtic’ to modern historians and archaeologists. The practice of constructing compact, defensible hilltop fortresses can thus be explained not as a fashion imported from Gaul, but as a Europe-wide phenomenon arising from a need felt by some communities to make their homes defensible against attack. The likeliest cause of such insecurity was a perception that society was becoming more dangerous or, to put it another way, that some communities were behaving more aggressively towards their neighbours. The first hillforts in Scotland were constructed in the Bronze Age, around the beginning of the first millennium BC, and were usually larger in area and fewer in number than their later Iron Age counterparts. Ramparts at some of the smaller forts nevertheless enclosed substantial surface areas containing many houses, giving these sites the character of ‘protected villages’. Other sites were even smaller, with only a few dwellings inside, but would have seemed more imposing to onlookers when located in positions of natural strength. The people who commissioned the construction of hillforts and other enclosed settlements were clearly capable of mobilising large labour forces. They were either powerful kin-groups wielding authority at local level, or entire communities undertaking co-operative projects. Earth, stone and timber provided the raw materials for rampart construction, but the actual designs varied considerably. The basic shape could be curvilinear or rectilinear, while the enclosing rampart might be a simple wooden palisade or a massive earthen bank. In some cases a palisade replaced an earlier earthwork, the former perhaps seeming – to modern eyes, at least – a less substantial, less effective type of defensive feature. A more elaborate technique involved adding timber to unmortared (drystone) walls, either by bracing with long wooden beams or by making a box-like framework. In Scotland, the distribution of these ‘timber-laced’ forts covers a wide geographical area, from Broxmouth in Lothian to Cullykhan, Craig Phadrig and Burghead in the north-east. Some were destroyed by fire, presumably at enemy hands, with temperatures rising high enough to fuse stone and timber together. The resulting process, known as vitrification, left the affected parts of the ramparts with a glassy appearance which so reduced their strength that a burned fort became unsuitable for reuse by its former occupants. Aside from such cases, many hillforts fell out of use within a few centuries of their construction, either through destruction or abandonment. Others continued to be occupied, with or without occasional breaks in habitation, into the first millennium AD. In a few cases the period of occupation was remarkably long and, as subsequent chapters of this book will show, a small number of northern hillforts were still being used as late as the Viking Age.

The larger hillforts constructed around 1000 BC were essentially substantial villages enclosed by ramparts. Their usefulness in military terms was minimal, mainly because they were too large for their inhabitants to defend. There were never many of them and most were abandoned when a trend for smaller forts began to gather pace in the final centuries BC. Some large sites nevertheless continued in use, serving Iron Age communities as major centres and perhaps assuming the role of local ‘capitals’. Their size added prestige to their occupants, even at those sites where an exposed or elevated position made habitation seasonal rather than permanent. When the Romans encountered such places in Gaul they frequently described them as oppida, a term found in Julius Caesar’s account of his Gallic campaigns. One of Caesar’s most celebrated victories came after a prolonged siege of an oppidum at Alesia, a site defended by strong earthworks similar in appearance to those at the major power centres of Celtic Britain. The impressive remains of British oppida can still be seen today, most notably in southern England at places like Maiden Castle and Cadbury Castle. Hillforts in Scotland were generally smaller, with even the largest barely matching the southern examples in size. This has not deterred archaeologists from applying the label oppidum to sites such as Eildon Hill North and Traprain Law, both of which are situated in the Lowlands. Indeed, it is possible that these two settlements served the kind of proto-urban function commonly associated with their Gallic and southern British counterparts. Whether this involved permanent habitation rather than seasonal or occasional ceremonial use is hard to discern from the archaeological evidence. Both were probably first inhabited in the Bronze Age, around 1000 BC. In each case, the earthworks enclosed a large area containing the houses of a substantial population. Settlement seems to have been continuous throughout much of the Iron Age, with a marked decline in activity in the first century BC followed by a period of recovery. Eildon was finally abandoned during Roman times but Traprain had a rather longer existence, its defences perhaps being refurbished as late as the fifth century AD.

Why were hillforts built? To this question there are no simple answers. Reasons for enclosing elevated settlements with ramparts and ditches were probably as varied as the number of sites. Some hillforts were surely constructed as protection from enemies, presumably at times of local insecurity when communities within particular districts faced real perils. The plethora of forts in the Scottish Lowlands might therefore suggest that this region was especially dangerous in the Iron Age, although the proximity of many enclosed sites to their nearest neighbours casts doubt on this explanation. Factors other than warfare and raiding could have been at work. A hillfort in a prominent location was an imposing feature in the landscape and may have served its occupants as a symbol of their group identity, or as a forceful marker of their territory. It is worthy of note that few hillforts in Scotland were still inhabited in the first century AD when the Romans invaded Britain. Perhaps, after fulfilling a range of purposes for many hundreds of years, their usefulness dwindled?

Hillforts appeared all over Celtic Britain and Ireland during the first millennium BC, but they were not the only defensible settlements constructed at this time. One type of dwelling found exclusively in Scotland is the ‘Atlantic roundhouse’, a class of stone-built structure which includes the distinctive brochs and duns. The type as a whole is most distinctively represented by the broch, a huge stone tower with walls so thick that chambers and stairways could be accommodated within them. A low door, invariably less than the height of an average person, provided the sole point of entry. The tower’s interior apparently supported one or more upper storeys, the topmost of which might be roofed with timber or open to the sky. Ruined brochs, some little more than traces on the ground, can be seen today in places as far afield as Shetland and Lothian. Their main concentration, however, is in the far northern Highlands and Western Isles. They were constructed during a period spanning, very approximately, the years 500 BC to AD 100. Why they were built is not known, but their enormous strength suggests that they were intended to present an impression of power. Whether they served as houses for high-status families, or as temporary refuges for entire communities in times of peril, is likewise an unanswered question. At some point in the early centuries AD their original or primary functions – whatever these were – evidently became redundant. Many brochs were subsequently abandoned or partly demolished, presumably because they no longer had relevance for the descendants of their builders. One of the most northerly examples, a coastal site at Jarlshof in Shetland, ceased to be inhabited around AD 200, at which point three round stone dwellings known as ‘wheelhouses’ were erected inside it and around it. The surrounding walls of the courtyard continued to offer protection long after the broch tower ceased to be a dwelling. Similar continuity occurred at other sites where, in some cases, a derelict broch became the central feature or landmark of a later village. The broch at Old Scatness on Shetland, built in the mid-first millennium BC, later become the focus of a surrounding settlement which was still occupied in Viking times.

Smaller than the broch was the dun, another type of stone-built ‘roundhouse’. Duns are found especially in Argyll and the Inner Hebrides and, in most cases, are of comparatively simple design. A typical dun consisted of a round stone wall enclosing a small area which was either roofed or open. Outside this very broad generalisation a great variety in shape, size and setting makes duns difficult for archaeologists to classify as a single settlement type. For instance, the distinction between a dun and a circular stone-walled house is frequently a matter of interpretation. The main period of dun construction straddles the later centuries BC and the early centuries AD. Occupants may have been single extended families belonging to a fairly prosperous tier of society, but their reasons for building and occupying stone-walled roundhouses probably varied widely. To what extent they regarded duns as defensive structures rather than as physical statements of land ownership or social rank is therefore impossible to ascertain. Factors influencing the dun-dwellers’ choice of habitation might, however, have been broadly similar to those that prompted other folk to build artificial islands on stretches of inland water. Just as hilltops appealed to many of the dun-builders, so lochs, rivers and estuaries attracted the attentions of others. Homesteads founded on man-made islands are found all over Scotland and are not confined to narrow periods of construction or habitation. Archaeologists call them ‘crannogs’, treating them as a single category of settlement regardless of their settings on rivers, sea firths or inland lochs. The oldest examples date from Neolithic times; the latest were occupied in the first millennium AD. Within such a wide chronological span it is inevitable that there are many different types but, like the duns, the generally compact size of crannogs suggests that each accommodated a single family. The most common type was a small island connected to the shore of a loch by a wooden causeway. A fine reconstruction of such a crannog can be seen and visited today, on the southern side of Loch Tay, close to the remains of actual examples from the Iron Age. It is clear from the reconstruction that these sites, despite their inaccessibility to land-based enemies, would have been easy prey to a sustained assault.

A simple deduction might be that the lower classes of Iron Age Britain lived in humbler abodes than their social superiors. This need not necessarily have been true in all areas, especially if imposing structures such as brochs were built and occupied by entire communities rather than by high-status families. Identifying houses of less prosperous character is, in any case, a difficult archaeological exercise, particularly when we consider that the excavated remains of both a simple cottage and a stone-walled barn might look identical. It seems clear, nonetheless, that the ubiquitous type of dwelling was the family homestead, usually circular or curvilinear in form, with walls constructed from timber or stone or a combination of the two. Some were scattered across the countryside as isolated farms, while others clustered together in small villages. Towards the end of the first millennium BC and in the early centuries AD some communities in the British Isles began to build underground structures, perhaps as storage for surplus agricultural produce. In archaeological terminology they are called souterrains, a French word describing their subterranean character. They were dug into the ground next to established homesteads and carefully lined with stone to create rooms and passages. Most were roofed with slabs or timbers, while some of the larger examples had additional sub-structures attached. Souterrains are found not only in what is now Scotland but also in Ireland, Cornwall and Brittany, their wide geographical distribution yielding a commensurate variety of designs. In Scotland they ceased to be used after the second century AD, for reasons that remain unclear. Their associated settlements frequently continued in use, so redundancy was not due to population decline or movement. Perhaps the descendants of the souterrain-builders ceased to produce enough agricultural surplus to justify the cost of maintenance?

Society and Culture

Barbarian society in Europe at the end of the first millennium BC was already shifting away from the egalitarian, co-operative systems of earlier times to a more hierarchical structure. This was not, however, a uniform process. Different groups tend to develop in different ways when separated from one another by geography, so some regions were quicker than others in moving towards social stratification. This was true of Celtic and Germanic peoples alike, with communities in some parts of Europe amalgamating into ‘tribes’, while others retained separate identities within a more localised network of allegiances. The resulting diversity in development means that the barbarian peoples of Gaul, Germany and the British Isles on the eve of their respective encounters with Rome presented a variety of political structures. In southern parts of Britain, where large hillforts almost certainly functioned as oppida or tribal capitals, society was more likely to have exhibited a well-defined hierarchy based on wealth and status. Northern Britain – including Scotland – was a region where forts and other enclosed settlements were generally smaller and less obviously ‘aristocratic’. Thus, it is impossible to distinguish in terms of social class the occupants of a small, fortified hilltop in Lothian from those of a Shetland broch. Nor can we be sure that the inhabitants of either of these sites possessed more wealth or greater status than the owners of a large drystone farmstead with associated souterrains. In those areas of northern Britain where social stratification and political centralisation had perhaps become manifest in the late Iron Age, power within a ‘tribe’ may have been wielded by one or more dominant kindreds whose claims to authority derived from ownership of land. Similar claims by rival kindreds undoubtedly sparked aggressive competition for territory, and this would have led to inter-tribal warfare. Roman sources seem to hint that hostility on this scale was not uncommon in southern Britain but it can barely be surmised for the North. Indeed, it is difficult to imagine the numerous small hillforts in the Scottish Lowlands being used as strongholds by what would have been a veritable plethora of Iron Age warlords. Perhaps there were occasional violent contests between communities in this region in pre-Roman times, but the scatter of forts suggests minimal political development and a resistance to centralisation. It therefore seems highly unlikely that any northern part of Britain had attained a level of development in which the topmost social tier was not merely a class of major landowners but a single individual, a paramount chieftain or king. Rome acknowledged the presence of kings among the large tribal amalgams of southern Britain, but there is no evidence that any persons of this rank existed in Iron Age Scotland. Archaeological evidence north of a line drawn between Tyne and Solway seems rather to argue against, rather than in favour of, a rigid social hierarchy and a shift towards political centralisation. This is not to wholly deny the existence of a wealthy, landowning, warrior-nobility in the North at the end of the first millennium BC. Nor should we presume that powerful leaders did not arise among the hillfort inhabitants and broch-dwellers around the time of Rome’s invasion of Britain in AD 43. The defiant northern warriors whom Roman forces subsequently encountered probably looked and behaved, outwardly at least, in a similar manner to the Gauls who had stood against Julius Caesar in the previous century. What we are unable to discern is their level of organisation, their social class, their group identities and their patterns of allegiance. Some of their leaders may have aspired to kingship, like their southern peers, but of this we cannot be certain.

At the lowest level of barbarian society stood those who lacked not only land but liberty as well. These were the unfree – primarily slaves – whose freedom had been forfeit from birth or through later misfortune. Above this tier was the freeman or ‘free farmer’, an individual whose social rank derived from land ownership. It may be that the more prosperous freemen in North Britain – those who owned more land than their neighbours – represented the class most closely associated with brochs, duns and crannogs. They presumably had opportunities for social advancement in districts where a hierarchy was already well established, or where upward mobility to positions of power was becoming an increasingly important aspiration. A poorer freeman owned less land and would have been at risk of losing his liberty altogether, especially if his smallholding was unable to sustain his family during hard times. In such circumstances, he might be forced to give up his freedom by becoming the ‘semi-free’ tenant of a wealthier neighbour. If large numbers of semi-free farmers were indeed present in the Iron Age, it is possible that they were the antecedents of the servile ‘bondsmen’ who seem to have constituted an agricultural peasantry in many Celtic regions of the British Isles in early medieval times.

The economy of pre-Roman Britain was based on agriculture, supplemented by hunting, fishing and specialised crafts such as pottery and metalworking. This was true of all regions, including even the North with its hilly terrain and poorer soils. It was also broadly true of all periods from the Bronze Age, through the Iron Age and Roman era, to the end of the Middle Ages and the dawn of industrialisation. Much of the Scottish landscape in prehistoric times was therefore a picture of fairly stable agriculture. An alternative image of rugged mountains and densely forested glens inhabited by scattered groups of untamed tribesmen persists today, but can be traced back to the prejudices of Roman writers. To Roman eyes it was easy to regard the remote northern parts of Britain as an archetypal wilderness beyond the reach of Classical civilisation. The region north of the Forth–Clyde isthmus was duly presented in Roman literature as a barbarous, chaotic and dangerous place. Such propaganda served a useful literary purpose by drawing a stark contrast with the civilised, Romanised, Latin-speaking provinces around the Mediterranean. The reality, of course, was quite different. Northern Britain on the eve of the Roman invasion was a country of tamed farmlands, the soils of which had been intensively cultivated for more than a thousand years. Highland areas were mostly wild and bare, but so were other mountainous regions of Europe where poor drainage and unproductive soils presented challenges to agriculture. In the lowlands and river valleys the ancient peoples of Scotland had long inhabited a cultivated landscape as populous as any in the British Isles. They grew barley, wheat and oats in small fields; they tended herds of cattle and pigs and, to a lesser extent, flocks of sheep; they levelled tracts of forest in the uplands to create additional land for grazing. So intensively cultivated were the valleys and other low-lying areas of the ancient Scottish landscape that, in the three centuries before the Roman invasion, space for new farms became scarce. Settlements began to appear on the higher land as people established homes on the hillsides to exploit treeless slopes for crop-growing.

Beyond the local agricultural base a web of trading networks linked all parts of the British Isles with one another and with Continental Europe. Most of these links were established in the Bronze Age or even earlier and provided the main channels of cultural interaction between the various Celtic peoples. It was via these same trade-routes that the distinctive artistic and metalworking techniques of the ‘La Tène’ style eventually spread to Gaul, Spain and the British Isles. La Tène is a Swiss village where, in the mid-nineteenth century, an impressive assemblage of ornate items was discovered. These displayed a distinctive style of decoration, characterised by spirals and curving shapes, which archaeologists now regard as the first flowering of Celtic art. By the end of the first century BC many regions further afield shared elements of this common ‘Celtic’ culture. In northern Britain the wealthiest members of society at this time advertised their high status through weapons, armour and domestic objects adorned with circular patterns and animal designs characteristic of the La Tène style. Some of these items arrived from Gaul, Ireland or southern Britain via seaways, rivers and ancient land-routes. Others were manufactured locally by talented artisans using skills first developed in Central Europe and transmitted to all corners of the Celtic world. By the beginning of the first century AD, the upper classes of northern Britain were actively participating in a sophisticated system of long-distance trade in exotic goods with other Celtic elites, and with the Roman Empire too. Prosperous landowners in Orkney, Perthshire, Moray and the Hebrides exploited trading networks to import high-status items such as jewellery for their own personal use. To pay for these expensive treasures they exported the home-produced commodities for which Celtic Britain was renowned: furs, skins, hunting dogs and slaves.

Like their neighbours in Ireland and Gaul, the people of Iron Age Scotland generated no documents of their own. They were not illiterate but rather pre-literate: their society functioned well enough without the need for written communication. Knowledge was preserved and transmitted orally, passing by word of mouth from generation to generation. The customs, traditions and history of each community were disseminated informally as folktales via the medium of storytelling. In addition to this informal transmission of lore, a measure of knowledge control may have been exercised by tribal elites seeking to present particular views of the past. For example, the oral declaration of a headman’s genealogy at public gatherings could have been one way of reinforcing his status within the community, especially if the alleged ancestors included local gods, otherworld figures and ancient heroes. Large public events probably incorporated religious rituals presided over by high-priests who were themselves members of an elite class and whose presence bestowed a sacred aura on the authority of a headman or chief. The venues chosen for such ceremonies would have included stone circles, monoliths and other monuments of antiquity, all of which – in the eyes of a superstitious populace – endowed the proceedings with the approval of revered forebears. In Chapter 4 the role of the pagan priesthood and the religious beliefs that sustained it will be examined more fully.

Conquest

In 56 BC, a Roman fleet commanded by Julius Caesar defeated the Veneti people of northern Gaul in a decisive naval encounter at Morbihan Bay, off the Atlantic coast of what is now north-western France. This battle finally brought to an end Caesar’s Gallic wars, a series of hard-fought campaigns which he later described in a detailed account. His victory at Morbihan had an additional significance: it brought the Roman military closer to Britain, a land whose people shared the Celtic culture of the newly conquered Gauls. To Caesar and his henchmen the British Isles were not a wild, windswept archipelago lurking on the edge of the known world, but an offshore corner of the European continent and a potentially profitable addition to the Roman economy. The coast of Britain was separated from mainland Europe by nothing more than a narrow channel of water, across which Caesar’s Gaulish foes had frequently received moral and material support from British sympathisers. Rome could no longer allow the Britons to lurk on the edge of the civilised world as a threat to the stability of newly conquered Gaul. It was only a matter of time before her military commanders devised a plan for invasion.

In 55 BC, and again in the following year, Caesar himself led minor expeditions across the Channel. His troops clashed with Britons on both occasions, but these encounters were not intended as a prelude to conquest. It was not until AD 43, during the reign of the emperor Claudius, that a full-scale invasion was launched. Four experienced legions – the Second, Ninth, Fourteenth and Twentieth – were selected to spearhead a strike force of 40,000 men. Crossing the Channel from Gaul, the invaders quickly assimilated the south-eastern tribes, among whom some surrendered without putting up much of a fight. Other tribes entered into treaties by which they became client states of the Empire under pro-Roman rulers. Resistance elsewhere was mercilessly swept aside: communities who tried to make a brave stand against the legions were subjugated by force and earmarked for Romanisation. In some areas, however, the Britons refused to give up their independence so easily. One defiant figure from this period was Caratacus, a chieftain of the Catuvellauni. The heartland of his people lay north of the Thames and had fallen to Rome within a few years of the invasion. Fleeing westward to Wales, Caratacus continued the fight by leading the Silures and Ordovices of Wales until his defeat by Roman forces in 51. Ten years later, in what is now Norfolk, the conquered Iceni rose in revolt under their warrior-queen Boudica. After Boudica and her immense army were vanquished by the Roman governor Suetonius Paulinus in 61, the southern Britons fell into line and never again rebelled against their conquerors. The Brigantes, whose territory encompassed much of northern England east of the Pennines, initially showed friendship to Rome before turning hostile after 69. Their name means ‘High Ones’, perhaps in the broad sense of ‘Highlanders’. It seems to be descriptive of their geographical situation rather than of their political unity, so we can probably envisage several ‘Brigantian’ tribal groups, each with its own identity and its own independent leadership. A faction among one of these groups rejected the pro-Roman sympathies of a queen called Cartimandua and rose in revolt around her former consort. The ensuing civil war gave Rome an opportunity to intervene on Cartimandua’s behalf and led ultimately to the conquest of the entire Brigantian zone. The emperor Vespasian, an ex-soldier himself, entrusted the campaign to Petillius Cerialis, the Roman governor of Britain and an experienced tamer of insurgents. By 73, Cerialis had crushed the resistance and brought all or most of the Brigantes to heel. Pushing northward and westward as far as eastern Dumfriesshire, he imposed Rome’s authority as far as the Solway Firth and placed a unit of troops at a new fort in Carlisle. This extension of Roman power northward to the Tyne and Solway brought the Empire face to face with the peoples of ancient Scotland for the first time.

In 77, when the governorship of Britain passed to Gnaeus Julius Agricola, only a handful of Britons remained in revolt in isolated pockets of conquered territory. Almost the entire area of what is now England, as well as large parts of Wales, lay under Roman rule. The next set of imperial objectives was clear: consolidation of recent territorial gains, destruction of lingering troublemakers, and a further northward drive. In the person of Agricola the military authorities had no better candidate to complete these tasks. His achievements in Britain were recorded by his son-in-law, Cornelius Tacitus, in a book bearing the simple title Agricola. Through the eyes of Tacitus, present-day historians are able to gain a valuable insight into how the Roman conquest of northern Britain was achieved. For Scottish historians in particular his narrative provides a unique window on their country’s ancient past. Care and caution should nevertheless be applied when reading Agricola, despite its author’s proximity to the events he describes. As a contemporary of the events and as a member of the Roman elite, Tacitus is obviously an important source, but he was rather too close to his subject to give a balanced account. Being married to Agricola’s daughter undoubtedly gave him a unique perspective, but an intense admiration of his father-in-law turned his narrative into a gushing eulogy. It is therefore through a rose-tinted lens that the modern reader must view this unique and valuable source of early Scottish history.

Agricola had prior experience of the Britons from his time as a junior officer during Boudica’s revolt in 61. He had also served as commander of the Twentieth Legion in the Brigantian civil war ten years later. Campaigns against rebellious tribes had taught him much about strategy and leadership as well as giving him an insight into native military organisation. When he became governor of Britain, he lost little time in drawing on the skills acquired during his youth by launching his first campaign in the same year. His targets were the troublesome Ordovices of North Wales. They were swiftly brought to heel and absorbed. In the following year he marched north to consolidate the Empire’s grip on the Brigantes, subduing them by constructing forts and roads across their territory. The farthest limit of ‘Brigantia’ lay between the Solway Firth in the east and the Cheviot Hills in the west. Beyond these frontier districts lay uncharted lands inhabited by other peoples whom Rome was soon to meet.

Agricola’s Northern Campaigns

Tacitus does not identify the tribes who dwelt north of the Brigantes. In the following century the Britons of the region between the Tyne–Solway and Forth–Clyde isthmuses were perceived by Roman geographers as being grouped into four large amalgamations: the Damnonii, Votadini, Selgovae and Novantae. Whether these four already existed in Agricola’s time, or whether they were formed in response to his campaigns, Tacitus does not say. All that can be deduced is that the people of this region were first subdued by Rome in 78 – when Brigantian territory was finally conquered – or in 79 when Agricola marched north to the River Tay. The latter campaign gave Roman troops their first sight of the untamed highlands of northern Britain, but Agricola halted his advance at the Tay estuary. In the next season of summer campaigning he consolidated earlier gains in the lowlands south of the Forth and Clyde, using the narrow isthmus between the firths as a natural frontier and guarding it with a chain of forts. All territory south of this line was regarded as part of the Empire. It represented the most northerly portion of the new province of Britannia