6.3 Indigenous communities

6.3.1 Introduction

The impact of the Roman presence and proximity on indigenous societies has long been a major research question. As sections 6.1 and 6.2 indicate, it should be seen in the much wider context of the complexity of identities and perspectives among frontier groups, not simplistically as "Roman" and "native". However, the impact on Iron Age societies is a major topic which merits specific consideration.

A number of datasets offer ways into this question, in providing insights into local societies and their workings over this period, such as changing settlement patterns or landscape use. However, proposals arising from these areas have been contentious and it is hard to demonstrate the impact of Rome specifically as distinct from wider indigenous processes, especially where dating evidence is vague. The key source material for investigating Roman impact is the presence of Roman finds on non-Roman sites in Scotland.

The long tradition of study was started by the enquiries of James Curle (1913) and then given a foundation with his major catalogue of Roman finds from non-Roman sites (1932). This was updated by Robertson (1970) and again in summary by Hunter (2001), while Wilson has produced a series of regional catalogues, looking at a broader range of material but including Roman finds from Iron Age contexts (1997, 2001, 2003, 2010).

The lasting impact of Rome has often been debated, with a major division between those scholars who see it as a fundamental phase in Scottish early history which induced substantial social and political reactions and change in many fields (e.g. Fraser 2009, 116-7), and others who see it as a passing moment of little consequence (e.g. Hanson 2004). Evidence from elsewhere beyond the Limes and from wider anthropological studies on the impact of frontiers, especially 'asymmetrical' ones in terms of organisational complexity of neighbouring groups, seems to support the former view, but the investigation of such contact situations requires a subtle and nuanced approach, for the reality will inevitably be complex.

There is a long tradition of identifying causative connections with Rome for any change in indigenous society which has a possible chronological connection to the Roman period. Rome has often been seen as a cover-all explanation of changing phenomena. In some cases this has been valid, but in many cases scientific dating has shown this not to be acceptable - e.g. the abandonment of hillforts, the building of stone-walled roundhouses or the construction of rectilinear enclosures.

Roman finds from the Iron Age crannog of Lochspouts, Ayrshire ©NMS

6.3.2 Differing methods of analysis

Early works did not consider mechanisms behind the movement of this material in any detail. Macinnes (especially 1984) provided a key shift in interpretation, considering the rich Roman finds from the lowland brochs in the context of prestige-goods economies and socially-restricted access to this material. This attempt to embed the understanding of Roman goods in local societies was a major and highly influential step forward.

The broader study of Roman goods beyond the frontier has been dominated by areas with rich burial finds, such as northern Germany and Scandinavia (Wheeler 1954; Eggers 1951; Lund Hansen 1987). In contrast, Scottish finds are predominantly fragmentary objects found among settlement material, which has led to their significance often being overlooked or understated. Indeed, Alcock, coming from an early medieval perspective, argued that much of this material was residual, and may have circulated as tokens or charms long after the Roman period (e.g. Alcock 1963; Alcock and Alcock 1990, 115-6). Issues of taphonomy and life-cycle remain critical to such discussions and Hunter (2007a, 11, 91) has argued against Alcock's minimalist perspective, seeing a clear selection of material types. These make little sense as fragments but form a coherent pattern if viewed as the debris remaining from complete prestigious objects. Thus have methods been developed to use the presence or absence of types, categories and the variety of material present to tease meaning from these settlement fragments (Hunter 2001).

There is no doubt that the taphonomic issues are complex, a topic considered for ceramics by Campbell (2011). In some cases Alcock was right to suggest that some material was clearly long-lived, and Wallace (2006) has usefully noted that elongated use-lives are found regularly, especially for samian, within the Roman world. Samian also seems to have been a particular target of reworking and reuse in indigenous contexts. The picture is undoubtedly complex, but close attention to material condition and context provide ways to take this topic forward.

Increasingly subtle approaches are allowing more detailed appreciation of the potential meaning, and current approaches take advantage of this long tradition of cataloguing to consider local populations as an active agent in this interaction, and Roman material as a powerful social tool and catalyst (e.g. Macinnes 1984; Hunter 2001, 2007a). Key issues for consideration are the range of material and its social impact. The widespread distribution of Roman finds indicates its desirability to local societies across Scotland (suggesting "resistance" is too simple a concept), but with clear signs of local variation which probably reflect local social differences and attitudes. There is a marked selectivity in the material, with a strong bias towards types which were locally useful in displaying social status (notably jewellery and feasting gear). There are also signs of emulation or other forms of copying, for instance in styles of finger ring or rare ceramic vessel forms which show the influence of Roman forms (Campbell 2011; Hunter forthcoming).

Although in some areas Roman material was clearly widespread, it was not always abundant, and in southern and eastern Scotland shows a marked focus on only a few sites. These may be seen as local or regional power centres or elite sites - although regional variation indicates that the nature of any such 'elites' varied widely. Here Roman material serves a valuable role in making local social systems visible.

Detailed analysis has suggested changing patterns through time, and arguably the deliberate 'targeting' of particular areas or groups for Roman diplomatic attention, possibly in the face of local instability; this has been linked to social collapse in NE Scotland and the rise of the Picts (Hunter 2007a, 2010). This is a useful reminder also of the varied relationships which must be expected between the Roman and indigenous worlds, as circumstances and motives changed and it is in the context conflict and coexistence, that attribution of social and / or economic value to this material should be expected (and hints of these are apparent).

The increasing evidence for Roman interference in local politics at some stages, and the apparent desire for Roman goods on the part of local inhabitants, suggests this relationship is likely to have had significant social effects, and the continuing study of these undoubtedly will see much controversy to come. Macinnes' discussion of the lowland brochs in relation to their clear access to Roman material remains seminal (1984) and these sites and Traprain Law (Jobey 1976; Hunter 2009) are pivotal to understanding interactions at the upper end of the scale. In the case of Traprain, the existing archive (both field records and finds) hold considerable potential for renewed analysis, while further fieldwork on the site would be of great value. The most recent campaigns have confirmed the virtual absence of Iron Age activity on the site before the Roman period; how then should its emergence and pre-eminence in the Roman Iron Age be understood? Should potential parallels for power centres emerging in the context of client kingdoms, such as the oppidum of Stanwick (N Yorks), be sought)?

It is equally important to seek out sites which can broaden horizons - as the case of Birnie (Moray) demonstrates, an outwardly unprepossessing site marked out by the chance metal-detecting find of a substantial hoard of denarii (Hunter 2007c). Indeed, the investigation of "stray finds" has a key role to play, in giving them an archaeological context beyond simply strays, and (in some cases) in providing springboards to much fuller investigation and consequent information.

Another factor in the use of Roman finds was their value as raw material sources. This is one end-point of the processes of adaptation, which could also include the reshaping Roman vessels to more locally-useful forms, for instance, or adapting samian sherds as polishers or pigment sources. Remelting and reuse was clearly demonstrated by Dungworth's analysis of copper alloys (1996), and is often suggested as the source for the glass for beads and bangles; in the later and post-Roman period, and it is very likely that silver was recycled (Stevenson 1956).

Scotland is one part of a much broader picture, and comparison with other situations around the Roman frontier, and indeed analogy to frontiers of other periods, are important to contextualise and interpret the Scottish evidence of purpose, function and interaction.

One of the late second century denarius hoards from Birnie, Moray.©NMS

An updated and discursive corpus of Roman material from non-Roman sites is a key desiderata; such a volume has been commissioned for the Romisch-Germanisch Kommission's "Corpus der Romischen Funde im europaischen Barbaricum", though further funding is still required.

Detailed study of specific artefact classes by specialists can cast important fresh light on apparently intractable or supposedly well-known material (eg Erdrich et al. 2000; Ingemark in press).

Lists of coin finds in Scotland have been an invaluable resource back to the days of Haverfield and Macdonald. The National Museum no longer employs a numismatist, leaving only one specialist in Scotland and putting this tradition at risk. Efforts are needed to ensure a continuing publication of the material.

The investigation of the impact of different frontiers (e.g. Hadrian's Wall cf Antonine Wall), the differential and long-term impact either side of a frontier (e.g. Hadrian's Wall), and broad comparative perspective to other frontier areas.

It is vital that the Roman material is considered in context, not in isolation - Roman material forms but one part of indigenous material culture and needs to be considered alongside this.

The life-cycle of the Roman material (covering evidence for its arrival, reception, modification, reuse, emulation and deposition) needs closer attention than it has traditionally received. There is a need for close study of taphonomy, from both object condition and site context, to understand life cycles of the artefacts.

Traprain is a pivotal site for understanding interactions with the Roman world. Full publication to modern standards of the existing assemblage, and further fieldwork to clarify the sequence and expand knowledge of the site, are long overdue. Why did Traprain become so prominent in the Roman period?

Significant stray finds should be followed up in the field wherever possible in order to retrieve their context. This need not always involve excavation - but study of existing aerial photos, geophysical survey results and fieldwalking can provide an understanding of the setting of such finds, and in many cases guide attention to sites which were previously unknown.



6.2 Social identities

6.2.1 Background

Carriden altar ©SCRAN and NMS

The theme of 'changing worlds' addresses the experiences and the impact of Empire on the daily life course of the inhabitants of northern Britain during the Roman period. It deals with the subtle negotiation of social identity, encompassing categories such as age, gender, ethnicity and status (Mattingly 2004). Increasingly Roman frontiers are viewed as pluralistic in nature, socially and ethnically diverse (e.g. Collins 2006; Cool 2004; Gardner 2001, 2007a, 2007b; Hingley 2004; James 2001; Okun 1991; Swan 1992, 1999, 2002; Wells  2005). This is not a new concept; there has long been discussion on who exactly peopled the frontiers (eg Curle 1911; Richmond and Steer 1957; Salway 1965; Birley 1980; Hanson and Maxwell 1986, 182-92). However, little of the recent work has directly considered the Scottish evidence, with the notable exception of Vivien Swan's work on Antonine pottery (Swan 1992, 1999, 2009). Yet there is substantial potential in this area. The very specific and well-understood temporal framework of Roman Scotland, with relatively clear phases of activity, has the great benefit of simplifying the datasets, creating a series of discrete case-studies which are not muddled by the complex rebuilding histories and extended periods of occupation common on Hadrian's Wall, for example. The biggest issue in addressing this material is to have the right questions in mind, as any enquiry will be limited by the way it is approached.

The epigraphic evidence has long been the first port of call for students of Roman Scotland, and it serves to highlight the diverse and complex reality of identity. The altar erected by the vicani, the inhabitants of the village, at Carriden attests to the presence of a civil settlement associated with the fort (RIB 3503; Richmond and Steer 1957). From Shirva (probably deriving from Auchendavy fort) comes the tombstone of Salmanes, erected by his father, who shared his name which is of Semitic origin, and Salmanes has been seen as a Near Eastern trader, since the inscription lacks any clear military link (RIB 2182). An altar at Westerwood (RIB 3504) was erected by the wife of a legionary centurion, Vibia Pacata; her name suggests she came from north Africa, while her husband was from Hungary. There is also a wealth of information relating to the soldiers themselves in the form of dedications and graffiti; Nectovelius, for instance, who was buried at Mumrills, served in a unit of Thracians but was himself a Brigantian (RIB 2142) while the dedications of Marcus Cocceius Firmus at Auchendavy emphasise the wide ranging career of officers in the Roman army (Birley 1953).

Place of origin is only one part of an individual's or unit's identity. As Swan's research (1992) has highlighted, it is not only a unit's origin that is of interest, but also its various deployments, each of which could influence it. Some units did maintain strong ties with their point of origin, notably Batavians (Roymans 1999; 2005), but, as the Mumrills inscription shows, while a unit may be raised in one area, it could rapidly lose any ethnic character in subsequent generations of more localised recruitment where they were garrisoned.

Epigraphy may provide a taster of the complexity, but a focus on it introduces a significant bias, as inscriptions are more common in certain periods and places, and are selective in terms of the class and status of those that commissioned them. There is a growing realisation that social identity (including ethnicity) had an impact on social practice in the fort which is archaeologically detectable (Bruhn 2008, Collis 2008, Cool 2004). This may be achieved by comparing categories of material (for example, by as types, proportions, and distribution) within and between forts, between forts and their annexes and vici, with surrounding farms, with neighbouring frontiers, and so forth.

The inscriptions noted above attest to the presence of women in and around Roman forts, a topic which saw early discussion, with Curle (1911) addressing the material from Newstead and Macdonald (1906) attempting to rationalise the presence of women's shoes at Bar Hill. While there are larger current debates about gender on the frontiers within Roman studies, and few accounts of Roman Scotland have attempted to address this issue in any great depth (cf Allason-Jones 1989, 1999; Allison 2006a, 2006b; James 2006; Van Driel-Murray 1995). There is a need to address the issue of women and children in the frontier zone by scrutinising the available material. Equally, they were not the only non-military individuals in the area, as the case of Salmanes suggests - a range of specialists such as traders and craft-workers as well as people drawn to the fort by family connections, and veterans who may have settled in familiar territory during the occupation (a group increasingly seen as key active social agents in affecting the impact of Roman culture; see Groot in press a) are to be expected. However, none of the Scottish vici can be shown to have outlived the military occupation.

Tombstone of the soldier Nectovilius, a Brigantian who was recruited into the Second Cohort of Thracians. From Mumrills ©NMS

Yet there are obvious traps here, notably a tendency to focus on the unique or different and to exclude the obvious from the discussion of social identity. When gender is discussed, the focus is on finding and defining women; the family elicits discussions of children; while with ethnicity the focus is on the exotic, such as the study of Africans and Syrians in the frontiers. A whole group of individuals too rarely feature in such debates: the 'common soldiery' (Haynes 1999). They are treated as the default setting, often layered with modern preconceptions. Yet there is, for instance, significant evidence to address gender in the frontier zone through the study of masculinity (Gardner 2007) and, as noted, diversity within the 'ordinary squaddie' should be expected and sought.

Other sources of evidence can cast light on frontier mentalities. Ferris (2000) has considered the iconography of Antonine Wall distance slabs, with their stereotypical depictions of naked, conquered barbarians. This offers insights into the perception of its formal enemy by the Roman military, although other evidence of interaction indicates an altogether more complex picture in reality (see 6.3 and 6.4 below).

6.2.2 Relationships

One way of conceptualising and investigating this complex topic is as a series of relationships (Table 3). Any such attempt at summary is inevitably a simplification, but it serves to highlight the diversity of relationships within any frontier community, taken here from the soldier's perspective. Although there are difficulties in investigating detail, many aspects of this are susceptible to archaeological investigation at some level.

Table 3: a model of relationships between people in the frontier zone, from a soldier's perspective

Relations of ... Such as...
Profession The army / wider military community The unit (at various levels - unit, century, contubernium) Rank Social class (careers of officers) Links to other units and/ competition, different unit origins, traditions and histories; detachments Veterans
Kin Current generation of family (common-law wife, children, nearby relations) Previous generations (veterans; more distant relations from homeland)
Sustenance Traders, craft-workers, suppliers of diverse origins and trades (shops, taverns, brothels etc) Slaves Indigenous population
Occupation Indigenous population (friendly, indifferent, hostile)

6.2.3 Advancing the debate

There are a number of ways forward in addressing social identities in Roman Scotland. The key thing is to approach the concept and data with an open mind and be willing to ask different questions of the existing data, building on the valuable datasets which are available. A good example is Swan's (1999, 2009) and Willis's (1997, 1998) work on pottery, which has highlighted its potential beyond mere dating evidence. The definition and identity of these groups, 'the army', 'civilians', 'locals', was not static; in these complex social settings, forms of hybridisation could occur between different social groups in what may be termed a frontier culture (Lightfoot et al. 1998; Hunter 2008), which may differ in different frontier zones.

A comparative approach is important here. How do different forts, or different areas of a fort, compare to one another? Can a difference between different garrison types, between primary / secondary Wall forts, or between different periods of occupation be seen? An area of great potential is studying the distribution of material. Hoffmann (1994) highlighted the potential of studying the distribution of finds in legionary barrack blocks, while Stoffels (2009) has looked at the distribution of locally-produced pottery within Roman military contexts; this approach could be readily applied to auxiliary forts to address issues of ethnicity, status, rank and gender. Are the high status goods coming from a specific area? Samian, for instance, was noted to be concentrated in the officers' quarters at Bearsden (Breeze 1977). The work of Pim Allison in using GIS to analyse spatial patterns in Roman forts is directly relevant here (Allison 2006a, 2006b, Allison et al. 2005). She has focused on highlighting the presence of women in these social spaces, but the techniques could be expanded to address all the elements of social identity; although an important caveat is the need for careful study of the taphonomic pathways of finds into deposits, which in many cases may reflect demolition activities rather than use of the space.

A number of recent excavations present a good opportunity to address these topics: Strageath, Cramond, Elginhaugh, Inveresk and Newstead. In the particular cases of Inveresk and Newstead, the availability of assemblages from both inside and outside the fort walls gives a tremendous resource for comparative study. There is a great potential in the archaeological record of Roman Scotland to address these issues and it should be embraced enthusiastically.

Use of the existing resources may be able to illuminate more complex understandings of social identities on the Roman frontier, with analysis of the material from different forts, comparing the types, amounts, proportions and distribution against one another.

For more recently-excavated sites, the spatial distribution of finds offers tremendous potential, comparing between areas and buildings to see if different social spaces can be identified. Suggested sites include Strageath, Cramond, Elginhaugh, and Bearsden.

Inveresk and Newstead, where excavation has ranged widely across the fort complexes, are particularly rich in potential for study. Publication is awaited for Newstead, and for one of the major Inveresk excavations. These are key sites for further analysis of existing datasets. Such an approach needs to put the Scottish material in context. Comparison is needed to the Hadrian's Wall area and to material from other frontiers around the Empire.