The quick and the dead in the extra-urban landscape: the Roman cemetery at Ostia/Portus as a lived environment
Introduction The urban cemeteries of the Roman world have often been described as ‘cities of the dead.’ The organisational and structural similarities between the so-called ‘streets of tombs’ which developed on the outskirts of major Italian urban areas such as Rome, Ostia and Pompeii, and the spaces within the city walls themselves have been highlighted by scholars wishing to demonstrate that the living city ‘found its double in these ‘cities of the dead’’ (Koortbojian, 1996: 232). In addition, the composition of the dead population residing in the necropolis reflected the heterogeneous community found within the city, encompassing both rich and poor, slave and free. Hopkins (1983: 241) has proposed, for example, that the internal organisation of the columbaria of the Imperial period paralleled that of the ‘men, women and children living like the proverbial sardines as slaves and ex-slaves in great households, or as tenants in multi-storeyed apartment blocks (insulae).’ Despite this, the location of cemeteries on the periphery of urban settlements (the result of strict legal restrictions governing the location of burial sites) has led to a continued perception of these spaces as marginal landscapes which were excluded from the routine activities and concerns of daily life. In fact, these extra-urban environments were as integral to processes of social negotiation and identitycreation amongst the living as other spaces in and around the ancient city. In the past, much attention has been focused on the ways in which the grave monuments and tombs located within the cemetery allowed the dead to affirm and advertise their identity and place within society. For example, Hope (1998) has shown how marginalized and socially or legally ambiguous groups, such as gladiators, used the medium of the funerary monument both to gain, and to subsequently express in a public form, a sense of legitimacy and group identity; something that was denied them in life. However, it is important to recognise that cemeteries were not solely the preserve of the dead. Not only were the living, for the most part, responsible for the erection of commemorative structures and tombs, but they also frequented the cemetery on a regular basis for a variety of religious, practical and personal reasons, during which they interacted with the structures within it and other members of the living community. The following examination of the public dining activities that regularly took place in the necropolis of Isola Sacra near Ostia is intended to illustrate how these, and the structures associated with such occasions, played an actively significant role in the lives and identities of the community who visited the cemetery, and elucidates some of the ways in which the manipulation of the built environment of the cemetery allowed members of that community to negotiate their place within wider society.
The cemetery and the suburb Between the landscape of the city and that of the countryside lay the suburb. Although the nature of this somewhat ambiguous space was neither purely urban nor rural it was certainly not devoid of life, activity or social significance in the lives of members of the urban community. Our inability to ascribe strict geographical limits to the Roman suburb has lent to it a greater sense of obscurity. In his discussion of ancient concepts of suburbium at Rome, Champlin (1982: 97) suggests that this area was always ‘a place of ambiguity, a border region, and its physical limits (to begin with the most obvious) defy precise definition.’ He proposes that it was ‘more a matter of shared attitudes than of location’ (ibid.: 99). The difficulties facing the modern scholar seem also to have been encountered by ancient writers. Dionysius of Halicarnassus (4.13), speaking of the extent of the Augustan city, writes:
‘The districts around the city, numerous and extensive as they are, are all developed despite the fact that an enemy could easily occupy them because they are exposed and unwalled. Suppose that you want a general notion of the size of Rome. Confusion is unavoidable – you will not even have any clear indication of how far the city extends or at what point it leaves off being city. It is the fact that town is so interwoven with country which gives the observer some idea of the endless urban sprawl.’
Dionysius’ description makes it clear that the elements of both city and countryside contained within the suburb merged to form a disorderly agglomeration with indistinguishable limits. This area was also evidently one of considerable activity. On the most basic level, it was through this region that any traveller must pass in order to enter or leave the city, and in the case of a thriving metropolis such as Rome, the suburbs were no doubt particularly active environments with people constantly moving to and from the city. Included within the ‘sprawl’ through which they passed were humble and luxurious dwellings, spacious horti, and marketgardens located within easy reach of the city markets where their fresh produce was sold. Particularly noxious or hazardous industries were also located on the margins of the city due to concerns about public health and safety, with the risk of serious fire posed by tile-factories, for instance, resulting in their exclusion from urban areas (Patterson, 2000: 93). In addition, these extra-urban areas were crossed by aqueducts and busy highways.
Areas of the suburb also became increasingly associated with tombs and grave monuments, and the development of extensive cemeteries in the suburbs of urban areas can be attributed to several factors. It was stated in the legal codes of the fifth century BC, the Twelve Tables (10.1), that burial could not take place within the city walls, and with a few exceptions (the cremated remains of Trajan, for example, were deposited in the base of his column) this law continued to be generally observed. In the course of time it became common for the wealthy to erect elaborate family tombs on their private suburban estates. In addition, tombs surrounded by financially productive gardens (cepotaphia) became increasingly popular and were typically located beyond the built-up area of the city where they competed for space and attention with various other forms of sepulchral monument (Purcell, 1987). One of the many ways in which funerary monuments and tombs were significant to members of Roman society lay in the fact that they could be used to directly communicate with the living and thus perpetuate the memory of the departed, ideally for eternity. Such structures could display, or even create, a particular image of the deceased and their family by emphasising certain aspects of their identity, wealth, social status, political power, or their illustrious family history. Such display naturally functioned most effectively in a visible, highly frequented location and the busy roads which passed through the suburbs therefore offered a perfect focal point for the development of cemeteries, allowing for continued easy access to tombs or graves by members of the family, in addition to increased visibility and opportunities for interaction with strangers. These cemeteries were therefore public environments that were intended to be seen and visited by members of the local urban community and strangers alike. In contrast to modern Western graveyards, there existed no walls to delimit their extent or gates to restrict access to specific times of the day. Although funerals were frequently held at night in early Rome this practice had long been abandoned by the late Republic. Cemeteries were, however, undoubtedly visited for a variety of reasons after dark. A guest at Trimalchio’s dinner party tells how one night, as he was passing through a roadside cemetery with a companion, they decided to stop and closely examine the tombs and their epitaphs shortly after which his friend was magically transformed into a werewolf (Petronius, Satyricon 61–62). Other ancient texts make it clear that nocturnal activities in the cemeteries were often rather lawless or unearthly in nature (see, for example, Horace’s description of witches digging up bones in the Esquiline cemetery (Satire 1.8)) but even these ensured that the necropolis was rarely unoccupied by the living. These ancient cemeteries were accessible at all times allowing people to come and go as they wished and travellers using the roads along which they were situated would have passed directly through them, possibly pausing to read an epitaph, admire a monument or take advantage of the shade they provided. In fact, these areas were so public that ancient authors such as Martial (Epigrams, I. 34. 8; III. 93. 15) describe how beggars, thieves and prostitutes often took up residence in and around the tombs which undoubtedly offered ideal protection and cover for their unlawful activities. Moreover, offerings left for the dead in the form of food were certainly welcomed by the hungry. In addition, during much of its active life the cemetery was probably in a perpetual state of development as new tombs were constructed and older ones modified and maintained. Teams of construction workers, architects and craftsmen were therefore frequent visitors to the necropolis, transforming a place of ritual remembrance into a busy building site. In cemeteries associated with relatively large populations, it can be assumed that funerals took place regularly, perhaps on a daily basis, with large or small groups of people in attendance for the relevant ceremonies.
The gravesite also formed the focus for several post-funeral celebrations such as on the ninth day of mourning (novemdialis) when the family returned to the grave to consume a ritual meal, an activity that was regularly repeated in order to commemorate birthdays, anniversaries and religious festivals. In May or June roses were placed on the graves as part of the Rosalia festival, and Ovid (Fasti, II. 533–542) informs us that the annual festival of the family dead, the Parentalia, took place in February. This week long festival culminated with the day-long public ceremonies of the Feralia, during which families brought simple offerings of food and flowers to the dead and, after the appropriate prayers had been said, reprised the meal originally consumed at the funeral. During such occasions they were apparently likely to become ‘boisterously drunk’ as they dined in the company of their dead relatives (Hopkins, 1983: 233). That the dead ancestors of the family were believed to be present at such occasions (see Cumont, 1922: 54), is attested by an inscription from Rome, cited by Hopkins (ibid.: 233), which ‘expressed the hope that the couple whom it commemorated would ‘come in good health to the funeral feast and enjoy themselves along with everybody else’ (CIL 6.26554).’ Furthermore, there existed a strong tradition that continued contentment and rest in the afterlife was dependent upon sustenance provided by the living in the form of food offerings or libations (Toynbee 1971: 51). It was believed that the shades of the dead who did not receive regular nourishment would escape the confines of the grave and threaten the well-being of the living. Epigraphic evidence attests to the strength of this belief and records the setting aside of money specifically for libations and the celebration of feasts. In addition, Ovid (Fasti, II. 547– 556) recounts a cautionary tale about how the neglecting of such rites once resulted in the shades of the dead rising from their graves in order to terrorise the living.
Roman urban cemeteries were consequently highly public spaces, much like the fluid and dynamic civic areas found within the towns and cities themselves, where certain members of the community erected statues and other monuments and took part in a variety of public activities in order to establish, negotiate or maintain their social position and identity. However, unlike these urban spaces, the cemetery and the specific activities which occurred there routinely included all members of society, regardless of their wealth or rank. The physical manifestations of funerary activities may have differed in their elaboration but crucially the ordinary person was not excluded from taking an active part in them, and it is this that is fundamental to our understanding of how these dynamic public environments were manipulated by the living community.
Dining in the necropolis of Isola Sacra The permanent structures associated with funerary dining that have been identified at the Isola Sacra necropolis provide an insight into one of the more specific manipulations of these extraurban landscapes. The cemetery of Isola Sacra, built predominantly between A.D. 100–250, is located between the two major imperial ports of Ostia and Portus, approximately 23 kilometres southwest of Rome. Excavations by Guido Calza in the early twentieth century uncovered a stretch of approximately 330 metres of the busy coastal road which linked the two ports and was flanked on both sides by tombs. These tombs were predominantly constructed by freedmen (ex-slaves) and their families, and take the form of brick-built house-tombs (see Fig. 2). The barrel-vaulted structures contain several niches in which cremation urns could be placed, in addition to larger spaces (arcosolia) cut into the lower parts of the interior walls which were designed to accommodate inhumations. The external façades of the tombs held inscriptions which stated the name of the owner and determined who could, and most importantly who could not, be buried within them. These were occasionally supplemented with a small relief image depicting an occupational scene from the daily life of the owner. A plaque on the façade of tomb 29, for example, depicts a blacksmith at work surrounded by the tools of his trade. Amongst these larger structures were located smaller chest-burials (cassone) and modest tile burials (cappuccina). Excavations in the 1970s and 1980s brought to light a further 650 anonymous depositions, some of which were marked only by the protruding neck of an amphora, which also served as a libation pipe, whilst others had no visible markers (Baldassare et al., 1985). The cemetery was evidently used by a variety of members of ‘lower-class’ society but who had a range of financial resources at their disposal.
Turning to the provisions for the frequent celebration of dining rituals in the cemetery, the three most archaeologically identifiable forms that these take are wells, ovens and two-sided dining couches (biclinia). Several tombs at Isola Sacra contain ovens (including tombs 34 and 89), the location of these suggesting that they served a practical role in the preparation of food for funerary banquets, although no direct evidence has been recovered for food or vessel waste in association with these features. The relatively small size of these structures and their location at the rear of the tomb enclosure further imply that these facilities were not designed as an element of display or for a specific ritual purpose. They were probably employed predominantly for baking bread and the heating of other food, and although their physical structure was hidden from public view, the smell produced by cooking would have signalled their existence. The investment in permanent oven structures also indicates that food preparation was a regular occurrence in the cemetery. The wells (found in tombs 16, 34 and 75) may have served several purposes, their water perhaps being used for cleaning the tombs, for the flowers placed on them during certain festivals, and for funeral banquets (Calza 1940: 332). These features are always located within the built structure of the tomb, usually inside an enclosure, and are not unique to Isola Sacra. Similar examples have been identified in the Via Laurentina cemetery a few miles away at Ostia (Floriani Squarciapino 1958).
Figure 1: House-tombs in the necropolis of Isola Sacra (author)
Figure 2: Masonry biclinium attached to the façade of tomb 15, Isola Sacra (author)
The most common structure associated with regular ritual dining at Isola Sacra takes the form of the two-sided masonry dining couch, or biclinium (see Fig. 2). Ordinarily, these sloping benches on which diners would have reclined flank the entrance to the tomb, and are often found in conjunction with cubes of masonry which have been interpreted as table supports. These benches and/or table supports are also occasionally found outside the so-called chest-tombs (including tombs 99, 72a, 73, 82, 60a and 60b). Similar structures have again been identified in the Via Laurentina cemetery, although these generally comprise three benches (a triclinium) like those traditionally found in domestic contexts. The triclinium belonging to Via Laurentina tomb 31, a small columbarium, is situated inside the tomb below the rows of niches and a wall painting depicting five men taking part in a funerary banquet. Dressed in white tunics, the figures are shown holding their glasses up in a toast (Dunbabin 2003: 128–129). There are names written above the figures (for example, Felix, Foebus, and Fortunatus) although it remains unclear whether the image represents the dead inhabitants of the tomb or the celebration by the living of a ritual funerary banquet.
Closer examination reveals other differences between Isola Sacra and Via Laurentina in terms of the permanent provisions created to facilitate ritual banqueting. Most significantly it is noticeable that the triclinia of the tombs at Via Laurentina are located inside the structure of the tomb, thus rendering them invisible to the passer-by, and perhaps affording a greater sense of privacy and intimacy between the living and the deceased inhabitant(s) of the tomb during feasting activities. Dining either within the confines of the main burial chamber or the tomb enclosure was evidently not perceived as a violation of the dead or contaminating for the living. In contrast, there appears to have been an overriding desire at Isola Sacra to place these facilities in a highly conspicuous and public position outside the tomb. This is particularly interesting given the fact that these tombs were built at a time when funerary display amongst the Roman elite was becoming increasingly internalised, with greater emphasis being placed on elaborate internal decoration than the exterior appearance of the tomb. It is possible that the owners of the Isola Sacra house-tombs, rather than emulating their upper class peers, capitalised on the opportunities offered by this reduced level of competition in order to exhibit their own identity and status. Some Roman tombs (such as the anonymous tomb located just outside the Porta Marina at Ostia) provided small seats on which travellers and visitors might rest and contemplate the monument, its message and their own mortality. Although ledges which may have been used for this purpose exist on the exterior of some tombs at Isola Sacra, it is evident that the biclinia are directly related to domestic dining structures and serve no other practical purpose other than to provide a surface on which diners might recline in accordance with proper Roman dining custom. There is a lack of permanent seating (or, more accurately, reclining) areas in association with some of the more humble chest-burials. This may have been overcome through the use of portable furniture brought to the site by the family when it was required. A similar situation may have existed for the poorest burials of the cemetery, or dining rituals may have been carried out in a less formal manner in the style of a picnic, the diners reclining on cushions placed directly on the ground.
Dining and identity At the same time as these structures were capable of facilitating important dining rituals in a practical sense, the fact that they, and the activities with which they were associated, were publicly visible suggests that they also acted in a symbolic or status oriented capacity. The owners of the tombs could have used the physical accoutrements of religious feasting occasions as a means by which to signal a particular social status, or aspirations to such a status. This was doubtless of heightened significance in the minds of the population of Isola Sacra which was predominantly comprised of members of the freedman classes and more humble levels of society, including many slaves. The ambiguous and unstable social and legal status of these people, created by factors such as their non-Roman origin and the circumstances surrounding their manumission and the receiving of Roman citizenship, seems to have led to an increased desire to publicly exhibit, and therefore confirm, their legitimacy as free and respectable members of society. Significantly, dining outside the tomb made this ritual activity visible to other living members of the community. Given the social and economic composition of the population using the necropolis and the presence of large, relatively substantial tombs adjacent to those of a more humble and simple nature, families holding banquets in view of other members of that community were easily able to demonstrate that they had the financial means at their disposal to afford the expense of a substantial banquet. The extent of this ‘boasting’ presumably varied with the elaborateness of the dinner in terms of the guests present (both their number and social standing), the food, number of attendants and the tableware on display. Economic and social differences between various groups of society would have been greatly emphasised by the visible juxtaposition of these elaborate banquets with the picnics occurring in association with the modest burials surrounding them. This would have been especially evident during major festivals such as the aforementioned Parentalia when members of each social group would have held some form of dining ritual at the grave and which provided an excellent opportunity for the wealthier groups to publicly ‘show off’. It is possible that the wealthier tomb owners also contributed to the feasts held by the relatives of their dependants who were also buried in the cemetery, again allowing them to publicly exhibit their generosity and confirm their position within the community. Difficulties may have arisen, however, if the families of dependants buried within the tomb of their patron wished to celebrate dining rituals at the same time as the tomb owners and some order of precedence may have been created in order to accommodate the large numbers of people, especially during major festivals. Celebrating funerary banquets within tomb enclosures at Via Laurentina may have brought the living into more immediate contact with the dead but it prevented them from communicating aspects of their wealth, social standing and identity to the community of the living.
Dining structures located outside the tombs also allowed the freedmen of non-Roman origin within the Isola Sacra necropolis to demonstrate that they frequently indulged in a Roman ritual custom, thus conspicuously aligning themselves with Roman traditions and religious beliefs.
They could regularly be seen to be actively observing traditional religious festivals and rites thereby making an unequivocal statement about their identity as citizen members of society. The use of an explicitly formal setting for this activity, in the form of a biclinium, further underlined their familiarity with Roman high status dining practice, allowing them to further assert their identity and position in society as Romans even if they were originally of foreign or slave origin. For people striving to consolidate or legitimise the position of their family within the community, and possibly to emphasise a newfound social or legal status, this provided an excellent opportunity for a socially acceptable public display of cultural knowledge and a visible embracing of Roman custom. As Dunbabin (2003: 88) observes, ‘dining could be a spectacle, something to be displayed before the eyes of the populace,’ and most importantly, ‘reclining to dine was in itself a recognised symbol of status.’ The dining activities and the structures provided to facilitate them would have been witnessed by other members of the community. However, the audience was widened still further to encompass the travellers using the road along which the Isola Sacra cemetery was situated. This road, which linked Ostia and Portus, would have regularly carried a large volume of traffic between the two ports and those partaking in feasting within the cemetery were therefore able to communicate with a large number of strangers in addition to the inhabitants of both towns. The religious significance of the act of dining outside the tombs was not diminished by their communicative role. It remained essential to observe the appropriate rites and, whilst the living reclined on the biclinia, the tomb acted as an extra ‘couch’ which united the two other benches and located the dead in the traditional place of honour at the head of a ‘triclinium’. The feasting activities were thus directed towards the dead as much as the living, however much the latter capitalised on the opportunity for a visible expression of their identity.
Interestingly, the practicality of using these dining areas for feasting may have been severely limited by their proximity to one another. The dining couches are often placed directly adjacent to, and sometimes even superimposed upon, those of the neighbouring tomb (see Fig. 3). During major festivals when these structures were presumably in use at the same time, the area would have become particularly crowded, and the diners may have been forced to recline virtually on top of one another. This apparent inconvenience indicates that practical concerns were considered far less important than the visible presence of dining facilities outside a monumental tomb and the information concerning their identity that they were capable of communicating.
The masonry table supports found in conjunction with many of the biclinia provide further grounds for believing that these exterior dining areas served more than a practical function. These simple blocks (approximately 0.5–0.75m2) have been traditionally interpreted either as small tables or the supports for a table surface that was brought to the site when it was required (Calza 1940; Angelucci et al. 1990). This interpretation is based on similar domestic structures recovered at sites such as Pompeii, and is supported by the discovery of a similar cube of masonry at Rimini, on the surface of which were found animal bones believed to have originally comprised part of a funerary meal or ritual offering (Ortalli 2001: 231). However, at Isola Sacra these blocks are located at the opposite end to that at which they are ordinarily found in a domestic context (where they occupy the space immediately between the couches) and at a considerable distance from the benches themselves, sometimes almost two metres away. Situating them directly outside the tomb entrance would have impeded access to the interior, but their current location is apparently beyond the reach of anyone reclining on the biclinia. (See Fig. 3) As a result it is unlikely that they were used in conjunction with the dining couches as functional tables.
Figure 3: Plan of tombs 77–80 at Isola Sacra. Biclinia and masonry blocks are shaded (after Baldassarre et al. 1996)
In a similar manner to that in which the biclinia symbolised the cultural knowledge, identity and status of the tomb owner, these structures could also be viewed as a status symbol. Their mere presence outside several of the more humble chest-burials of the cemetery implies to the onlooker that theoretically a formal, and potentially elaborate, dining ritual could take place there, whether or not it actually did. If portable furniture was brought to the site and arranged around the cube of masonry, a formal meal could also be consumed in accordance with strict Roman dining customs. The construction of these simple masonry blocks would have been relatively inexpensive for those owning larger tombs and within financial reach of those owning smaller structures such as the chest-burials who could not afford to construct a complete biclinium and, in terms of display, could theoretically have served the same purpose. These cubes of masonry, which appear fairly innocuous to the modern eye, would have been instantly recognisable to contemporary members of society as a basic component of Roman formal dining furniture. Although not as overtly clear as a biclinium, their implication for the hosting of formal dining rituals would have held particular importance for those who wished to affirm or negotiate their place in respectable society but who could not afford to construct an entire biclinium. The owners of the chest-burials, already located somewhere on the social ladder between the wealthiest and the poorer groups using the cemetery, would therefore also have been able to demonstrate to the community, and strangers visiting or passing through the necropolis, their potential ability to host such activities as formal funerary banquets.
Conclusions There is strong evidence which suggests that regular funerary dining occurred at Isola Sacra, including the preparation and consuming of food in a formal manner. It is evident that such activities and the structures connected with them were also capable of conveying a particular image to the broader community. A wide variety of public activity undoubtedly took place within the cemetery involving members of different social classes and that was visible to those simply passing through. These interactions provided invaluable opportunities for display and social advancement. In some instances the mere demonstration of a theoretical ability to hold a formal Roman-style banquet, symbolised by the presence of a masonry block outside a simple tomb or chest-burial, could communicate and advertise the status and aspirations of the deceased and, importantly, their surviving family. In a world in which social status and identity were considered vital, various opportunities were seized upon to promote or display this. Funerals and funerary monuments have already been shown to be fundamental to this process, and it is therefore natural that funerary banquets, dining rituals and their associated structures were also deeply embedded within it.
This paper has highlighted just one example of the ways in which the living manipulated the physical environment of the Roman cemetery in order to negotiate their own place within the wider community whilst observing specific religious occasions and commemorating the dead. It was the fact that the cemetery was a dynamic environment that facilitated this manipulation and allowed for direct communication with other members of the living community. Although primarily a space focused on the world of the dead, within which activities took place in order to honour, appease or immortalise the departed, the environment of the cemetery was very much a public one that was understood and employed by the living for their own ends. The urban cemetery, located within the indefinable yet vibrant suburb, may initially appear to be a marginalized space inhabited by bones, dust and beggars, but it remained actively and intimately involved in the social sphere of the city itself. To describe these landscapes as ‘cities of the dead’ is to misunderstand their dynamic role in the ‘cities of the living.’
Department of Archaeology, University of Sheffield
Acknowledgements I wish to thank Maureen Carroll for her assistance and encouragement in the preparation of this paper. Thanks must also go to Cat Howarth for her support and productive comments on an earlier draft, and to Ben Croxford for organising the original session at which it was presented.