Searching for Identity in Italian Landscapes

Martin Sterry

Introduction The study of identity has, arguably, seen little direct attention in Roman landscape archaeology even though it has become an issue of rising importance in other branches of Roman archaeology (e.g. Mattingly 2004.). Here it is considered that people have multiple and fluid social identities that are expressed or performed in different arenas (Meskell 2001). Certainly, Witcher (2006a) has highlighted the lack of theoretical and interpretative development in Roman landscape archaeology, particularly in the survey projects that drive this field. He specifically targets the debate on identity (ibid. 41, 58–59) and draws attention to the lack of engagement with this debate on the part of landscape archaeologists. Whilst some landscape archaeologists have considered the issue of identity (e.g. Petts 1998; Alcock 1993; 2002; Roymans and Gerritsen 2002), predominantly they are concerned with the development of a pre-defined community identity. rather than with the multiplicity of competing identities that can exist within a community. Such an understanding of identity and landscape is contradicted by recent work in archaeology and anthropology that has shown how identity is intimately linked to the landscape (Bender and Winer 2001; Low 2002; Blake 2005). I argue that identity is fundamental to an archaeological understanding of all areas and periods because it is the means by which we characterise the people of the past and interpret their relationships and actions. It allows us to move away from bland and arguably meaningless generalisations of peasant and elite, Roman and native, colonizer and colonized, towards a more nuanced understanding of past social realities. Indeed, if we do not engage with the metanarratives of identity (Meskell 2001: 188–9) then our ideas of landscapes become disconnected from the very people that lived in and created them. This paper is a brief attempt to respond to Witcher’s challenge, and to offer a route for understanding and utilising the concept of identity within survey projects, and particularly in the historical context of Roman expansion.

In Italy, survey projects are the most prominent form of archaeological investigation with regard to the protohistoric, Roman and early Medieval periods, yet also they are one of the most widely criticised for the conservatism of their approach (Alcock 1993; Alcock and Cherry 2004; Witcher 2006a). Site density and simplified site hierarchies form the basis for discussions of population dynamics and the economic well-being of different areas without engaging with the “richness and complexity of the available information” (Alcock and Cherry 2004: 5). This approach inevitably leads to a survey being discussed as an irreducible unit, often as a valley. Too often survey projects, if they consider identity at all, reports deal with identity in a simplistic manner, along the lines of – “the area of the survey lies within the territory of group x” or “the Roman presence in the landscape expanded busily in the second century A.D.”. These assumptions are largely derived from the secondary historical literature and treat identity as an historical fact, not a theme for discussion. However, as this paper will seek to show, identity has an important role in clarifying the interactions people have between each other and their environment. We cannot hope to fully understand the economic fortunes or any other aspect of an area without attempting to understand the people – who they were and what choices they made. In this paper I will first outline my views on culture histories and maps in reference to central Italy in the later first millennium B.C.; I will then develop my ideas via a case study based on the Iuvanum Survey Project from central Italy that shows that identities can indeed be accessed, considered, and debated through a landscape approach which integrates field survey data in conjunction with excavations and standing remains.

Culture Maps and Ancient Historians Maps of tribal territories such as Salmon’s map of central Italy are derived solely from the works of ancient authors such as Vergil, Ptolemy, Pliny and Strabo (Salmon 1967: 24–5; Fig.1). However, the descriptions of ethnoi given by these authors are conflicting (Dench 1995:175–217) and as they refer to a certain moment or perceived history from an individual point of view, they are not applicable to the loosely dated material typically recovered by field survey. Yet Salmon’s map has provided a substitute for a discussion of identity for more than one survey, such that in the modern region of Molise, landscape units have been described as ‘Pentrian territory’ and the ‘Frentanian [lower] valley’ (Lloyd 1995: 182, 199). Peter Wells (2001: 25–27) and others have argued against describing material culture in such ethnic terms as ‘a Celtic sword’ or ‘a Germanic fibula’ on the basis that it tells us little about the people that used them. Surely it follows that describing an area and everything within it as ‘Pentrian’ or ‘Samnite’ is not only unhelpful but obscures the wide range of identities that people may have held. My complaint is not that the geographical correlates are imprecise, but rather that given the complexity of lived social identities these terms obscure as much as they reveal. Ethnicity in particular is a highly political concept, and individuals will describe themselves and each other differently depending on who is asking and when.

Figure 1. The Samnites and their neighbours c. 350 B.C. (adapted from Salmon 1967: Map 1).

Siân Jones (1997: 30–31) has criticised the creation of culture maps for their inaccurate conceptualisation of ethnicity as commensurate to the politicised groupings of tribes and chiefdoms. They present an essentialist view such that an ethnic group may be defined by a combination of shared history, mythology, language, religion and culture. The evidence for this is dominated by the literary record, but, as Kathryn Lomas (1997: 2) has argued, there are very few self-perceived views from most areas of Italy since most of the sources date to the period of the early Empire and are all external perceptions of local cultural identities (see also Bradley 2000). Ancient historians have been largely successful in emphasising the diverse nature of Italic communities, their willingness to accept outsiders and their changing political makeup (e.g. Bradley 1997; Dench 1995). However, there are key concepts that the historians have been slow to relinquish. Their approach is very much dominated by ideas of territorial division but with little reference to its purpose, use and effects (see Delaney 2005), and the nature of their evidence – i.e. literary sources, epigraphic sources and elite material culture – restricts their discussion to the social groups that could have created this evidence: predominantly the social elites. The sense is that they desire only to fuzz the boundaries of culture history, rather than to reconceptualise what identity is and how it was used by all members of society.

I do not advocate dismissing the evidence of the texts as there is much work in Italy and other parts of the Roman Empire that is very useful and informative. But it must be recognised that these provide only one view at one scale and that this scale is different to that at which many landscape studies work, particularly those based on survey projects. After all, field data reveal all levels of social scales, even those not mentioned in texts, and can potentially cover all parts of a particular landscape, not just the cemeteries and sanctuaries linked with elite display

Identity and Ethnicity In other archaeological spheres and the social sciences in general, identity is discussed in a very different manner. Critically, ethnicity is not the same as identity: Lucy (2005: 95) thus describes it as an aspect of a relationship. It can be used as a tool, consciously or otherwise, to emphasise perceived cultural differences and to differentiate between ‘us’ and ‘them’; therefore, like the relationship of which it is part, it is malleable and negotiable. Depending on the nature and importance of the situation it can be utilised in ways that are linked to other aspects of identity, such as social status, religion or gender, or not at all. Each individual will have different notions of who is part of their ethnic group and who is not, depending on who they are dealing with: a leader drawing power from their followers, a farmer negotiating the use of land with his neighbours, a merchant trading with disparate groups over long distances, or a spouse drawn from an outside community.

Ethnicity waxes and wanes in importance in relation to a range of social factors including political, economic, ideological and demographic pressures (Díaz-Andreu and Lucy 2005: 9–12). The goal should not be to speculate what the ethnicities of different groups were, because these are not distinct entities, and their numbers, spatiality, and temporality will vary depending on the individual concerned. There is no irreducible ethnic ingredient that can be uncovered. The question is how far archaeology can go towards teasing out such levels of identity. Siân Jones has convincingly argued: “There is no unambiguous ethnic association [to material culture] because no such single social reality has ever existed” (1997: 140). It is not possible or particularly useful simply to look at an artefact and say this was made, used or deposited by a member of an ethnic group on the basis of its form, date or location.

Archaeological studies of identity have been successful in emphasising the diversity of ideas and of experience in the Roman world, but they are often lacking in spatial context or narrowly focussed in the micro-scale of cemeteries or household. This can surely be expanded to the regional scale through the study of landscapes and relationships of the past, identities are not shared, but individual, complex and ambiguous, contrary to what is implied by many field projects. Furthermore, ethnicity should not be confused with territories or state boundaries, but should be considered in a world of multiple potentially conflicting identities. This is further problematised by the possibility that an individual may manipulate their identity as he or she moves about and interacts with different people. For example, a young Samnite warrior may portray himself as a native of his town or state in opposition to Rome when at home, but on campaign in Spain he may portray himself as part of the Roman state in opposition to Celtiberian tribes. As epigraphs show, ‘Romans’ themselves often identified most with their towns or regions than with the ‘state’. However, ignorance does not make these distinctions negligible. At different scales, different aspects of identity become more prominent as is apparent in the ancient sources. It is this type of ambiguity that should be applied to the landscapes forged and worked in central Italy, an ambiguity which stresses a plurality of views, interpretations and ways in which identities were used. The question thus becomes: how does identity affect the way that people act and interact, through forming social relations and in constructing their own landscapes?

Identity and Landscape To take this discussion further, landscape must be more closely defined. Landscapes are the world as actively understood, experienced and engaged with by individuals; landscapes are individual and subject to constant reinterpretation (Bender 2006: 303–4). If identity as discussed above influences what people do, where they go and who they interact with, then it will in turn influence their landscape. The relationship is reflective because the landscape, the world that people perceive, will constrain and influence where people go and what they do. Whilst a Roman colony may be created in a specific form partly because of the identities of the colonists (as suggested for Ostia by Bispham 2000), it will be continuously reinterpreted in multiple and potentially conflicting ways by all the different people that come into contact with it. The physicality of the colony has a role to play in framing relationships and disputes between those who might view it as a place to live, an overnight stop, a trading centre, a new religious focus, a place of control or a disruption of the old order. Over time, however, the colony’s role in peoples’ landscapes will change and the material of the colony will likewise change with new ideas and new activities. But whilst everyone’s perceptions will be unique there should also be similarities in their interpretations. The ambiguity and plurality of these experiences can be managed in archaeological interpretations, but it is important not to dismiss them. It is linking this kind of approach to what the pot-sherd can potentially tell us that is the key to exploiting the survey evidence.

Some of these issues are starting to emerge in other work. As already mentioned, Witcher (2006a) has argued forcefully for a new interpretative framework for the study of Roman landscapes. His approach has been to characterise a large number of surveys in Etruria to look at social change on a very broad scale both spatially and temporally. He argues that it might be possible to see changes in social relations and social tension from the overall quantity of material in circulation (Witcher 2006b). This is one possible way of considering the development of identities through time, but it still does not fully address the criticism that identity should not be rooted in places, but rather in the people that moved through the places. Examples of how this can be considered can be found in the literature of landscape archaeology outside of the Roman sphere. One valuable example relates to the 19th-century homestead landscapes of southern Colorado. Church (2002) describes how the region was colonized by Anglo and Hispanic families. Their different identities affected their perceptions of how to utilise the land: the Anglo-Americans arriving from the east in the 1860s followed homestead law in which each settler was awarded a 160 acre plot on a grid system that emphasised uniform land-use with little concern for available water; conversely, the Hispanics, whilst also bound by homestead law, clustered in plazas along rivers with access to several plots so that they could work the land in a mixed fashion. However, by the 1880s the distinction between the ethnic groups was less pronounced: Anglo land holdings developed to include access to water and instead the difference between open ranching and bounded agriculture became the important distinction of identities and conflict as both groups sought access to water. This example demonstrates how land-use was originally determined by the identity of the settlers, but that over time the ways in which people used the land developed their identities, leading to a conflict between ranchers and farmers. The landscapes recoverable archaeologically are expressive of identity, but it is a recursive relationship – as the landscapes were expressed so they changed the identities, leading to new landscapes. Bringing these ideas back into the realm of Roman landscape archaeology, it should be recognised that a more textured approach is required, one which draws out the multiple landscapes that are connected to different identities and overlap through their relationships and conflicts (Thomas 2001: 181–2). Thus, instead of snapshot-like maps of sites at different periods, there needs to be an attempt to show the developments and constant changes in the material record as peoples’ landscapes reflect their changing experiences of the past.

In Italy, there are clear aspects of the environment that would have defined identity which have seen discussion. Centuriation, the regulated divisioning of land plots, is perhaps the most obvious as it links concepts of an ‘imperial’ state-ordered identity and territory to the archaeological evidence for landscapes (Attolini et al. 1991; Dall’Aglio and Marchetti 1991). Yet, if a closer look was taken of land-divisions would they have always been used in the way that they were originally intended? They may initially have been seen as a symbol of the colony, with owning and working plots a means of personal identification, but not all colonists were good at farming, and hereditary transfers, sales, and illegal squatting could have changed the associations with identity. As colonists learnt their landscape, they might have altered their plots to make better use of the environment: irrigating fields with floodwaters, or providing shelter for livestock, thereby disrupting the regimented vision of a centuriated landscape. Others may have viewed the boundaries in different ways, walls as barriers to the free movement of flocks or perhaps a ditch might be seen as a temporary haven by a slave. It would be a mistake to consider identity only in relation to the most visible elements of Roman colonisation. By considering the way that people moved through, used, and lived in their landscapes, it is possible to reflect on aspects of their identity beyond the dualism of Roman and native – such as different labour roles, social status and religion.

Identity in Field Surveys The environment in which people live creates formal associations that frame the identities people have (Meskell 2001: 189), affecting what they do and the landscapes they make. This could be attributed to specific aspects of identity or more generally to how identities related to different groups or communities. The critical point is to consider the environment not as a distribution of sites or settlement against a geological background, but as an arena in which activities take place at different times involving different groups. The environment needs to be peopled in order to better our understanding of identity.

The weakness of many field surveys is thus the way their interpretations interchange between sites and people, creating a rather static social order based on a site hierarchy. To assume certain communities with monolithic identities and normative boundaries inhibits any recognition of the concepts of identity that I have discussed (see also Cohen 1985), and serves to eliminate those that are not tied to one specific site – e.g. itinerant potters, merchants and transhumants. No-one would suggest that there was no interaction between these groups, but it is rarely addressed as more than a side issue. Instead it is more appropriate to consider that the ‘community’ relates to the social or settlement centre in the same way that the household relates to the house (Van Dommelen et al. 2005: 56–57; see also Allison 1999).

The notion of communities in the landscape, deserves some exploration and in recent years there has been a flurry of scholarly attention in archaeology (e.g. Canuto and Yaegar 2000; Gerritsen 2003; 2004) such that a whole section was devoted to it in Papers in Italian Archaeology VI (Attema et al 2005.). Yaeger and Canuto (2000: 5–9) argue for a model based on social interactions that is articulated in Yaeger’s study of Mayan San Lorenzo (2000) through the role of activity groups (such as feasting and house building) and shared understanding of practices ( such as through the shared use of a quarry). Gerritsen has built on this work to argue that communities should be analysed as a “symbolic construct of identity” (2004: 147); in particular he is concerned with the idea of belonging: “how localities are part of one’s identity” (ibid.). This allows for an effective analysis of the development of a particular community over time as his analysis of Iron Age Southern Netherlands shows (Gerritsen 2003; Roymans and Gerritsen 2002). Whilst Gerritsen argues that the local community is cross-cut by other identities, this aspect fails to come out of his case study and the focus on a single ‘community identity’ could be argued to reify the culture maps that were the starting point for this critique. Perhaps more helpful is the approach of Tyler (2006), who sees the construction of communities as a device for exercising power and maintaining social networks. Interests, values and practices are shared through these networks using typical forms of behaviour and sanctions against those who act outside the community. An individual’s history, interests and practices mean that they will belong to a range of communities whose memberships (e.g. familial, religious, professional or political) may overlap, but are not identical and are therefore not concentric in their organisation (Lewis 2006: 70). Communities should not be directly linked to identities, so one community does not equal one identity. Nonetheless, they are useful for providing a framework in which identity is used. For example, the community utilising a sanctuary might contain landholding members and agricultural workers whose interactions with each other can be traced at the sanctuary, in the fields and with members of other communities.

I would argue that the key is to move from an archaeology of settlement structures to one of activities and relationships, with sites as the evidence, but not the limits, of this archaeology (e.g. Terrenato 1998). Identifying the who, when and where of the activities carried out in the landscape will show how different people and groups of people came into contact with each other and the characteristics of these meetings (Neustupný 1998: 21–33). The rhythm of these activities – the length of time they take, their frequency and how they relate to one another – is very important to help understand the changing nature of identity and how it can be utilised in different ways by the same people. From this, the social networks that existed can be drawn out and the construction of communities better understood in order to examine their internal connectivity and relations with other communities.

The Case Study of Iuvanum, Abruzzo My own doctoral research aims to consider changes in landscapes, landuse, people and materials across central Adriatic Italy in the Iron Age to Roman periods. Core to my research is a questioning of identity. From materials so far gathered, here I draw upon one case study which helps demonstrate how this proposed new approach to tracing ‘identities’ might be applied. It should be noted that analysis is still at a preliminary stage. Situated in the southern Abruzzo region of central Italy, the site and environs of ancient Iuvanum is rich in archaeological evidence for many periods and recently has been subject to a fieldwalking survey – the Iuvanum Survey Project (ISP) (Bradley 2005; Menozzi 2003). The survey, along with the excavation of Iuvanum itself, has recovered a wealth of information for understanding the development of a Hellenistic sanctuary site into the municipium of Iuvanum. Characteristic of the majority of the sites recovered was evidence of activity from at least the Hellenistic period to the Imperial period, but with important developments across time.

The ceramics recovered by the survey provide excellent evidence of different consumption practices that provide a useful starting point to begin to seek out communities and identities. Throughout, coarsewares and regionally produced finewares dominate the assemblages, but the proportions in which they occur, their forms and the items that are imported give some clue to the heterogeneity present across this region.

In the Hellenistic period the sanctuary acted as a religious focus around an important spring (Fabricotti 1990), it is likely that markets and fairs were held in this place and it may have been a stopping-off point for transhumants and travellers (Pasquinucci 1990). However, some other aspects of power seem to have been linked to the monumentalisation of three nearby hills with short stretches of polygonal walling – Monte Pidocchio, Monte di Maio and Montenerodomo. Whilst the evidence for sites in the region during this period is wide-ranging there are few patterns that can be drawn from the assemblages. Small cups and skyphoi produced in regional Black Gloss appear to be an important component, and assemblages recovered from the hilltop sites and Monte Pidocchio in particular reflect this interest. A few sites do stand out, especially that of ISP5006, some 5km to the south of Iuvanum, which is very large and has higher proportions of regionally-produced fineware and storage dolia. This may have been one of the villages widely attested for this area (e.g. Lloyd 1995) and/or the residential area of a prominent family estate in this region. The rest of the assemblage is constituted of bowls, jars and open vessels in a mixture of impasto, coarseware and buffware. The interpretation for this period may then be one of a community or communities with different centres of interaction – the sanctuary, the hillforts, and larger farms (perhaps some of the larger survey sites). Therefore, we might think of a fractured group dynamic with a strong emphasis on rural activities, such as farming, that might have been used in the creation of identities with associations with the sanctuary creating a different identity apart.

In the Roman period a clear distinction emerges between those sites close to the town (within 1.5km) and those outside. The ‘inner’ sites have significantly higher percentages of amphorae, Italian Terra Sigillata and other imported wares (Fig. 2), similarly the assemblages have higher numbers of forms that could be used for serving food and drink or specialised food production – platters, flagons, jugs, and also the specialised clibanus (used in bread baking) (Fig. 3). In contrast, the ‘outer’ sites have more balanced assemblages that seem more utilitarian in nature, and it is interesting that the evidence for weaving (nine ceramic loomweights and spindlewhorls) also derives chiefly from these outer sites. It may be hypothesised that after the formalisation of the site of Iuvanum as a town with a monumental forum, basilica, workshops, and tabernae, there was a direct effect on the nature of rural settlement. Two possibilities stand out: that these represent the dwellings of prominent individuals with more expensive and cosmopolitan consumption tastes for whom proximity to the town was more important than to the production areas of the countryside. Alternatively, the area could plausibly have become more frequented by travellers with the reorganisation of territories in the first century B.C.; some of these sites may then have had a service role, especially site ISP2013 for which there were a significant number of flagons, jugs and amphorae. Either way the building of the town can be linked to a refiguring of landscapes for many of the inhabitants of this area as they started to use it more frequently and for more activities. This should not be seen as a totalising phenomenon though, as the largest and richest assemblage still comes from the site (ISP5006) that was so prominent in the Hellenistic period, and perhaps this relates to continuous ownership of a farm/estate and the adoption of partially static identities linked to property and place.

It is also apparent, although beyond the scope of this paper, that in addition to changes in imports there were general changes in cooking and serving practices that played an important role in figuring people’s identities and the places from which they procured their resources. From the point of view of the community in the Imperial period, there seems to be a stronger focus based around the site of Iuvanum, but this is one which is more heterogenous in nature that can perhaps be attributed to new market-focussed identities developing from the changing political and economic situation of the region.

Discussion and Conclusion Survey data need not be homogenised and described in landscape units. The recent surveys nearing final publication, such as the Sangro Valley Survey (Lloyd et al. 1997) and the Iuvanum Survey Project (Bradley 2005; Menozzi 2003; Menozzi and Fossatoro 2006) with their more intensive survey and greater analysis of the finds, provide an ideal opportunity to develop this approach. Rather than dealing with the specifics of habitation around Iuvanum, I have tried to demonstrate that identities played a role in the perception and changing use of the sanctuary and later town. The use of the town re-figured and re-oriented the identities and communities of the people around it. Distinctions appeared between the consumption activities of those living close to and in the town, as well as those living further away. There were new places and ways of interacting despite the presumed continued access to and circulation of key resources, such as the clay needed for pottery production. To further draw out the differences in identities it might be useful to consider the territorial aspects of the town, the survey sites and the communities that occupied them, and the changes in who was excluded, how practices were controlled, and how people passed through the area. This can be done through looking at the cultural and natural environments of sites, thorough analysis of archaeological assemblages and faunal or botanical data, or by one of a host of other archaeological methods. The case study presented is very much a work in progress and no doubt ideas and problems will develop as my research evolves, but even at this stage it is clear that Roman landscape scholars need to move beyond traditional assumptions of who they are looking at in the field.

Figure 2. Imported ceramics per hectare in the Iuvanum region during the Roman period. Based on unpublished data from the Iuvanum Survey Project (permission obtained from Bradley and Menozzi, figure drawn by author).

Figure 3. The relative percentage of flagons, jugs, platters and clibani in the Iuvanum region during the Roman period. Based on unpublished data from the Iuvanum Survey Project (permission obtained from Bradley and Menozzi, figure drawn by author).

I have argued here that the study of Roman landscapes, particularly in Italy, does not adequately confront the issue of identity. However, this view could equally be reversed to suggest that discussions of identity do not adequately take into account how these identities are manifested, associated with and used in the landscape. The archaeology cannot be led by text-driven discussions of ethnoi and cultural maps because they are inherently flawed with their lack of time depth, their limited point of view, the fallibility of the sources, and their literary scale. The conception of identity as composite, layered and conflicting and rooted in activities and relationships is not novel, but needs fuller appreciation. As has been argued in other areas of archaeology and anthropology we need to look in more complex fashion at complex ideas of identities and landscapes together. This will encourage more humanised approaches to the discussion of economy, land-use and demography to create a more textured, ambiguous and conflicting vision of the past where not everybody did the same thing.

School of Archaeology and Ancient History, University of Leicester

Acknowledgements I wish to express my sincere gratitude to Guy Bradley and Olivia Menozzi for their permission to utilise unpublished data from the Iuvanum Survey Project. My thanks to Neil Christie, Dave Edwards, Colin Haselgrove, Thomas Leppard and David Mattingly for reading earlier drafts of this article and providing invaluable help and advice.