The Iron Age in Northern Britain

Britons and Romans, Natives and Settlers

Dennis W. Harding

• Publisher: Routledge
• Year: 2017
• eBook ISBN: 9781317296492
• Edition: 2
• Topic: Social Sciences
• Subtopic: Archaeology
• Index: Social Sciences

The Iron Age in Northern Britain examines the archaeological evidence for earlier Iron Age communities from the southern Pennines to the Northern and Western Isles and the impact of Roman expansion on local populations, through to the emergence of historically-recorded communities in the post-Roman period. The text has been comprehensively revised and expanded to include new discoveries and to take account of advanced techniques, with many new and updated illustrations. The volume presents a comprehensive picture of the 'long Iron Age', allowing readers to appreciate how perceptions of Iron Age societies have changed significantly in recent years. New material in this second edition also addresses the key issues of social reconstruction, gender, and identity, as well as assessing the impact of developer-funded archaeology on the discipline. Drawing on recent excavation and research and interpreting evidence from key studies across Scotland and northern England, The Iron Age in Northern Britain continues to be an accessible and authoritative study of later prehistory in the region.

Part I
The legacy of the past

1
Introduction

The archaeological framework

The chronological framework

Any regional or chronological synthesis should begin by defining its chronological and spatial parameters. Those parameters need not be regarded as having significance beyond limitations of convenience, though one might prefer wherever possible to adopt meaningful boundaries or chronological horizons to ones that are purely arbitrary. Furthermore, chronological thresholds that might be applicable to one region of Northern Britain will not necessarily or automatically be apposite for other regions, compounding the problems of devising a workable system of classification and terminology.

For Southern Britain, the situation, superficially at any rate, is not so acute: at least there, the Roman conquest serves as terminus ante quem for studies of the Iron Age. It could be argued, of course, that the native response to conquest was integral to an understanding of Iron Age communities, and that Britain under Roman occupation should be regarded as an important sequel to an understanding of pre-Roman Iron Age society. But as a threshold of convenience, AD 43 has at least some historical validity. There is no comparable historical threshold for the Iron Age of Northern Britain. The Romans may have established a permanent frontier across the Tyne-Solway isthmus in the late first and early second centuries, but their occupation of southern Scotland was intermittent, amounting to little more than eighty years in all. Beyond the Forth-Clyde line, Roman influence was still more limited, and in Atlantic Scotland, it was virtually non-existent.

For much of the Atlantic north and west, there is no historically defined break in the sequence of Iron Age settlement from the second half of the first millennium BC until the Norse settlements of the later first millennium AD. Important changes can certainly be observed in the settlement sequence, notably from the ‘monumental’ phase of architecture, represented by brochs or ‘complex Atlantic round-houses’, into the non-monumental period of building represented by ‘cellular’ and related structures. But there is no clear archaeological evidence to associate this progression with any radical change in population or culture. For these regions, therefore, it makes sense archaeologically to consider the ‘long’ Iron Age, in which ‘earlier’ represents a span of time that in Southern Britain would cover the whole of the pre-Roman Iron Age and in which ‘later’ is applied to the first millennium AD from around its second quarter. Some authorities have preferred to adopt the threefold system of early, middle and late Iron Ages, in which the ‘middle’ component for many years constrained the occupation of brochs to a limited span of two or three centuries around the turn of the millennium. Partly because of that artificial constraint on the dating of Atlantic round-houses and their antecedents, the simple division between ‘earlier’ and ‘later’ is here preferred. There is, however, at present no universally accepted scheme of classification, and chronology remains flexible. In any event, with the now-routine application of radiometric dating, and a shift away from the older cultural-historical paradigm, complex classifications are no longer a priority of archaeological synthesis. On these grounds alone, the simplest and most flexible system is probably to be preferred.

In the eastern lowlands of Scotland, a simple earlier-later division of the Iron Age will serve, so long as we recognize that not all changes in field monument or artefact types need have been synchronous and that these changes may still have taken place within an essential framework of continuity. Field monuments such as souterrains, popularly regarded in their classic stone-built form as ‘Pictish’, in fact are largely abandoned in eastern Scotland by the third century, though elsewhere in Scotland and Atlantic Europe they may have a longer currency. Continuity of settlement, despite disruptions, was almost certainly a keystone of the archaeology of southern Scotland and the Borders too. In the post-Roman centuries, these regions were not immune from external influences or even intrusive settlement, but unlike the later Norse settlements, these are not so readily distinguishable, in terms of distinctive settlements and burials or diagnostic material types, from the native communities with which they interacted. Claims may be made for recognizing innovative settlement types among the Anglian settlers of south-eastern Scotland, but in the west, it is generally recognized that the historically recorded Gaelic settlement of Dalriadic Scots is very hard to equate with any innovative class of archaeological evidence, structural or material. In southern Scotland as in northern England, the Roman period may afford a more obvious interface between ‘earlier’ and ‘later’ Iron Ages, but in terms of native settlement, continuity of traditions is still apparent if not dominant.

A starting date for the Iron Age around the seventh or sixth centuries BC is just as arbitrary as any historically derived horizon and could lead to an interminable and ultimately fruitless debate regarding the beginnings of iron technology and when it impacted significantly upon communities in later prehistory. The origins of iron technology in Europe can certainly be traced back to the second millennium BC, with significant occurrences of iron artefacts in the Urnfield culture of the later Bronze Age in Central and Western Europe. For Britain, and more especially for Northern Britain, the incidence of iron artefacts or evidence of ironworking was extremely limited but in the past twenty years has burgeoned from settlement excavation, especially in Scotland. As a label of convenience, the point at which Northern British prehistory becomes ‘Iron Age’ in any meaningful sense hardly matters. But as an indication of the level of control that Iron Age communities could exercise over resources, some locally available but others necessitating longer-distance communications and access to supplies, the presence or absence of metalworking and its products may be an important measure of power or wealth. Students of the Iron Age in Northern Britain habitually refer to the presence on settlement sites of copper or bronzeworking as an index of social status, only very occasionally discussing the role of metalworking or metalworkers within the social hierarchy. Knowledge of iron technology was probably current in Northern Britain from around the mid-first millennium BC or slightly earlier, as in Southern Britain. At Hunterston in Ayrshire at least, its introduction appears to follow in short sequence on from Late Bronze Age metalworking. So the term ‘Iron Age’ here is shorthand for a period of a thousand years and more, from at least the mid-first millennium BC to the later first millennium AD.

The structure of the following chapters, therefore, is based essentially on a broad division between the earlier Iron Age and later Iron Age, the comparative adjectives being used simply to underscore the essential continuity of settlement. Between the two, the Roman interlude should not be the determining factor in classification, important though its impact may have been upon those limited regions of Northern Britain upon which it registered. In most treatments of the ‘long’ Iron Age, an astonishing methodological volte face is performed by archaeologists as they move from the earlier Iron Age to the later Iron Age. Having treated the former in anonymous, sub-Childean cultural terms, they turn without explanation to discuss the later Iron Age communities of Scotland and northern England as Picts, Angles, Britons and Scots in the context of named kings, named sites and dated events, all derived from documentary sources, as if the archaeology had suddenly been relegated to a supporting role in amplifying recorded history. It is not my purpose in challenging this long-established and enduring practice to deny the value of documentary sources, nor indeed of linguistic, numismatic, onomastic or epigraphic sources, in reaching a better understanding of early societies. I do, however, object to the tacit assumption that documentary sources are in some way more reliable than archaeological evidence, particularly where the two might be in apparent conflict, and to the implication that history as a discipline is more rigorous in its methodology than is archaeology. In attempting to articulate its theoretical principles beyond the ‘common sense’ paradigm, archaeology may have engulfed itself in a morass of jargon and some pretentious and vacuous theory, but its basic principles for evaluating artefacts in context have been long established and are as fundamental as those used by historians in evaluating their documents. One of the attractions of archaeology is its inter-disciplinary character, and the relevance of related disciplines, including history, in the social, economic and cognitive reconstruction of past communities should be an asset rather than a liability. But these sources too must be subjected to the methodological scrutiny of the discipline concerned and not simply used as a basis for unsupported assertion, as too often are ethnographic or theoretical models derived from other disciplines when used by archaeologists.

The geographical context

If our chronological parameters necessarily must remain flexible, what about our geographical limits? Since the Act of Union, Northern Britain has been a euphemism for Scotland, but in the present context, that political border makes no archaeological sense, and plainly our remit must also include parts of northern England, at least for the pre-Roman period. Crucially relevant is the territory assigned in documentary records to the Brigantes, whether or not this proves archaeologically to be a meaningful entity. Archaeological distributions show a divide between coin-using societies in southern England and their non-coin-using neighbours to the north and might support a provisional boundary somewhere along the Trent, looping around the southern Pennines towards the Wirral peninsula. It would be absurd to exclude Yorkshire, Durham and Northumberland on the one hand, or Lancashire and Cumbria, however intractable the Iron Age in these regions has hitherto proved, on the other, from any treatment of Northern Britain in the Iron Age. For the period of the Roman occupation, and more especially for any consideration of the question of ‘Romanization’, these regions will also be crucial. For the post-Roman period, on the other hand, it is necessary to impose some limitations, and it is not here proposed to extend into a consideration, for example, of Anglian northern England, except in so far as it intrudes into the native ‘British’ areas of southern Scotland.

In view of the importance of the Irish Sea as a means of communication with other regions of Atlantic Europe, it follows that our study should include the Isle of Man, too frequently neglected in studies of the British Iron Age. A major programme of excavation was achieved during the Second World War by Gerhard Bersu, the excavator of Little Woodbury in the late 1930s and of Traprain Law among other Scottish Iron Age sites in the immediately post-war years, whilst he was interned there as an alien. But the Isle of Man has not featured prominently in British Iron Age studies in more recent years.

North of Hadrian’s Wall Piggott (Piggott, S., 1966) extended Hawkes’ (1959, 1961) scheme of provinces and regions to include Tyne-Forth, Solway-Clyde, North-Eastern and Atlantic Provinces, a model that has proved remarkably resilient. Its most obvious limitation was its apparent presumption that major rivers and estuaries formed cultural boundaries, rather than serving potentially to unite the communities on opposite shores. Whilst this might be challenged in the cases of the Tyne, Forth, Solway and Clyde, it must be acknowledged that the communities who could see each other’s shores across the Moray Firth show archaeologically very little evidence of inter-communication. A second limitation of Piggott’s scheme was that it effectively ignored Perthshire and the significant distribution of crannogs and land-based duns of the central Highlands. The unity or diversity of Atlantic Scotland as a cultural province will also need to be examined.

In fact, we should hardly expect any static scheme of geographical provinces to serve as a means of articulating the dynamic processes of cultural progression over a millennium or more. Necessity demands some division of the material for convenience of discussion, but the option preferred here is drawn from the basic geological and physiographical divisions of Scotland, which, unlike northern England, create broadly diagonal zones across the landmass. The geological zones of the Midland Valley and the Southern Uplands themselves evidently include subtle regional variables which will locally have affected settlement patterns, whilst climatic variables between east and west introduce further factors that will have impinged upon settlement. Over time, too, patterns of human settlement will have been subject to dynamic change for a variety of reasons, not least those imposed by human rather than environmental constraints. Nevertheless, these broad divisions may be preferred in the presentation and evaluation of the archaeological data for want of other, more compelling regional groupings.

The nature of the archaeological evidence

Archaeological evidence can be divided into three broad categories. Artefacts are represented by material remains such as pottery, personal ornaments, weapons or tools. Sites, including fortifications, domestic buildings or tombs, might also be regarded as artefacts on a larger, non-portable scale. A third category, ecofacts, includes environmental evidence of, or relevant to, human occupation. The Iron Age in Northern Britain could be characterized as site-dominated, that is to say that the classification and interpretation of the evidence has been largely based upon field monuments – hillforts, homesteads, brochs, long cist cemeteries or whatever – rather than by artefact assemblages. Accordingly, prehistorians have written about the ‘Castle Complex’, the ‘Hownam Culture’ or the ‘Arras Culture’ but never of a cultural grouping designated by a material type. The point is not simply semantic because the choice of definitive trait also determines the significant thresholds for interpretation. What is taken to matter in Atlantic Scotland, for example, is when monumentality in broch construction gives way to non-monumental cellular building, not when a given style of pottery or metal type is replaced by another, which may not be a coterminous event at all. Environmental archaeology has likewise always been subordinated to a supporting role. No-one has yet proposed a classification of the Northern British Iron Age based on the ‘spelt wheat zone’ as opposed to the ‘barley zone’, though this distinction might be argued as having as great a significance as have differences in structural types.

The relative abundance of settlement sites, however, tends to obscure the fact that those that survive archaeologically may represent only part of the original settlement spectrum. Monumental stone forts and brochs have an obvious durability, as have any settlements with upstanding earthwork enclosures. Even where these have been obliterated by generations of agriculture, earthworks may be traced through air-photography. Yet archaeologically arbitrary development such as road construction continues to expose sites, the ephemeral nature of which has left no trace on the surface or from the air. Discussions of the social hierarchy of brochs, duns and wheelhouses in Atlantic Scotland, or presumptions of a ‘trend towards enclosure’ in the lowlands and Borders, may be based upon only a very partial selection of evidence that happens to have survived. Factors affecting survival may be many, not least the building materials used and the method of construction, but the dice of survival is obviously loaded against sites that have no enclosing works, or which may have been occupied seasonally or temporarily, or with buildings lacking substantial or earth-fast foundations.

Similar strictures could be applied to funerary sites. Archaeologists, vexed by the absence of formal cemeteries for much of the Iron Age in Northern Britain, resort to explanations that would leave no archaeological trace, such as cremation and scattering, as a possible dominant rite. But where there is an identifiable rite, such as the square-ditched barrow cemeteries of eastern Yorkshire, this is assumed to be the norm rather than just one selective mode of disposal. Chariot burials are regarded as high-status tombs, while burials that have minimal grave-goods are tacitly assumed to be of lesser status. Given their numbers, they plainly come closer to a ‘norm’ than most Iron Age burials. But they might still represent a selective section of society, and other rites of disposal like fragmentation and integration into settlement contexts (Harding, 2015) might have been the regular practice for a significant part of the population. In earlier prehistory, long barrows and chambered tombs are widely accepted as a standard means of burial, notwithstanding the fact that the total known numbers could not accommodate more than a fraction of the total population of the Neolithic. If it is accepted that they were a very particular form of élite burial within a wider spectrum of practices then it follows that the norm has still to be identified archaeologically. As ritual and ceremonial monuments, it is possible that disposal of the dead was not even their primary purpose, but just one of the associated rites.

Artefacts too are subject to differential survival, depending upon what they are made of and the matrix and circumstances in which they were deposited. The contrast between the material assemblages from the land-based and underwater excavations of the island dun and its associated structures at Loch Bharabhat in west Lewis (Harding and Dixon, 2000) was striking and instructive. Pottery survived from both land and underwater contexts in broadly equal measure, but it was only in the anaerobic conditions underwater that organic artefacts and materials survived, including a range of domestic utensils that seldom survive on dry-land sites. Perishable materials such as wood, leather or textiles were evidently used as containers and for other purposes, doubtless throughout the Northern British Iron Age. From dry-land contexts at Dun Bharabhat, no bone survived, so that stone artefacts and pottery were virtually the only indicators of domestic culture. In fact, the range and quality of ceramics in the Iron Age of Atlantic Scotland, by contrast to most other regions of Northern Britain and Ireland, which are virtually aceramic, is remarkable. In other parts of the Old World, aceramic cultures are hardly known in societies as advanced as those of the Iron Age, so that the contrast in Northern Britain demands some explanation.

In contrasting the Iron Age material assemblages of the British Isles (and of Ireland) with those of continental Europe, we should remember that the better-preserved and more distinctive continental assemblages are derived in significant part from cemeteries rather than from settlement sites. Not only are the prospects of survival of intact artefacts in graves far greater than in domestic contexts, but it is also possible that grave-goods were in other respects special or different from those in domestic circulation. The apparent poverty of material assemblages from Britain, including Northern Britain, in contrast to those of continental Europe may therefore reflect the contrasting contexts of deposition rather than cultural isolation from the presumed ‘mainstream’ of continental Europe.

Archaeologists by convention order their sites and artefacts into classes or types as the first step in the process of interpretation. There is nothing fundamentally wrong with this procedure, since it enables the evidence to be examined, like with like, and spatial distributions to be compared and contrasted. The problem arises in the subjective assessment of likeness and determining which traits are important in making meaningful comparisons. At the level of defining cultural assemblages, attempts have been made in the past to distinguish between descriptive types, those basic structural or artefactual types that are characteristic of the assemblage under examination, but which may equally be common to several or many different assemblages, and diagnostic types, that is, those that are exclusive to a particular assemblage, and therefore uniquely proclaim its context and date. The latter may appear to be more useful than the former, in that they are strictly definitive, but the problem arises as to what it is that we imagine we have defined. The cultural model developed by Gordon Childe quite expressly assumed that archaeological distributions, defined by recurrent structural or artefactual types, should be coterminous with prehistoric ethnic groups or historically recognized populations. In reality, however, population dynamics and interaction with neighbouring groups over time generally obscure any simplistic patterning, even if the cultural mo...

Table of contents

• Cover
• Title
• Copyright
• CONTENTS
• List of figures
• List of colour plates
• Preface
• PART I The legacy of the past
• PART II The earlier Iron Age
• PART III The Roman Iron Age and its impact
• PART IV The later Iron Age
• PART V Review and conclusion
• Abbreviations
• Bibliography
• Index