1. Introduction

The process described in this article was born of a need identified in the course of my doctoral research to express visually the spatial aspect of material culture patterning. My research explores dress and personal ornament in the Latial Iron Age cemetery of Osteria dell'Osa to the east of Rome (see Figure 1). This article presents preliminary results from the GIS component of my study. The cemetery of Osteria dell'Osa was excavated between 1971 and 1986, and published in exceptional detail in 1992 (Bietti Sestieri 1992a). The meticulous inclusion of a broad range of data, and the detailed illustrations of burials and grave goods, make this site highly suitable for testing the application of GIS to the analysis of published data, despite the lack of true geo-referencing.

Figure 1
Figure 1: Map of Italy indicating location of Osteria dell'Osa (Image courtesy Google Earth™ mapping service)

Using this important cemetery as a case study, my research project examines variation in the clothing and personal ornament of individuals, exploring the premise that such varied visual identities correspond to distinctions at the level of ideal social structure. It is proposed that gender, age, status and kinship are signalled through modes of appearance, but these are not discrete identities: the high status of a female child, for example, is unlikely to be proclaimed with the same material vocabulary used for an adult male of comparable standing. Likewise, different kinship groups might utilise unique symbols to materially confer status: a female who had come of age might be bestowed with an item of jewellery that was stylistically unique to her family. Such items proclaim the dual identities of kinship and adulthood.

Distinctions between gender, age and materially manifested status can often be discerned through conventional statistical analyses: exclusive correlation of an artefact type with a sex category, for example, reveals not only social acknowledgement of sex, but exemplifies the social norms for its expression. Vida Navarro's (1992) study of sex and fibula type in southern Italy typifies this approach.

Kinship, however, is not so easily identified, especially when kinship relations are formed through marriage or other social alliances. Indeed, even hereditary relationships are extremely difficult to prove. Brothwell (1981, 109) argues that biological data may not, in fact, be useful in determining family relationships, as only rare physical anomalies such as brachydactyly (having short fingers) will be apparent skeletally, and too little is known about family transmission of normal skeletal features. There have been significant advances in DNA analysis since the publication of Brothwell's argument, but the cost of such analyses is still too prohibitive for many projects and many sites lack biological material of suitable quality.

Some scholars refute the value of archaeology to reconstructing kinship ties. Allen and Richardson (1971) argue that without understanding the rules of descent or residence we cannot hope to identify kinship groups, or confidently assign members of a common burial group to the same lineage. The location of adult residence is determined by a range of factors that transcends biological relationship, and may be influenced as much by individual choice as by prescribed rules (Allen and Richardson 1971, 47). We may not be able to identify the membership rules for an archaeologically identified corporate group. Allen and Richardson (1971, 49) assert that 'it is unwise to assume that the rule of descent produces non-overlapping, unambiguously affiliated collections of individuals'. We must neither discount kinship on the basis of material dissimilarity, nor automatically assume it on the basis of common patterning and location. It is, however, possible to identify corporate aggregation in the mortuary record, even if the precise genealogical relationships are unknowable. Consideration of spatial patterning is essential to this process.

The assumption that kinship relations will be visible in cemetery layout is predicated on the ethnographic observations of Saxe (1970, 170), who argues that where control of limited resources is mandated by lineage, a discrete burial ground will be maintained to emphasise ancestral links. This conclusion is supported by the findings of Carr (1995, 165), whose study of non-state societies drawn from the Human Relations Area Files reveals that in 27 per cent of observed groups burial location was determined by kin affiliation. This finding is also supported by the earlier research of Goldstein (1976, 60), who echoes Saxe in positing a link between bounded burial areas and membership of a group that controls limited resources.

For the purposes of this analysis Bietti Sestieri's (1992a) division of the cemetery into discrete clusters based on kin affiliation is accepted (see Table 1). There is no obvious spatial separation of burial groups, but Bietti Sestieri (1992b, 146) identified distinct groupings by considering relative positions, anthropological data, grave structure and corredi. The names of the clusters are derived from the lowest and highest burial numbers within the group. For example, cluster 340-579 contains 33 burials, including burials 340 and 579. The 31 burials in this cluster and between these numbers are not sequentially numbered because their excavation and numerical designation preceded division of the cemetery into clusters.

Table 1: Phase IIA burials by sex, age and cluster

Phase IIASexAge Class
ClusterMature Male Mature Female Unknown Infant Child Juvenile Adult Old Adult Unknown Total Burials
1-60         0
165-19363 7 4 1 2 3 6 16
34-219          0
340-5791 1    11 2
402-450 2 3 1  1 4 1 6
458-51713    4  4
503-578         0
64-2063 5 2  2 6 2 10
71-435 3 3 3 2 1 3 3  9
88-497 813532 183 26
Central damaged         0
North814 10 1 5 3 13 9 1 32
North/Central damaged         0
South 12 8 1  1 15 4 121
South damaged6 3 1  1 81 10

Table 2: Phase IIB burials by sex, age and cluster

Phase IIBSexAge Class
ClusterMature Male Mature Female Unknown Infant Child Juvenile Adult Old Adult Unknown Total Burials
1-60 10 26 10 3 5 2 27 4 5 46
165-193  1      1  1
34-219 2 7 2  1  8 1 1 11
340-579 8 15 6 5 1  18 5  29
402-450 8 10 2  1 1 14 4  20
458-517 7 12 3 2   15 5  22
503-578 16 16 10 8 1 1 25 6 1 42
64-206 3 6 2  2  6 3  11
71-435 10 10 1   1 9 11  21
88-497 4 5 1  1  8 1  10
92-410 2 4     5  1 6
Central damaged 1 2     2  1 3
North 4      4   4
North/Central damaged 3 10 5 4 1  12 1  18
South 1  2 2   1   3
South damaged 3 2     5   5

Table 3: Phase II (unspecified) burials by sex, age and cluster

Phase II unspecifiedSexAge Class
ClusterMature Male Mature Female Unknown Infant Child Juvenile Adult Old Adult Unknown Total Burials
1-60          0
165-193 1 1     1 1  2
34-219          0
340-579 1  1 1   1   2
402-450 3 2 1 1   4 1  6
458-517  7 2 1   5 2 1 9
503-578 1      1   1
64-206  1     1   1
71-435  2 3 2    1 2 5
88-497  2     2   2
92-410  3     2 1  3
Central damaged  2     1  1 2
North          0
North/Central damaged  1 1    1  1 2
South  2 2 1 1   2  4
South damaged 2 3 6   1 4  6 11

There is not necessarily a common biological link between every individual within a cluster. However, following Bietti Sestieri (1992b, 146), we can accept the corporate membership of each individual within the group, but not necessarily be able to describe his or her relationship with the others.

The division of the Osteria dell'Osa necropolis into discrete clusters is not unique, and has been applied at other Latial sites such as Caracupa (Angle and Gianni 1985).

In a mortuary sample comprised of many family groups, patterning within groups will not be statistically obvious if the spatial aspect is neglected. There is a risk that such small-scale patterning will escape notice if it connects a small group of individuals of disparate age, gender and status. If spatial proximity is noted, however, relationships are immediately obvious. In the absence of unambiguous biological or genetic data, spatial association of burials with material or ritual connections is often considered an indication of kinship relations (see Allen and Richardson 1971). The ability to make these relationships visually explicit through GIS programmes is an invaluable adjunct to conventional database analyses.

The introduction of GIS analysis into my research on dress in the central Italian Iron Age has provided a new analytical perspective, and permitted more comprehensive exploration of the reasons for material variation in dress in the cemetery of Osteria dell'Osa. This article illustrates the potential of GIS to illuminate spatial relationships by mapping a range of data on a digitised site plan.

GIS applications have traditionally attracted users with specialist geographical knowledge that equips them to take advantage of full programme functionality. It is widely recognised that GIS programme interfaces are not intuitive to non-specialist users (Traynor and Williams 1995, 288). However, non-specialists from a range of fields are increasingly making use of these applications (Traynor and Williams 1995). Such users can take advantage of particular features without the need for extensive training, fully geo-referenced data, or even an understanding of the programmes' functionality beyond their own requirements. The greater accessibility of GIS was anticipated by Burrough and McDonnell (1998, 296) who describe a move towards the development of more user-friendly interfaces that enable non-specialists to engage directly with programmes using their own data. The development of more user-friendly interfaces has allowed a degree of innovation and flexibility that allows much broader application of GIS and the development of procedural solutions that circumvent the higher functions of the software. This project presents one way in which GIS functionality has been appropriated and modified to achieve good results for a non-specialist user using non-geo-referenced data.


© Internet Archaeology/Author(s) URL:
Last updated: Mon Jun 30 2008