3
Questioning fieldwork
The archaeological study of London is propelled by discovery, where the urgency of collection dominates research practices that remain true to the inductive and empirical values that first inspired antiquarian survey. Busy excavators tend to assume that we can afford to postpone interpretation until some happy future of greater leisure and more comprehensive knowledge. As a consequence, the occasional advance of understanding can seem the fortuitous consequence of accidental encounters with the past rather than the product of purposive research. The fieldworker’s dependency on the serendipity of discovery is sometimes seen as intellectually sterile by an academic community that renounced inductive approaches in favour of more theoretically aware exercises in hypothesis building and testing. The Roman archaeologist and philosopher R.G. Collingwood was an early critic of ‘blind digging’ undertaken in the hope of discovery rather than to study problems, recommending a theory-informed ‘question and answer’ approach. 1 This is the object of Karl Popper’s refutationist form of scientific inquiry: where theories are advanced as stepping stones towards better theories, in which new understanding is won by testing ideas to destruction, refinement or acceptance. 2 Popper’s critical concept of ‘falsification’, or the ability to disprove, makes the primary hypothesis a vital part of research, since arguments that are not vulnerable to error cannot be tested. If we accept this position unreservedly, then data-driven or ‘inductive’ philosophies self-evidently lack validity. 3
This puts rescue archaeology at a disadvantage. It is unavoidably the case that opportunities to excavate in the City of London are the product of development decisions made with scant regard to the needs of archaeological sampling. Even where the response to such opportunity is guided by research questions these tend to be concerned with finding things rather than understanding them. The continuing expectation is that knowledge is won by gathering facts, from which generalizing explanations might later be drawn. This remains an inductive process.
Matters are further complicated by the way in which the post-war growth of urban rescue archaeology inspired the development of new techniques designed to improve the accuracy of stratigraphic recording on open-area excavations. The recruitment of a burgeoning army of dedicated field archaeologists stimulated the emergence of what Richard Bradley has described as a craft tradition, where the goal of methodological consistency within an increasingly technical profession was served by an architecture of codified recording systems that separated interpretation from description. 4 Ideas explored in field projects of the early 1970s, most notably by Philip Barker at Wroxeter, were drawn together by the Museum of London’s Department of Urban Archaeology (DUA) in landmark excavations of Roman sites in Milk Street and Newgate Street, where the contribution of Steve Roskams was particularly influential. 5 The DUA system codified on these sites requires each digging archaeologist to separately identify, record and excavate every unique ‘stratigraphic unit’ or ‘context’ encountered. These contexts include the physical remains of past buildings, cut features such as ditches, pits and wells, as well as erosional and other sedimentary layers. All descriptions of these contexts—their attributes, components and spatial coordinates—are recorded on standardized pro-forma sheets cross-referenced to scale plans, and located in stratigraphic order as components of a sequence diagram (or ‘Harris matrix’). A complex archaeological site contains tens of thousands of these stratigraphic units, generating an equivalent number of written descriptions, drawings, and catalogued finds assemblages. The order of stratigraphic deposition presents a relative chronology, from which the wider reconstruction of phases of change can be attempted. These activities combine to witness changing uses of urban space, whilst the stratified rubbish describes changing patterns of consumption and discard.
When a large workforce is engaged in recording archaeological stratigraphy in this fashion the process of description risks becoming unwieldy. Engineering considerations can reduce a single archaeological site into dozens of different areas, each dug to a separate timetable. Sites are consequently explored as a host of disconnected fragments, not viewed as coherent archaeological horizons. Large projects will involve fifty or more professional archaeologists recording islands of isolated stratigraphy, working at different speeds in different areas. The traces of a single past event, such as the demolition of a wall, might be encountered in numerous places recorded by different people at different times. Each record might reach different conclusions as to what the evidence meant within its local context. This complicates the task of reassembling the records into a single account, which exercise invariably awaits ‘post-excavation’ study when specialist scientific reports on the finds recovered are also available. The problem with a system that assumes that post-excavation data will improve understanding, as it so often does, is that it encourages an over-worked field team to marginalize the role of interpretation, replacing strategic reasoning with routine collection.
We therefore find that at the very time that university-based research turned against earlier positivist and empiricist traditions, fieldworkers on development-led projects became more entrenched in such practices. The tendency to privilege the documentation and recovery of material evidence as an end in itself was reinforced by the urgency of rescue archaeology and the emphasis placed on the chimerical concept of ‘preservation by record’. 6 The planning regime put in place to protect buried archaeological remains was aimed at their conservation, but accepted that development needs sometimes made it necessary for archaeological deposits to be removed. The system therefore established mechanisms for a ‘second best’ solution that involved recording remains in such detail that the archive of records and finds became a surrogate for that part of the archaeological heritage being lost. The principle goal of ‘preservation by record’ can sometimes involve the creation and maintenance of a record, rather than achieve advances in understanding.
The consequence is a disciplinary fracture between problem-oriented research and data-led cultural resource management. 7 Frustration has understandably crept into academic debate over the direction taken: especially with the proposition that excavation is an objective exercise of gathering empirical data best kept free of interpretive presuppositions. 8 A problem with this critique, however, is that it risks devaluing the potential of the information obtained from rescue excavations. In his studies of the formative periods of Romano-British studies Richard Hingley has shown how it was this gradual accumulation of knowledge that gave impetus to new ways of thinking about the Roman past, amongst other things inspiring Haverfield’s work on Romanization. 9 Such reasoning continues to play an important role. 10 The philosopher James A. Bell has described the intellectual space left to inductive research, acknowledging that it remains a useful method when individualistic elements are drawn on in proposing archaeological explanation. 11 The very presence of unique components makes it difficult to develop models that yield testable explanations. This is the situation that applies in the study of Roman London: where separately observed sequences of unique events allow the reconstruction of an explanatory narrative history. What matters is that these explanations are consistent with the wealth of available data, and don’t selectively ignore or disqualify findings that fail to fit preconceptions brought to the research. Much has been achieved in the excavation of Roman London: achievements made because of, and not in spite of, the research practices of urban rescue archaeology and a centuries-old concern with the scientific description of London’s Roman finds.
Studying change
The main goal of this book is to use this wealth of descriptive documentation to understand how and why London changed through time. The reconstruction of historical narrative was once a principal goal of Romano-British studies, best realized by Sheppard Frere in his confidence that ‘it is folly not to use the material to construct a history, however provisional’. 12 More recent scholarship has preferred the different challenges of thematic and theoretical exploration, exploiting the contribution that material culture can make to our understanding of the past. 13 But it remains the case that most archaeological writing starts as narrative, setting disparate activities into a sequential order to aid explanation. As Ian Morris has observed, archaeological data allows us to write chronologically tight, sequential stories that draw on the centrality of the event as an analytical category. 14 The tighter the chronology the greater the opportunity to shift from elusive grand theory onto the microanalysis of events that illuminate aspects of the past. Our goal is to understand the factors and forces that link events one to the other, exposing possible causes of change. 15 Our point of departure, therefore, is a site specific sequence of events, from which we can ask how and why London differed to other Roman cities and what this tells us that cannot be learnt elsewhere.
London’s exceptionally tight chronologies are the product of deep stratigraphies that establish relative sequences, the exceptional potential of absolute scientific dating, and an abundance of typologically distinct classes of finds. The opportunity to compare and contrast relative and absolute dates across hundreds of different excavation sites refines and enhances these frameworks. Some of the most valuable information comes from tree-ring studies, or dendrochronology, since this introduces calendar dates to the narrative. Dendrochronology emerged as a significant archaeological tool at the end of the 1920s, and was successfully applied to waterlogged oak in London from the mid-1970s. Oak was widely used in London: in waterfront revetments, piled foundations, and lining wells and drains, where anaerobic and water-logged conditions aid survival. Over 1,000 structural timbers from Roman London have been dated by measuring tree-ring growth. 16 In many constructions, timbers were used in large numbers without trimming away sapwood and bark. Usually these show none of the ‘shakes’ that occur when timbers are allowed to dry out before use, the tool marks show no sign of the grain tearing that happens when seasoned timbers are reworked, and there is no evidence of woodworm or decay in the vulnerable sapwood. 17 These features show that the oak was used ‘green’, most probably in the season after it was felled. Where several fresh timbers were used in a single construction, the weight of corroborating evidence means that we can be reasonably certain of the exact year in which the works took place. Even where timbers are divorced from their original context by later reuse they still date an earlier procurement exercise. It is difficult to over-state the research potential of knowing exactly when so many different building programmes began.
London’s pottery assemblages also allow us to reconstruct surprisingly tight chronologies. Vast quantities of tablewares, cooking vessels, storage containers, and specialist products were used and broken. Many tonnes of sherds have been recovered and classified by form and fabric using consistently applied typologies. 18 More than 230 Roman fabric types are represented within the Museum of London’s pottery recording systems. These reflect on the different sources and circumstances of manufacture, and include elements of design introduced by potters to distinguish different wares. At a rough estimate, about one million separate Roman assemblages have been studied since the Museum of London started archaeological explorations in 1973, with over 440,000 different ‘spot-dating’ records entered into the Museum’s database since 1995. In addition to identifying the forms and fabrics present, these records usually describe the weight, sherd count, estimated number of vessels (ENVs), and estimated vessel equivalents (EVEs) allowing for advanced statistical analysis.
Changing networks of production and supply create differences between assemblages of different dates. Pottery forms change through time, as sources of manufacture change, as products are added to and dropped from the repertoire, and as new kilns come into use. Date ranges for different periods of pottery currency are established from excavations that also draw on scientific and epigraphic dating, and are refined through the comparative review of hundreds upon thousands of different relative sequences. Although deposits can be reworked and rubbish recycled, the sheer volume of these stratified assemblages allows us to establish confidently when certain classes of material first appeared in London. What this means is that the absence of the more ubiquitous types of pottery from a large assemblage signals that it was formed before such pottery came into widespread use.
The study of so many sites, some containing tens of thousands of sequenced stratigraphic activities, has allowed the identification of a series of ceramic phases. Five major phases of ceramic supply have been recognized within the first 120 years of Roman London’s history (to c. ad 165), and five more from the subsequent 250 years. 19 Pottery does not offer the absolute dating of dendrochronology, but its ubiquity makes it the most useful of all dating tools. Whilst the individual potsherd carries little weight, there is a security in numbers. The horizons of change represented by London’s different ceramic phases are therefore key to the narrative presented here. Further opportunities for close dating are provided by a select number of dated inscriptions and writing tablets. Coins also provide absolute dates, although these are less useful than might be expected since most issues remained in circulation long after they were minted. When viewed within stratified sequences these absolutely dated finds only suggest a date after which the layers were deposited, but how much after is not always clear. Although reworking and residuality can result in deposits being formed from misleadingly old rubbish, these anomalous deposits can be identified from their stratigraphic context. In the vast majority of cases finds dating fixes building events to within a decade or so although there were some periods of poorer supply, particularly during the later years of Roman London, where our dating frameworks are more uncertain.
This precision of archaeological dating presents an exciting challenge. Since we know exactly when parts of London’s fabric changed we can treat these as historical events, setting local episodes of re-planning and redirection within their broader context. The narrative that emerges hints at London’s vulnerability to ‘acts of god’—such as war, famine, pestilence, and natural disaster—which checked growth and interrupted archaeological sequences. This too is an important challenge. Koenraad Verboven has encouraged historians to study the role of such stochastic shocks on institutional change, and the reading of particularities in London’s archaeological record allows for exactly such study. 20
Objectscapes and landscapes
The description of London’s ceramic phases is, of course, far more than a dating tool. This artefactual evidence charts changing patterns of supply and consumption that were a product of the social and economic developments we seek to understand. Each major change in the range of goods deployed in London begs questions as to how and why it occurred. Explanation cannot assume that this evidence was uniquely the outcome of market forces, where suppliers changed the direction and nature of trade in adapting to consumer demand. Political considerations also shaped London’s access to the different forms of goods that have become archaeological finds.
The study of material culture is central to archaeological study. People surround themselves with possessions to sustain status, both materially and symbolically, and patterns of consumption can be interrogated against their implications for questions of identity and belief. The complexity of power relations and the sheer density of social networks in Rome’s expanding empire created space for many different forms of belonging. New combinations of objects—dress accessories, tablewares, and burial goods—reveal the spread of new customs. This has generated a significant body of research relevant to the study of Roman London that can only be touched on here. 21 Much of this has involved explorations of human agency in social practice, drawing on the theories of Pierre Bourdieu and Anthony Giddens to explore how individual actors contribute creatively to processes of change. 22
Identities were also expressed and shaped through urban design, and studies of how space was constructed and experienced forms an important strand of contemporary research into the ancient city. 23 The first attempts to reconstruct London’s Roman town plan emerged from the late sixteenth-century interest in cartography to map knowledge, and subsequent research has focused on reconstructing the urban layout as a means to understanding the origins and functions of the city. Studies of London’s urban form also build from an interest in town-planning as social engineering, where modern planners took inspiration from the ideals of orthogonally planned Greek and Roman cities. 24 Analysis now addresses issues of movement and connectivity, whilst recognizing the importance of urban design in the reproduction and manifestation of imperial power. 25
The design of Roman London was the product of many distinct phases. Its environment was the ordered creation of considered interventions, accommodating major changes in population scale and density, but designed with close regard for the topography of the site. The hundreds of excavations undertaken allow a detailed understanding of when and how London’s architecture evolved, although the fragmentary nature of archaeological survival and the limitations of access leave many gaps. Some four-fifths of the reconstructed plan of the Roman city depends on an informed exercise in joining the dots. There is sufficient, however to show how profoundly and frequently the landscape was transformed, and this allows new readings of the ideological meaning of the city.
A good example of the potential and pitfalls of such study is found in John Creighton’s comparative study of Romano-British town planning. 26 He suggests that irregularities in London’s plan came about because its architecture was determined by acts of individual patronage, mirroring elite practice elsewhere. This would suggest that there was no grand design for the city. In Creighton’s view, London developed as the product of social competition rather than externally imposed Roman ideals. The argument is elegant, but lacks supporting evidence. On the contrary, most of London’s irregularities seem likely to have been the product of different episodes of planned urban growth. It is consequently possible to describe a form of architectural rhetoric, in which individual monuments introduced new arguments and practices to city living but were shaped in response to earlier arguments already written into the landscape. Place-making framed the new within the old, naturalizing the artificiality of construction through corporeal anchors that respected inherited and sacred landscapes. In trying to understand these architectural arguments, we find scant evidence for the creative agency of most Londoners in the process of design. London’s public architecture and urban shape can be credited to the patronage of senior officials in the administration, perhaps acting on imperial instruction. Investment in London was born of Imperial attention not the product of local competition, and London’s architecture reveals a history of Roman intervention.
The relevance of ‘Romanization’
A principle concern of Romano-British studies involves the study of how ideas drawn from the Graeco-Roman world came into currency in Britain, inspiring a provincial culture that was visibly and materially Roman but locally conceived and elaborated. Martin Millett’s 1990 book the Romanisation of Britain involved an explicit shift away from the event-based narrative approach that previously prevailed, and has framed debate for a generation. In Millett’s view, the process of acculturation, or Romanization, depended on the involvement of Britain’s land-owning elite society, who benefitted through their active and creative engagement with the ideas that supported the political authority of imperial Rome.
The emphasis on Romanization as process helped scholars move beyond earlier top-down descriptions of how Rome was supposed to have brought civilization to its provinces. As Ray Laurence has observed, Millett’s model was closely influenced by Moses Finley’s earlier work on the ancient economy. 27 In this model, Roman territorial annexation introduced a monopoly of force and imposed taxes, but otherwise relied on a federation of diverse self-governing communities. These communities were governed through cities, where local propertied aristocrats controlled economic and political activity. Cities were nodes through which Rome raised taxes and administered justice, where political careers were forged, and where rituals and ceremonies gave temporal power divine mandate. These activities were entrusted to a local elite, co-opted as magistrates and public officials. Rome relied on institutionalized competitive office-holding within such communities, and on the attendant networks of patronage that embedded local communities within wider imperial landscapes. Land was the principal source of wealth, whether in the form of rents, taxes, or agricultural surplus. Power consequently resided with those who owned land, and this also served as a qualification for participation in civic life. Drawing on these political and economic understandings Millett was able to describe an exercise of Roman imperial rule that remained largely decentralized, at least in territories outside of military control. 28 In his minimalist vision, Millett describes Britain as a place that held relatively little interest to those who governed the empire, except as a place where political prestige might be won, and who preferred to leave administration in local hands. Rome is argued to have gained and held onto its empire with remarkably little in the way of centralized imperial administrative bureaucracy. Government was achieved by co-opting local, provincial elites who ruled as agents of Rome. This empowered local aristocrats, who governed and raised taxes on Rome’s behalf.
A generation of scholarship has explored this ‘Romanization’ model to the point of exhaustion. It isn’t possible to do justice here to the many competing and complementary perspectives developed by way of critical approaches to ‘Romanization’, including those rooted in post-colonial theory, world-systems theory, globalization, and other explorations of relationships of power and social identity. 29 This study is taken in a somewhat different direction by the nature of London’s archaeology. London was a new foundation, planted beside the Thames early in the history of the Roman administration (Chapter 5). There is little to suggest the presence of local aristocrats ready to be drawn into government and the city relied on an immigrant community of officials, soldiers, and merchants. The evidence marshalled here suggests that the provincial administration was more actively involved in London’s affairs than usually assumed. This makes it a place where we can trace the consequences of Rome’s imperial rule. Far from witnessing the light touch of a distant governing power, the study of London describes twists and turns in the direct exercise of Roman authority. The decisions of emperors and their representatives had consequences. This gave agency to the powerful few, where London’s experiences and manifestations of Rome were imported and imposed. The government may have sometimes been creative, adding value and generating surplus, but at times—perhaps most of them—was brutal and exploitative.
The Romanization debate is concerned with the ways in which locally based elite society found it advantageous to work within Roman cultural norms, themselves of Hellenistic inspiration, whilst elaborating new ways of being Roman. This shifted the focus of attention away from Roman rule as a purportedly civilizing force and onto the creative role of provincial society, but it has also moved the issue of military and political command to the margins of debate. This study offers a different take on the relationship between colonial forces and the colonized, one sympathetic to David Mattingly’s description of an oppressive imperial presence. 30 The ‘trickle down’ and ‘upward mobility’ issues addressed in discussions about Romanization don’t really apply. This was a place dominated by the political establishment, enjoying lifestyles that were radically different to those of subject populations within the urban hinterland. 31 London was, it is argued, a creature of Roman political purpose. As in more recent times, it can be described as a city unmoored from the territories it governed, identified with the rule of interconnected ‘elites’ of commercial, political and social classes who served their own interests ahead of all others. 32
An alien city
Visions of London’s origins have variously lauded its mythical antiquity, the civilizing glories of Roman imperial expansion, the proud mercantilism of its thriving port, and the exotic diversities of an open and multi-cultural city. Shadows of these arguments can be found in the chapters that follow, but they say more about the presumptions of the present than they do about the constitution of the past.
London was the largest and arguably the most Roman city of Britain. Throughout much of our period it was a principal point of entry into the province, and London’s fortunes were to ebb and flow with the tide of imperial interest. London is important, therefore, not just as the principal town of Roman Britain but also as a part of the Roman Empire. The city was founded on the borders of existing polities, exploiting a previously marginal location to stand free of pre-existing political structures. It served Roman colonial needs: a city mediating between Rome and a conquered territory; the product of imperial patronage and an instrument of imperial control. London was consequently its own place with its own identity, unfamiliar to both Rome and to Britain.
Britain was a frontier province with an unusually large garrison, and remained an active sphere of military operations long after the Claudian conquest. Campaigns were intermittently underway throughout the first century, and well into the second. They were a recurring feature of many periods thereafter. Since London lay within southern Britain, it is usually considered as part of a self-governing ‘civil’ zone behind areas under direct military occupation. It needs to be remembered, however, that London fell within the command structure responsible for military operations. Politically and strategically, it was a frontier city and its port continued to play an important role in official supply well into the third century. This meant that power was exercised by senior government officials, military commanders, and the extended household of the emperor and his servants. There was space within these structures for competition and political diversity, but they marginalized civic institutions and local magnates. London was dominated by immigrant officials and their agents, with no evident role for local land-owning aristocracies, and witnessed an imposed colonial power. Because of its particular role to the colonial government, London became a hub for the introduction and elaboration of cultural practices of Mediterranean and continental inspiration. These differed in important detail from more exclusively military styles of living found in the camps, giving Roman London a discrete and individual cultural identity. As a consequence, it remained an exotic ephemera: neither Roman nor British, military nor civilian, mercantile nor administered. These are false distinctions to draw. What it was, however, was a place apart with its own story. To this our attention can now turn.