Introduction: the origins of oppida
Understanding the genesis of the massive centres known as oppida has been a long-standing focus for researchers of the European Iron Age. Recognition, in the 1980s, that many may have been preceded by unenclosed agglomerations led to this being considered an earlier stage of social and economic centralisation.1 Although Late Iron Age Britain saw the development of monumental complexes, similar to oppida elsewhere, the Middle Iron Age (c. 400 BC – 100 BC) appears to have been devoid of such agglomerations.
This paper reviews evidence for pre-oppida agglomerations and the alternative settlements that some have suggested fulfilled similar roles in Britain. It argues that Britain witnessed an alternative trajectory in the origins of oppida, suggesting they emerged from, and primarily remained, assembly places. These differences may reflect the heterarchical nature of social organisation, with British oppida an indigenous solution to increasing population and changing social relations in the final decades of the Iron Age. These are part of the varied responses by Iron Age societies across northern and central Europe to a range of similar pressures reflecting different underlying social structures and the agency of communities and individuals.
Oppida and agglomerations
The discovery in the 1970s and 1980s of large, unenclosed settlements, particularly Levroux, Basel and Aulnat-Gandaillat, indicated that the enclosed Late La Tène oppida, which had been the focus of investigation since the late 19th century, were preceded by an earlier phase of social centralisation.2 Dating largely to the Middle La Tène (3rd-2nd c. BC) such complexes appeared to have been focused on specialised, artisan activities. Developed largely from studies of Levroux and Aulnat, a theory emerged which argued that unenclosed agglomerations were the precursors to many oppida,3 potentially the first stage upon an evolutionary road to urbanism.
Agglomerations are now understood to have been widespread across Gaul4 and have been identified farther afield in central Europe.5 Studies have added further complexity, recognising that some agglomerations were contemporary with oppida, as multi-polar complexes or extra mural settlements,6 with a recognition that the model of agglomerations as evolutionary precursors was far from universal.7 Use of the term agglomeration has subsequently become increasingly varied, although a group of ‘agglomération artisanale’ remain identifiable, defined by craft-working, the presence of specialists, and a focus on regional exchange.8
Despite an increasing wealth of information on the Iron Age in Britain, largely due to development-led archaeology, there remains very little evidence of agglomerations comparable to those on the continent. In the late 1st c. BC and 1st c. AD, however, Britain saw the appearance of large centres, usually referred to as oppida, which shared similarities with their continental counterparts, including Roman imports, rich burials, and huge systems of ramparts. This raises questions as to why Britain did not also see the development of earlier specialised settlements, implying perhaps there was something distinct about how British oppida emerged or that other settlements may have filled the roles played by continental agglomerations.
Previous approaches: proxies and models
The agglomerations which occur in Britain in the Middle to Late Iron Age have been classified using a variety of terminologies (fig. 1), although their coherence as categories of monument has been subject to debate.9 Traditionally,10 three types of larger agglomerations can be identified in Britain: hillforts: well-defined enclosures, predominantly dating from the Early and Middle Iron Age (800-100 BC) which rarely exceeded 20 ha in size.11 In many areas hillforts were largely abandoned by the 1st c. BC, and although they continued later in some regions, such as south-western Britain, their roles appear to have changed; enclosed oppida (fig. 2): distinguishable from hillforts by their location in low-lying areas and apparently occupied slightly later, during the Middle to Late Iron Age (c. 200 BC to 1st c. AD); territorial oppida (fig. 3): consisting of sprawling earthworks that cover large areas of landscape, often hundreds of hectares, but which lack evidence for large areas of dense occupation. Instead, activities within them, including artisanal areas, elite occupation, and funerary monuments, were dispersed amongst seemingly open landscapes. Some analyses have questioned the relevance of the label oppidum for these complexes, their lack of dense occupation or evidence for large-scale trade contrasting with many continental centres.12 This has led to their description instead as ‘polyfocal complexes’, reflecting their dispersed nature,13 or as ‘royal sites’, recognising their roles as centres of kingship and their similarities to Irish complexes of that name, such as Navan, County Armagh.14 As will be explored below, such debates reflect the problematic nature of attempting to neatly categorise agglomerations in Britain.
The ‘territorial oppida’15 emerged in Britain in the final decades of the Iron Age. Silchester and Camulodunumprobably began as major centres in the late 1st c. BC16 while others, such as Bagendon, flourished slightly later in the early-mid 1st c. AD.17 Despite debate as to whether the British complexes should be called oppida, they fulfilled many of the same roles as their continental counterparts, as socio-political centres, and may reflect similar transformations in society.18
Until relatively recently, models of oppida development in Britain regarded their emergence as stimulated by the impact of trade with the Roman Empire, which increased from the late 2nd c. BC and particularly after the mid-1st c. BC, as southern Britain came into direct contact with Rome after Gaul’s annexation.19 Barry Cunliffe also saw this as a process of social evolution, beginning with the competitive growth of hillfort territories.20 For him, this represented a move from smaller-scale societies to larger ones, culminating in a transition from chiefdoms to states, represented by a shift from hillforts to oppida in the 1st c. BC.
Cunliffe21 recognised that the process of centralisation was more complex, however. Many of the hillforts, which he regarded as dominating Middle Iron Age societies, were abandoned, or had significantly declined, by the early 1st c. BC, while oppida were hard to date earlier than the very late 1st c. BC. It was also clear that oppida emerged in different areas of the landscape to earlier hillforts, often in low-lying locations. Cunliffe22 argued, therefore, that the transition from hillforts to oppida was via ‘enclosed oppida’, large enclosures often situated in river valleys, which emerged in the early 1st c. BC (fig. 2).23 He also argued that some upland ‘enclosed oppida’ were situated close to the later territorial oppida which replaced them in the 1st c. AD, seen for example in a sequence from Wheathampstead to Verlamion24 and Bigberry to Canterbury.25 This model echoed, but effectively reversed, that suggested for sites like Levroux in Gaul, arguing instead for a shift from a nearby hilltop enclosure to a later, low-lying ‘territorial oppidum’.
Julian Henderson also argued that Britain followed a similar path to the rest of Europe with increasing craft specialisation across the Middle to Late Iron Age.26 He argued for evidence of the emergence of ‘industrial villages’ in the Middle Iron Age, focusing on evidence for glass manufacturing at Meare, Somerset, and iron production at Hunsbury, Northamptonshire, with more such settlements emerging in the Late Iron Age, most notably at Gussage All Saints, Dorset.
This model of Iron Age Britain has come under sustained critique during the last 30 years. Hillforts, such as Danebury, certainly represented agglomerations; at its peak, in the 2nd c. BC, Danebury was 16 ha with a population of perhaps 250 inhabitants.27 There is little evidence, however, that such hillforts acted as economic hubs or were elite centres at the top of settlement hierarchies.28 Neither do hillforts appear to have had control over production, for example of metalwork or ceramics, or had higher levels of regionally derived materials.
Suggestions that British oppida development was stimulated by trade from Mediterranean Europe have also been challenged. Most oppida produce only small numbers of imports, often found in high status burials; these might better be regarded as diplomatic gifts or evidence for elite exchange networks, rather than large-scale trade.29 Critiques of Cunliffe’s narrative have tended to regard developments in Britain as very different from those in Gaul, considering British oppida more akin to Irish ‘royal sites’30 and developing in relation to elite power dynamics rather than the impact of trade.31
While recognising such differences we should be wary of assuming that Britain was somehow less developed than continental Europe. Southern Britain shared, sometimes contemporaneously, many of the changes seen on the continent, such as the development of coinage, and witnessed a similar trajectory in settlement patterns. By the late 2nd c. BC imports of Dressel 1 were entering southern Britain, if not in the quantities seen in Gaul, in numbers that emphasise significant interaction across the Channel, some of which had taken place for some time.32 The adoption of cremation burial by communities along the south coast began by the late 2nd c. BC, suggesting that Britain was already undergoing transformations before the impact of Rome in the 1st c. BC and 1st c. AD.33 In light of these considerations, it becomes more important to explain how and why British oppida developed in the ways they did and why this differed from other parts of Europe.
Settlement and demography in Later Iron Age Britain
Before exploring the development of oppida in Britain, it is important to consider the context in which they emerged, particularly the long-term settlement and demographic changes in the later 1st millennium BC.
Across most of southern Britain, Early Iron Age (800-400 BC) settlements were predominantly unenclosed, frequently consisting of clusters of post-built roundhouses. These often shifted across the landscape as houses were rebuilt. From the beginning of the Middle Iron Age (c. 400 BC), these settlements were often replaced by more permanent, enclosed farmsteads, usually encompassing one or two roundhouses. This was part of a broader move towards enclosed settlements across much of Britain, coinciding with a spate of hillfort construction or refurbishment in areas such as central-southern Britain, northern England and the Welsh Marches.
As well as witnessing changes in settlement form, the Middle Iron Age also saw an increase in the overall number of settlements, as seen for example in the Severn-Cotswolds.34 Similar increases have been identified in Wessex35, Yorkshire36 and the East Midlands.37 These regional developments seem to have been part of a wider pattern identified in an assessment of summed radiocarbon dates. This indicates an increase in settlements throughout the British Isles beginning around the 4th c. BC.38 Although such assessments have problems, for example potentially under-estimating settlement from the Early Iron Age where fewer dates have been procured, combined with regional assessments they support an impression that settlements increased from the Middle Iron Age.
Alongside changes in settlement patterns, there is evidence for the intensification and extensification of the use of landscape. Some regions saw the reappearance of large-scale field systems which integrated enclosures and trackways.39 This is matched by expansion into the fenlands of Somerset, East Anglia and south Wales.40 In some regions, this corresponded with the appearance of new settlement forms, such as the so-called ‘banjo’ enclosures (named for their peculiar form) representing the exploitation of landscapes which appear to have been used less intensively in the Early Iron Age.41 This more extensive use of the landscape may explain the emphasis on enclosed settlements at this time, indicating communities’ desires to stress their tenure of the landscape.42
The application of Bayesian analysis to radiocarbon dates is allowing greater refinement of when these changes took place.43 Assessment of the initial dating of enclosures in the Severn-Cotswolds, for example, although based on a relatively limited corpus, suggests many small enclosures emerged around 300 BC.44 Assessments of some other regions indicate that many appeared closer to 200 BC, suggesting this may have been a long process of filling up a densely settled landscape.45
Developments in Britain appear to have been part of a wider process of settlement expansion and a move towards enclosed settlements from c. 300 BC seen, for example, in northern France46 and the Low Countries.47 The widespread nature of these changes suggests large-scale factors, related to increasing population and climate trends, were partly responsible. Although climate proxies are imprecise, and the impact of localised factors need to be considered, they indicate that Britain (and much of northwest Europe) experienced a warmer phase after c. 300 BC,48 following a relatively cool and wet period (c. 800-400 BC).49 Temperatures appear to have increased further during the so-called ‘Roman warm period’ from the turn of the millennium. The resultant longer growing seasons are likely to have had an incremental impact on productivity, leading to an increasing population.50
This increase in settlement was contemporary with other transformations in society. From the 4th c. BC regional ceramic types emerged. This marked the increasingly interconnected nature of communities, exchanging a variety of material culture, including ceramics, millstones and glass beads.51 These were once seen as evidence for emerging socio-ethnic groups or ‘tribes’.52 There is, however, little evidence for overarching hierarchy and these societies probably represented loose networks, rather than socio-cultural entities.53 The lack of variation between settlements suggests that these societies remained heterarchical,54 with limited status distinctions between communities and mechanisms to prevent fixed hierarchies emerging.55
Many regional assessments suggest the increase in settlement continued through the Late Iron Age, for example in the Severn-Cotswolds and parts of southern England.56 Some regions do have evidence for disruption of settlements, however, Wessex being one example.57 For Essex, some have even argued for population decline in the 1st c. BC.58 Certainly, a decrease in activity is evident on some settlements, for example in the East Midlands.59 The extent to which this marked a major population decline, rather than shift in settlement form or location, is questionable, however. In some regions the decline may reflect a reliance on identifying Late Iron Age phases through wheel-thrown ceramics, which many communities did not have access to.60 It may also indicate changes in settlement form, especially where there is clear continuity between Middle Iron Age and Roman occupation. The absence of Late Iron Age houses may also be due to a move to building forms that are less detectable archaeologically, either as sill-beam or cob structures.61
Evidence for specialised communities in Britain
Despite some variation, southern Britain saw a relatively steady increase in settlements from c. 300 BC. These were situated in a more interconnected landscape of exchange networks with regional identities emerging. Societies remained heterarchical, with little evidence for centralised elites. How, then, did these societies manage an increase in population, and to what extent did this lead to greater centralisation or specialisation of communities in the Late Iron Age?
To answer this, it is important to determine whether there were a range of settlements that acted as central places prior to the territorial oppida. One group is the ‘enclosed oppida’ proposed by Barry Cunliffe62 to have fulfilled such a role. The corpus of such sites has not extended much beyond Cunliffe’s earlier list, although enclosures at Abingdon, Oxfordshire; Cassington, Oxfordshire; Stonea, Cambridgeshire and Burgh, Suffolk are sometimes included (fig. 2). Silchester, Berkshire is sometimes referred to in these terms too.63 Haselgrove has also compared Stanwick’s earliest phase (The Tofts) to low-lying enclosures, such as Oram’s Arbour.64
Although our understanding of many of these sites remains limited, it seems that (contrary to Cunliffe’s claims) they were not a unified phenomenon.65 Most are larger than earlier hillforts, but some hillforts, such as Maiden Castle, Dorset, are not much smaller, while other hillforts, like Ham Hill, Somerset (fig. 2), which was occupied into the Late Iron Age,66 blur the distinction between these categories.
Evidence that these enclosures developed after the abandonment of hillforts but prior to the ‘territorial oppida’ is by no means clear either. At Salmonsbury and Abingdon, the enclosures seem to have encompassed pre-existing Middle Iron Age settlements.67 At Salmonsbury, the enclosure may represent the nucleation of earlier settlements, rather than a new central place. Both Salmonsbury and Abingdon also show continued occupation in the Late Iron Age, although its character is hard to determine. Others that emerged in the Middle Iron Age, such as Oldbury, Kent, are hard to distinguish from hillforts.68 Oram’s Arbour seems to have begun in the Middle Iron Age too, probably in the 2nd c. BC, but was abandoned well before the Roman conquest.69 Examples that have been regarded as precursors to the territorial oppida, such as The Ditches, Gloucestershire, close to Bagendon,70 seem in fact to be integral parts of these Late Iron Age complexes. Others, like Wheathampstead, Hertfordshire, are questionable as coherent enclosures at all.71
Understanding the internal occupation of these enclosures can be difficult, but geophysical survey at Salmonsbury and aerial photography at Dyke Hill indicates a relatively dispersed occupation consisting of small enclosures and roundhouses. There is little to imply that they had different roles to other hillforts or agglomerated rural settlements, with no greater evidence for specialised activities, such as metalworking. The location of some close to major rivers, which may have been routeways, has been used to argue for their role in controlling long distance exchange. However, by the Middle Iron Age most communities were engaged in regional exchange, as noted above, and none of these sites has yielded significantly greater quantities of such material. The presence at some, such as Salmonsbury, of Late Iron Age imports tells us instead about their roles in the final decades of the Iron Age.
A number of these settlements were replaced by small Roman towns (e.g. Salmonsbury, Abingdon, Dyke Hills), even a civitas capital in the case of Oram’s Arbour (Venta Belgarum), which may signify their social importance at the time of conquest. However, a link between Roman settlement status and the social importance of any Iron Age precursor is questionable. Roman small towns were preceded by a range of Iron Age sites, including sanctuaries and small settlements, or represented entirely new establishments.72 Inferring the status of Iron Age activity risks uncritically back projecting the Roman geography of Britain.73 Rather than a coherent group, the ‘enclosed oppida’ seem to represent a variety of settlements that were part of a trajectory towards a densely settled, and enclosed, landscape in the final centuries of the 1st millennium BC. There is, however, little to suggest they performed the same roles as ‘agglomération artisanale’ found in continental Europe.
If hillforts and enclosed oppida were not central places for exchange or administration, did other settlements fulfil these roles? There are a few examples of Middle Iron Age settlements which appear different to both hillforts and small farmsteads that included specialised artisanal activities. Glastonbury Lake Village and Meare, in Somerset, were located in what was then fenland. Occupied between c. 175 and 50 BC,74 the character of the sites varies. Glastonbury is the largest and most likely to have been permanently occupied, with a population of perhaps 200 at its height.75 Meare seems to have focused on specialist production, particularly glass bead manufacture, and may have been a seasonal exchange centre.76 While Glastonbury has evidence of metalworking, there is insufficient evidence to argue it was home to specialist craftworkers, however. The Somerset lake villages are yet to be paralleled elsewhere. Specialist communities may have existed in the East Anglian fens,77 but most seem to have been seasonal or ritual in character rather than permanent agglomerations.
The glass making at Meare was part of the emergence of places that were foci for specialised activities, including activities such as salt making and ceramic production.78 Most, however, were seasonally utilised, with little evidence for permanent, specialised communities or direct control of production. Many, like Meare, appear to have been deliberately located away from densely settled landscapes. The so-called ‘lake villages’ did not represent the appearance of specialised, industrial villages,79 therefore, but the exploitation of marginal landscapes in a more systematic manner and use of such areas for production and exchange in what may have been more neutral locations.80
Agglomerations of Middle Iron Age date can be identified elsewhere, but these were rural in character. Large clusters of unenclosed roundhouses are well known in the Thames Valley and East Midlands, where they are sometimes given the epithet ‘village’. For example, Crick Covert, Northamptonshire, may have included 240 inhabitants in a settlement covering c. 12ha and was potentially part of a larger group of integrated settlements (fig. 4).81 Others can be identified in the Thames valley, at Mucking, Essex, and around Stanton Harcourt, Oxfordshire.82 Some may have had populations in the hundreds, although they were dispersed across loosely connected settlements. However, there is no evidence these were specialised communities, and most have relatively limited evidence of long-distance exchange.
Although the presence of agglomerations focused on craftworking and exchange is lacking for most of the Iron Age, evidence from the Late Iron Age is more complex. One agglomeration with some similarity to those in Gaul is Braughing, Hertfordshire. This unenclosed settlement was located c. 30km from the oppidum at Verlamion (St. Albans). Although relatively poorly understood, activity of Late Iron Age date appears to have covered c. 170ha.83Partridge, 1981; Potter and Trow, 1988. Within this, however, only smaller areas of occupation are identifiable, most notably at Skeleton Green and Puckeridge, which may have covered as much as 10ha.84 At its maximum extent this agglomeration would be on a par with some of the larger examples in Gaul; however current evidence indicates that occupation shifted within this area overtime, rather than it being a coherent agglomeration.85 The complex received imports, including Dressel 1 amphorae, in the late 2nd and early 1st c. BC, and has yielded large numbers of coin-moulds, suggesting it was an important social and exchange centre. While Braughing emerged earlier than the oppidum at Verlamion, which developed in the early 1st c. AD,86 occupation at both was concurrent, with Braughing finally developing into a small Roman town.
Recent discoveries close to the oppidum at Stanwick have revealed an Iron Age settlement at Scotch Corner c. 6km to the south-east, comprising a range of small enclosures, situated on the Roman road which reflected an existing important north-south route.87 The presence of pellet or coin moulds (surprising considering there is no tradition of coin making in the region) and other craft activity suggest an important settlement. However, the Middle Iron Age phase appears rural in character while the later phases are contemporary with occupation of the oppidum at Stanwick, suggesting the two may have been part of an integrated complex rather than sequential.
Agglomerations may have existed elsewhere but there is limited information concerning them. At Weston-under-Penyard, Herefordshire, a small Roman town was preceded by a settlement with evidence of iron-working and a significant Late Iron Age coin assemblage, but it lacks evidence of dense occupation.88 Eastern England also has several nucleated settlements which emerged in the late 1st c. BC. Heybridge, Essex, for example, is represented by a cluster of roundhouses and roads which morphed into a small Roman town. It is hard to determine whether it had any specialist roles, although the presence of a sanctuary at its core may indicate it acted as local centre.89 Other enigmatic Late Iron Age agglomerations are known at Sleaford, Dragonby and South Ferriby in Lincolnshire, with evidence of coin moulds and imports, some argued to be ‘ports of trade’.90 Other agglomerations may have existed beneath Roman towns, for example at Gloucester and Caistor St Edmunds,91 as has been argued for the supposed oppida beneath Roman civitas capitals such as Leicester and Canterbury.92 But these settlements remain enigmatic, largely only recognisable through assemblages of Late Iron Age coins or brooches, with earlier structural remains presumably truncated by Roman and later activity.
Some of the examples above may represent agglomerations partly focused on craft activities and involved in consumption of Roman imports, in some ways similar to the areas identified within territorial oppida such as Bagendon and Camulodunum, which combined craftworking and feasting. Most, however, seem unlikely to have been large agglomerations but were more ephemeral gathering places. Rather than seek to classify these agglomerations morphologically, or rank them, it seems better to regard these as a range of loosely defined centres which emerged in the final decades of the Iron Age. Evidence from Scotch Corner reminds us that unenclosed, potentially specialist settlements may await detection, but they are unlikely to be of the scale of those in Gaul.
British oppida and their origins
The discussion above emphasises that is unlikely that specialised agglomerations existed prior to the late 1st c. BC. The unenclosed agglomerations that have been identified are either aggregated rural settlements or are of Late Iron Age date, with the latter better considered alongside the broader oppida phenomenon in Britain. There is also little evidence that a coherent group of ‘enclosed oppida’ fulfilled central place or specialist functions in the centuries prior to the appearance of the territorial oppida. How then did Late Iron Age oppida emerge from the intensively occupied Middle Iron Age landscapes and why did Britain not see a phase of artisan and exchange focused agglomerations? Recent work is beginning to provide evidence which may help to address some of these issues.
Many British oppida appear to have been located in landscapes that were lacking in Middle Iron Age settlement.93 Verlamion, Silchester and Bagendon, for example, all appear to have been situated in areas of the landscape that were less densely settled in the Middle Iron Age. The mystery around such oppida has been: what drew people to these locations if not the presence of a pre-existing settlement?94 Previous discussions suggested this was due to a radical power shift in society.95 Recent assessment, however, indicates that, although these landscapes were less intensively occupied, the idea that they were unexploited is questionable.96 This research is important in explaining how these landscapes were used and their significance prior to the Late Iron Age. Large-scale geophysical survey and excavation at the oppidum at Bagendon, for example, has revealed two previously unknown, banjo-type enclosures within the centre of the complex.97 These were constructed around 300 BC, contemporary with the spate of enclosures appearing elsewhere in the region. They seem to have been contemporary with linear features that represent precursors to the more substantial oppidum earthworks (fig. 4). The unusual form of these enclosures and their location in what remained a relatively wooded landscape suggests they were not typical settlements. Evidence from isotopic analysis also indicates that already in the Middle Iron Age people and horses were coming to the site from significant distances.98 Combined with the fact that the enclosures were abandoned when the complex was transformed into a centre of long-distance exchange and regional power in the Late Iron Age, this suggests that these enclosures were already hubs of interaction for communities from the wider region.
Evidence from elsewhere indicates other oppida originated as earlier social centres. At Silchester, a Middle Iron Age enclosure is suggested as a precursor to the oppidum.99 Although the character of activity within the enclosure remains uncertain, it seems to have been used seasonally, rather than representing a permanent settlement. At Stanwick occupation pre-dating the outer ramparts dates back to c. 80 BC. This consisted of a small enclosure and unenclosed settlement overlooking a marshy area, possibly a ritual focus.100 At Chichester, Nicky Garland argues, the oppidumdeveloped from a steady increase in settlement with various significant social foci.101 Both Stanwick and Chichester emerged, therefore, not in empty areas but from increasingly densely occupied landscapes, but like those at Bagendon and Silchester they appear to have been focal places for these rural communities.
The apparent emptiness of the landscapes in which many oppida emerged was thought to indicate that these locations were ritual foci or elite centres.102 Evidence that these locations were often occupied in some manner has challenged this perspective,103 although the unusual character of these locations remains. Indeed, this early activity at British oppida may explain how the highly contested landscapes of the Middle to Late Iron Age were organised. Late Iron Age oppida appear to have been situated at what were already important assembly places or nodal points in otherwise densely settled landscapes, places where rural communities gathered for exchange and interaction. Rather than mark a fundamental dislocation from it, oppida seem to have emerged from the existing social landscape.104
The early origins of oppida as assembly places rather than major settlements largely reflects their role in the 1st c. AD. At most British oppida much of their interior appears devoid of dense occupation. At Bagendon, for example, the massive earthworks encompassed c. 200 ha which included an area of metalworking and coin minting, covering c. 16 ha, and discreet enclosures which might represent elite residence (fig. 5). The rest of the interior appears to have been devoid of structures, suggesting a role for the corralling of animals or temporary gathering of people.105 Few other oppida have been subject to systematic survey, but a similar absence of dense settlement and presence of large open areas seems to be the case at Camulodunum and Verlamion (fig. 3).106 Most oppida can be regarded as polyfocal, with activities including craftworking, elite enclosure and burial, scattered across a wide area.107
Further evidence that oppida were not densely occupied is indicated by their apparently limited impact on rural settlement. In the Bagendon environs, for example, few settlements were abandoned at the time Bagendon was occupied. While some regions witnessed settlement dislocation, it is hard to relate this to the development of oppida, most of which flourished later, in the 1st c. AD.
The implication is that Late Iron Age oppida in Britain were always primarily assembly places, rather than large settlements. It also seems that most territorial oppida probably represent only those centres which were the most successful as socio-political centres at the time of the Roman conquest. Many other Late Iron Age centres, like those identified above, may have been assembly places of similar significance, but few developed as anything other than local foci. This reminds us that other Late Iron Age centres also existed, for example at Saham Toney, Norfolk, and the complex at Loose, Kent, which are rarely labelled oppida but may have had similar roles and been part of this diverse picture.108 An analogy for this array of social centres might be seen in the range of locations used as early medieval assembly places, with locales operating at different social scales and accommodating varied numbers of people.109 We should hesitate to regard this as simply reflecting a neat model of pagi and civitates.110 Instead, the variation outlined above more likely reflects the fluidity and instability of Late Iron Age societies that is increasingly understood from other evidence, such as coinage distribution.111
The lack of evidence for agglomerations in Britain similar to those in continental Europe does not reflect a less developed economy, instead it emphasises social differences. The nature of Late Iron Age oppida in Britain indicates that while power was increasingly centralised, and perhaps held by smaller groups, there remained an impetus not to coalesce people within these centres. While power was expressed at oppida, probably through periodic gathering, feasting and elite burial, these centres were never proto-urban settlements. The reason for this difference does not seem to have been economic; the densely settled landscapes of Later Iron Age Britain could have supported large agglomerations, and inter-regional exchange already existed. Instead, it seems to have been a deliberate choice to ensure power was partly retained by rural communities. Rather than a catalyst for dispersed Iron Age societies becoming more hierarchical,112 the emergence of oppida from existing foci and their continued role as assembly places suggests that societies continued to be based on negotiated forms of power.113
Significant questions remain to be answered to allow us to confirm this narrative. Settlement chronologies remain imprecise; a more systematic use of radiocarbon dating, alongside Bayesian analysis, is likely to reveal greater complexity, with the potential to tease out the ebb and flow that may have existed in the development of oppida. More precise dating of rural settlements, in particular, would enable a clearer understanding of their relationship to the oppida.
There is also a need to better understand the oppida themselves. Despite recent studies,114 we have limited understanding of the interior of most complexes, including the nature of the apparently ‘empty’ areas. Better understanding of the other Late Iron age centres is also required. While ‘enclosed oppida’ were not a coherent category of settlement, the chronology, internal structure, and role of many examples remains unclear. Similarly, the chronology and layout of agglomerations such as Braughing remains to be fully understood.
Only when we have a better understanding of these issues can we fully appreciate how oppida emerged from Middle Iron Age landscapes. It seems likely that such studies will begin to recognise localised trajectories reflecting Iron Age communities’ adaptation to pressures of increasing population and changing internal power dynamics, as well as to external forces, such as the expanding Roman Empire. This may explain the different forms of social centre we see in the Late Iron Age. These contrasts are unlikely simply to map onto levels of interaction with Roman trade115 or the impact of Roman power,116 but may also reflect pre-existing social organisation and the agency of individuals and communities in experiments of centralisation.
- Armit, I., Swindles, G., Becker, K., Plunkett G. and Blaauw, M. (2014): “Rapid climate change did not cause population collapse at the end of the European Bronze Age”, Proceedings of the National Academy of Sciences of the United States of America,111,(48), 17045-9.
- Atkinson, M. and Preston, S. (2015): Heybridge, a Late Iron Age and Roman Settlement: excavations at Elms Farm 1993-5, Chelmsford.
- Barral, P. and Nouvel P. (2012): “La dynamique d’urbanisation à la fin de l’âge du Fer dans le Centre-Est (Bourgogne–Franche-Comté) ”, in: Sievers, S. and Schönfelder, M., ed. : Die Frage der Protourbanisation in der Eisenzeit/La question de la proto-urbanisation à l’âge du Fer, Bonn, 139-164.
- Bevan, A., Colledge, S., Fuller, D., Fyfe, R., Shennan, S. and Stevens, C. (2017): “Holocene fluctuations in human population demonstrate repeated links to food production and climate”, Proceedings of the National Academy of Sciences of the United States of America, 114(49), E10524-31.
- Bradley, R., Haselgrove, C., Vander Linden, M. and Webley, L. (2016): The Later Prehistory of North-West Europe. The evidence of development-led fieldwork, Oxford.
- Brudenell, M. (2021): Late Bronze Age to Middle Iron Age Resource Assessment. East of England Research Framework [online] accessed 7 Feb. 2022, [online] https://researchframeworks.org/eoe/resource-assessments/late-bronze-age-to-middle-iron-age/ [consulted on 13/12/22]
- Bryant, S. (2007): “Central places or special places? The origins and development of ‘oppida’ in Hertfordshire”, in: Haselgrove, C. and Moore, T., ed.: Later Iron Age in Britain and Beyond, Oxford, 62-80.
- Buchenschutz, O. (1995): “The significance of major settlements in European Iron Age society”, in: Arnold, B. A., Gibson, D.B., ed.: Celtic Chiefdom, Celtic State: the evolution of complex social systems in prehistoric Europe, Cambridge, 53-63.
- Champion, T. (2007): Prehistoric Kent, in: Williams, J. H., ed.: The Archaeology of Kent, Woodbridge, 67-134.
- Coles, J. (1987): Meare Village East. The excavations of A. Bulleid and H. St George Gray 1932-1956, Somerset Levels Papers, 13, Exeter.
- Coles, J. and Minnitt, S. (1995): Industrious and fairly civilized: the Glastonbury Lake Village, Somerset Levels Project, Exeter.
- Collis, J. (1984): Oppida. Earliest towns north of the Alps, J.R. Collis publications, Sheffield.
- Collis, J. (2019): “Oram’s Arbour, Winchester: a new interpretation”, in: Romankiewicz, T., Fernández-Götz, M., Lock, G. and Buchenschutz, O., ed.: Enclosing space, opening new ground. Iron Age studies from Scotland to Mainland Europe, Oxford, 65-85.
- Collis, J., Krausz, S. and Guichard, V., ed. (2000) : “Les villages ouverts en Gaule centrale aux IIe et Ier s. av. J.-C.” in : Guichard, V., Sievers, S. and Urban, O., ed.: Les processus d’urbanisation à l’âge du fer. Eisenzeitliche Urbanisationsprozesse: actes du colloque Glux-en-Glenne, 8-11 juin 1998, Bibracte 4, Glux-en-Glenne, 73-82.
- Creighton, J. (2006): Britannia: the creation of a Roman province, London.
- Creighton, J. and Fry, R. (2016): Silchester: changing visions of a Roman town. Integrating geophysics and archaeology, the results of the Silchester Mapping Project 2005-2010, London.
- Cunliffe, B. (1976): “The origins of urbanisation in Britain”, in: Cunliffe, B. and Rowley, T., ed.: Oppida in Barbarian Europe, BAR Supplementary Series 11, 135-162.
- Cunliffe, B. (1988): Greeks, Romans and Barbarians: Spheres of Interaction, London.
- Cuniffe, B. (2005): Iron Age communities in Britain, 4th edition, London.
- Davies, J. A. (2008): The Land of Boudica, Oxford.
- Davis, O. (2013): “Re-interpreting the Danebury assemblage: houses, households, and community”, Proceedings of the Prehistoric Society, 79, 353-375.
- Elsdon, S. M. (1997): Old Sleaford Revealed: A Lincolnshire settlement in Iron Age, Roman, Saxon and Medieval times: excavations 1882-1995, Oxford.
- Evans, C. (2013a): “Delivering bodies unto waters: a Late Bronze Age mid-stream midden settlement and Iron Age ritual complex in the Fens”, The Antiquaries Journal, 93, 55-79.
- Evans, C. (2013b): Process and History. Prehistoric communities at Colne Fen, Earith. Bronze Age fieldsystems, ring ditch cemeteries and Iron Age settlement, Cambridge.
- Evans, C., Appleby, G. and Lucy, S. (2016): Lives in Land. Mucking excavations by Margaret and Tom Jones, 1965-1978: Prehistory, context and summary, Oxford.
- Fell, D. (2020): Contact, concord and conquest: Britons and Romans at Scotch Corner, Northern Archaeological Associates Monograph 5, Barnard Castle.
- Fernández-Götz, M., Wendling, H. and Winger, K., ed. (2014): Paths to complexity: centralisation and urbanisation in Iron Age Europe, Oxford.
- Fichtl, S.  (2005) : La Ville Celtique. Les oppida de 150 av. J.-C. à 15 ap. J.-C., 2nd edition, Paris.
- Fichtl, S. (2013) : “Les agglomérations gauloises de la fin de l’âge du Fer en Europe celtique (IIIe–Ier siècle av. J.C.)”, in :Garcia, D. ed. : L’habitat en Europe celtique et en Méditerranée préclassique, Actes Sud, Paris, 19-43.
- Fitzpatrick, A. (2001): “Cross-Channel exchange, Hengistbury Head, and the end of hillforts”, in: Collis, J. ed.: Society and settlement in Iron Age Europe. Actes du XVIIIe colloque de l’AFEAF, Winchester 1994, Sheffield, 82-97.
- Fitzpatrick, A., Haselgrove, C. & Hamilton, D. (2017): “Radiocarbon dating and bayesian modelling of the Late Iron Age cremation burial cemetery at Westhampnett (West Sussex/GB)”, Archäologisches Korrespondenzblatt, 47(3), 359-382.
- Foulds, E. M. (2017): Dress and Identity in Iron Age Britain: A study of glass beads and other objects of personal adornment, Oxford.
- Franconi, T. and Gosden, C. (2021): “Long-Term Interactions between Society and Ecology”, in: Gosden, C., Green, C., Cooper, A., Creswell, M., Donnelly, V., Franconi, T., Glyde, F., Kamash, Z., Mallett, S., Morley, L., Stansbie, D. and Ten Harkel, L., ed.: English Landscapes and Identities: Investigating Landscape Change from 1500 BC to AD 1086, Oxford, 107-148.
- Fulford, M., Clarke, A., Durham, E. & Pankurst, N. (2018): Late Iron Age Calleva. The pre-conquest occupation at Silchester Insula IX., London.
- Garland, N. (2020): “The Origins of British oppida: understanding transformation in Iron Age practice and society”. Oxford Journal of Archaeology, 39,(1), 107-125.
- Gerritsen, F. (2003): Local Identities. Landscape and Community in the Late Prehistoric Meuse-Demer-Scheldt region, Amsterdam.
- Hamilton, D. and Haselgrove, C. (2019): Exploring settlement dynamics through radiocarbon dating. ”, in: Romankiewicz, T., Fernández-Götz, M., Lock, G. & Buchenschutz, O., ed.: Enclosing space, opening new ground. Iron Age studies from Scotland to Mainland Europe, Oxford, 111-120.
- Hamilton, D., Haselgrove, C. and Gosden, C. (2015): “The impact of Bayesian chronologies on the British Iron Age”, World Archaeology, 47, 642-660.
- Hamilton, D. W., Sayle, K. L., Boyd, M. O. E., Haselgrove, C. C. and Cook, G. T. (2019): “ ‘Celtic cowboys’ reborn: application of multi-isotopic analysis (d13C, d15N, and d34S) to examine mobility and movement of animals within an Iron Age British society”, Journal of Archaeological Science, 101, 189-198.
- Haselgrove, C. (1982): “Wealth, prestige and power: the dynamics of late Iron Age political centralisation in south-east England”, in: Renfrew, C. and Shennan, S., ed.: Ranking, Resource and Exchange, Cambridge, 79-88.
- Haselgrove, C. (2000): The character of oppida in Iron Age Britain”, in : Guichard, V., Sievers, S. and Urban, O., ed.: Les processus d’urbanisation à l’âge du fer. Eisenzeitliche Urbanisationsprozesse: actes du colloque Glux-en-Glenne, 8-11 juin 1998, Bibracte 4, Glux-en-Glenne.
- Haselgrove, C., ed. (2016) : Cartimandua’s Capital? The late Iron Age royal site at Stanwick, North Yorkshire, fieldwork and analysis 1981-2011, Council for British Archaeology Research Report 175, York.
- Haselgrove, C. and Guichard, V. (2013): “Les Gaulois sont-ils dans la plaine? Reflections on settlement patterns in Gaul in the 1st century BC”, in: Krausz, S., Colin, A., Gruel, K., Ralston, I. and Dechezleprêtre, T., ed.: L’Âge du Fer en Europe: Mélanges Offerts à Olivier Buchsenschutz, Paris, 317-328.
- Haselgrove, C. and Millett, M. (1997): Verlamion reconsidered, in: Gwilt, A. and Haselgrove, C., ed.: Reconstructing Iron Age Societies, Oxford, 282-296.
- Haselgrove, C. and Moore, T. (2016): “Iron Age and Roman settlement in the Stanwick Environs”, in: Haselgrove, C., ed. 2016, 358-374.
- Hawkes, C. F. C. and Crummy, P. (1995): Camulodunum 2, Colchester Archaeological Report 2, Colchester.
- Henderson, J. (1991): “Industrial specialisation in late Iron Age Britain and Europe”, The Archaeological Journal, 148, 104-148.
- Hill, J. D. (1995): “The Iron Age in Britain and Ireland”, Journal of World Prehistory, 9,(1), 47-98.
- Hill, J. D. (1996): “Hillforts and the Iron Age of Wessex”, in: Champion, T., and Collis, J., ed.: The Iron Age in Britain and Ireland: Recent Trends, Sheffield, 95-116.
- Hill, J. D. (2007): “The dynamics of social change in Later Iron Age eastern and south-eastern England c 300BC-AD43”, in: Haselgrove, C. and Moore, T., ed.: The Later Iron Age in Britain and Beyond, Oxford, 16-40.
- Hill, J. D. (2011): “How Did British Middle and Late Pre-Roman Iron Age Societies Work (if they did)?”, in: Moore, T. and Armada, X.-L., ed.: Atlantic Europe in the first millennium BC: Crossing the divide, Oxford, 242-263.
- Hughes, G. and Woodward, A. (2015): The Iron Age and Romano-British settlement at Crick Covert Farm, Northamptonshire. Excavations 1997-1998, Oxford.
- Huisman, F. (2019): Wild wetlands and domestic drylands? Prehistoric communities of the East Anglian Fens in their broader regional context (c. 4000 BC – 100 AD). PhD Thesis, Durham.
- Jackson, R. (2012): Ariconium, Herefordshire. An Iron Age settlement and Romano-British ‘small town’, Oxford.
- Lambrick, G., Robinson, M. and Allen, T. (2009): The Thames through Time. The archaeology of the gravel terraces of the Upper and Middle Thames. The Thames Valley in Later Prehistory: 1500BC-AD50, Oxford.
- Malrain, F., Zech-Matterne, V. and Blancquaert, G. (2015): “Apprehending continuty and discontinuity in Iron age soil occupation and rural landscape through a collective database”, in: Danielisová, A. and Fernández-Götz, M., ed.: Persistent economic ways of living. Production, distribution and consumption in Late Prehistory and Early History, Budapest, 137-144
- Marshall, P., Brunning, R., Minnitt, S., Bronk Ramsey, C., Dunbar, E. and Reimer, P. J. (2020): “The Chronology of Glastonbury Lake Village”, Antiquity, 94 (378), 1464–1481.
- Masefield, R., Chapman, A., Mudd, A., Hart, J., Ellis, P. & King, R. (2015): Origins, development and abandonment of an Iron Age village (DIRFT Vol. II), Archaeopress, Oxford.
- Moore, T. (2006): Iron Age societies in the Severn-Cotswolds. Developing narratives of social and landscape change, BAR British Series no. 421.
- Moore, T. (2007): “Perceiving communities: exchange, landscapes and social networks in the later Iron Age of western Britain”, Oxford Journal of Archaeology, 26,(1), 79-102.
- Moore, T. (2011): “Detribalizing the later prehistoric past: concepts of tribes in Iron Age and Roman studies”, Journal of Social Archaeology, 11,(3), 334-360.
- Moore, T. (2012): “Beyond the oppida: polyfocal complexes and late Iron Age societies in southern Britain”, Oxford Journal of Archaeology, 31,(4), 391-417.
- Moore, T. (2017): “Alternatives to urbanism? Reconsidering oppida and the urban question in Late Iron Age Europe”, Journal of World Prehistory, 30,(3), 281-300.
- Moore, T. (2020): A Biography of Power. Research and excavations at the Iron Age oppidum of Bagendon, Gloucestershire (1979-2017), Oxford, Archaeopress.
- Moore, T., Braun, A., Creighton, J., Cripps, L., Haupt, P., Klenner, I., Nouvel, P., Ponroy, C. Schönfelder, M. (2013): “Oppida, agglomerations and suburbia: The Bibracte environs and new perspectives on Late Iron Age urbanism in central-eastern France”, European Journal of Archaeology, 16,(3), 491-517.
- Moore, T. and González-Álvarez, D. (2021): “Societies against the Chief? Re-examining the value of ‘heterarchy’ as a concept for examining European Iron Age societies”, in: Fernández-Götz, M. and Thurston, T. L., ed.: Power from Below in pre-modern societies. The dynamics of political complexity in the archaeological record, Cambridge, 125-156.
- Partridge, C. (1981): Skeleton Green, London.
- Potter, T. and Trow, S. (1988): “Puckeridge, Braughing”, Hertfordshire Archaeology, 10.
- Poux, M. (2014): “Enlarging oppida: multipolar town patterns in Late Iron Age Gaul”, in: Fernández-Götz, M, Wendling, H. and Winger, K., ed.: Paths to Complexity: Centralisation and Urbanisation in Iron Age Europe, Oxford, 156-196.
- Ralston, I. (2019): “The hillforts of Britain and Ireland – the background to the Atlas Project: an overview of the number of hill- and promontory-fort sites”, in: Lock, G. and Ralston, I., ed.: Hillforts: Britain, Ireland and the Nearer Continent. Papers from the Atlas of Hillforts of Britain and Ireland Conference, June 2017, Oxford, 9-27.
- Roberts, I. (2010): Understanding the cropmark landscapes of the Magnesian limestone, Archaeological Services WYAS, Leeds.
- Rogers, A. (2008): Religious place and its interaction with urbanization in the Roman era, Journal of Social Archaeology, 8(1), 37-62.
- Salač, V. (2012) : “Les oppida et les processus d’urbanisation en Europe centrale”, in: Sievers, S. and Schönfelder, M., ed. : Die Frage der Protourbanisation in der Eisenzeit. La question de la protourbanisation à l’âge du Fer, Vor-u. Frühgesch 16, Bonn, 319-346.
- Sealey, P. (2016): Where Have All the People Gone? A Puzzle from Middle and Late Iron Age Essex, Archaeological Journal, 173(1), 30-55.
- Semple, S., Sanmark, A., Iversen, F., Mehler, N., Hobaek H., Ødegarrd, M. and Skinner, A. T. (2020): Negotiating the North. Meeting-Places in the Middle Ages in the North Sea Zone, London.
- Sharples, N. (2010): Social relations in Later Prehistory. Wessex in the first millennium BC, Oxford.
- Sharples, N. (2014): “Are the developed hillforts of southern England urban?”, in: Fernández-Götz, M, Wendling, H. and Winger, K., ed. (2014), 224-232.
- Smith, A., Allen, M., Brindle, T. & Fulford, M. (2016): The Rural settlement of Roman Britain, London.
- Truscoe, K. (2021): The Landscape Settings of Three Iron Age ‘Territorial Oppida’ in Southern Britain: A Study Carried out Using Aerial Photographs and Lidar, PhD Thesis, University of Reading.
- Webley, L. (2015): “Rethinking Iron Age connections across the Channel and North Sea”, in: Anderson-Whymark, H., Garrow, D. and Sturt, F., ed.: Continental Connections: Exploring Cross-Channel Relationships from the Mesoithic to the Iron Age, Oxford, 122-144.
- Collis, 1984, 77-81.
- Fichtl, 2013, 77.
- Collis et al., 2000; Buchenschutz, 1995, 60.
- Barral and Nouvel, 2012; Fichtl, 2013.
- Salač, 2012.
- e.g. Moore et al., 2013; Poux, 2014.
- Fichtl, 2013.
- Haselgrove, 2000; Moore, 2020, 547-69.
- Cunliffe,1976; 2005.
- For more detailed definitions see Ralston 2019.
- Woolf, 1993; Haselgrove, 2000.
- Haselgrove, 2000; Moore, 2012.
- Hill, 1995; Haselgrove, 2016.
- Haselgrove, 2000; Cunliffe, 2005, 159.
- Hawkes and Crummy, 1995, 174; Fulford et al., 2018.
- Moore, 2020.
- Moore, 2017.
- Haselgrove, 1982; Cunliffe, 1988.
- Cunliffe, 2005, 593.
- Cunliffe, 1976, 145.
- Cunliffe, 2005, 402.
- Cunliffe, 1976; Haselgrove, 2000, 103.
- Cunliffe, 1988, 155.
- Cunliffe, 1976, 151.
- Henderson, 1991.
- Davis, 2013.
- Hill, 1996; Sharples, 2010.
- Fitzpatrick, 2001; Hill, 2007; Sharples, 2010, 170-171.
- Hill, 1995.
- Creighton, 2006.
- Webley, 2015.
- Fitzpatrick et al., 2017.
- Moore, 2006; Moore, 2020, 517.
- Sharples, 2010.
- Haselgrove and Moore, 2016.
- Brudenell, 2021.
- Bevan et al., 2017.
- e.g. Roberts, 2010.
- Moore, 2006; Huisman, 2019.
- Sharples, 2010.
- Moore, 2007.
- Hamilton et al., 2015; Hamilton and Haselgrove, 2019.
- Moore, 2020.
- Hamilton and Haselgrove, 2019, 117.
- Malrain et al., 2015; Bradley et al., 2016, 267.
- Gerritsen, 2003.
- Franconi and Gosden, 2021.
- Armit et al., 2014; Bevan et al., 2017.
- Lambrick et al., 2009, 379.
- Moore, 2006; Foulds, 2017.
- Cunliffe, 2005, 592.
- Moore, 2007.
- Moore and González-Álvarez, 2021.
- Moore, 2007; Hill, 2011.
- Smith et al., 2016, 81; Garland, 2020; Moore, 2020.
- Sharples, 2010; Haselgrove and Guichard, 2013, 321.
- Sealey, 2016.
- Hughes and Woodward, 2015.
- Moore, 2020.
- Bradley et al., 2016; Moore, 2020.
- Cunliffe, 1976; 2005.
- Haselgrove, 2000.
- Haselgrove, 2016.
- Collis, 2019, 83.
- Sharples, 2014.
- Lambrick et al., 2009, 362.
- Champion, 2007, 118.
- Collis, 2019, 83.
- Moore, 2020.
- Bryant, 2007, 73.
- Smith et al., 2016, 395.
- Moore, 2011.
- Marshall et al., 2020.
- Coles and Minnitt, 1995, 204.
- Coles, 1987; Coles and Minnitt, 1995.
- Evans, 2013a; Evans, 2013b.
- Moore, 2007.
- contra Henderson, 1991.
- Moore, 2007.
- Hughes and Woodward, 2015; Masefield et al., 2015.
- Lambrick et al., 2009, 108 ; Evans et al., 2016.
- Bryant, 2007, 63.
- Bryant, 2007, 64.
- Bryant, 2007.
- Fell, 2020.
- Jackson, 2012.
- Atkinson and Preston, 2015.
- Cunliffe, 1976; Elsdon, 1997; Cunliffe, 2005, 195.
- Davies, 2008, 124; Moore, 2020, 527.
- Cunliffe, 2005, 195; Champion, 2007, 121.
- Hill, 2007.
- Creighton and Fry, 2016, 339.
- Hill, 2007.
- Garland, 2020.
- Moore, 2020.
- Creighton and Fry, 2016, 340.
- Haselgrove, 2016.
- Garland, 2020.
- Haselgrove and Millett, 1997; Haselgrove, 2000; Rogers, 2008.
- Garland, 2020, 121.
- Garland, 2020; Moore, 2020.
- Moore, 2020.
- Bryant, 2007; Truscoe, 2021.
- Moore, 2012.
- Champion 2007 ; Davies, 2008, 124.
- Semple et al., 2020.
- cf. Cunliffe, 2005.
- Moore, 2011.
- cf. Hill, 2007; Garland, 2020, 121.
- Moore and González-Álvarez, 2021.
- Creighton and Fry, 2016; Moore, 2020.
- Cunliffe, 1988.
- Creighton, 2006.