Reconstructing Contact Space Biographies in Sudan During the Bronze Age

Abstract

Traditional models of interaction in northern Sudan have innate Egyptological, elite, and urban biases which have relegated certain areas to mere peripheries of more ‘established’ and ‘central’ sites. In order to reach a higher resolution understanding of cultural dynamics and diversity of ancient Nilotic groups, the DiverseNile project has established the bespoke concept of Contact Space Biography which we present in the following article. We challenge existing approaches to cultural contact in the region by adopting a bottom-up approach which moves away from well-established categorisation of sites in our study area. In particular by reconstructing landscape biographies of the Bronze Age in the Middle Nile beyond established cultural categories in order to provide new insights into the ancient dynamics of social spaces, which include landscape features and non-human activities. In the following we instead consider such areas as complex social spaces intertwined with, an often changing, landscape by presenting our findings from the study of cemetery and settlement sites. Overall, the concept of Contact Space Biography effectively combines models of contact spaces, the idiosyncrasies of a changing landscape and the technological and industrial prerogatives of those living in and accessing this region.

1. Introduction

This paper introduces the concept of Contact Space Biography established during the ongoing ERC funded DiverseNile project through our work in Upper Nubia, now Northern Sudan. The main aim of DiverseNile is to reconstruct landscape biographies of the Middle Nile during the Bronze Age beyond established cultural categories in order to provide new insights into the ancient dynamics of social spaces. Historically, the region has often been viewed from an Egyptological perspective as a zone of Egyptian influence or colonialism. However, far from being a one-way exchange dominated by Egypt, various interactions and mutual influences are evident in the Nile Valley (S. T. Smith, 2003; van Pelt, 2013; Morris, 2018; S. T. Smith, 2021). The long and diverse history of relationships and cultural encounters between Egypt and Nubia are increasingly being addressed from a post-colonial perspective and with bottom-up approaches (van Pelt, 2013; Matić, 2017; Spencer et al., 2017; Howley, 2018; Lemos, 2020; Lemos & Budka, 2021; Budka & Lemos, 2024). It is now well established that the former colonial perspectives neglected the complexities of the two-way encounters traceable in the region and the cultural mixing, intermingling, entanglement, and blurred cultural boundaries this inevitably leads to (S. T. Smith, 2003; van Pelt, 2013; Edwards, 2004; Lemos & Tipper, 2021).

1.1. The DiverseNile Projet and the MUAFS Concession Area

The Munich University Attab to Ferka Survey Project (MUAFS) concession area covers a region of over 337 km² in Upper Nubia (Figure 1) and includes the districts of Attab, Ginis, Kosha, Mograkka, and Ferka.

This region has a long-standing history as a contact zone between different cultural groups. During the Bronze Age this included the Kerma Kingdom (c.2500–1500BC)—established in Upper Nubia—and New Kingdom Egyptians (c.1550–1069BC), in particular during the overlapping 18th Dynasty period (c.1550–1292BC) and the Ramesside period (c.1292–1096BC). Our findings also suggest the presence of other cultural groups from Egypt and Nubia such as ‘Pan-Grave’ (c.1850–1600BC) communities. Geographically, the region is also in the ‘periphery’ of two major urban sites: Sai which was occupied in the Egyptian 18th Dynasty and the subsequent Ramesside period, and Amara West in use during the Ramesside period. The location of the MUAFS concession area, next to the Dal cataract region and upstream of the rocky outcrop of the Batn el-Haggar, makes it a geological boundary zone as well as a cultural contact zone.

The DiverseNile project addresses cultural diversity across a range of themes (e.g., social complexity and settlement patterns; intersocial aspects [such as gender, age, ethnicity, origin]; subsistence strategies; cooking practices/nutrition; health and fitness; technologies and technology transfer; trade and contacts as well as religious practices) in the MUAFS concession area for the Bronze Age [for more information on the project and the research area see (Budka, 2019; Budka, 2020a, 2024a). This approach allows us to consider concepts that go beyond identity by also incorporating external factors such as the local landscape, in order to assess specific living conditions in the Attab–Ferka region during the Bronze Age in direct comparison to urban centres such as Amara West and Sai Island.

1.2. Contact Space Biographies

In order to reach a higher resolution understanding of the cultural dynamics and diversity of ancient Nilotic groups, the DiverseNile project draws upon the concept of contact spaces after Stockhammer and Athanassov’s meaning of ‘social spaces where human actors meet, perceive and constitute otherness, clash, and grapple with each other’ (Stockhammer & Athanassov, 2018, p. 106). Combining this with a landscape biography approach, we consider the individual life cycles of all cohabiting actors in the region, in particular humans, fauna and flora, as well as human-made technologies (Kolen & Renes, 2015, p. 32). Peripheral areas, like the Attab–Ferka region, are perceived as complex social spaces intertwined with the landscape. It should be stressed that ‘periphery’ is not used in the sense of traditional theoretical models of core-periphery and world system theory (Stein, 2002) (for theoretical flaws of this theory after Wallerstein (Wallerstein, 1974) see also (Sulas & Pikirayi, 2020)), implying one superior power with asymmetrical control and high development, but ‘peripheral’ in terms of location in the hinterland of urban sites (see also (Walsh, 2022)).

The combination of the contact space and landscape biography models results in a novel approach that we call Contact Space Biography. Biography is used, in this context, as a metaphor revealing ‘the variety of relationships between people and things in different cultural contexts’ (Gosden & Marshall, 1999, p. 177) and integrating the effects of various environmental conditions on these relationships. Like the archaeological concept of the biography of place (Raja & Sindbæk, 2020) we combine multiple techniques to assess individual sites and societies, for instance, improved geoarchaeological methods enable a more rigorous reconstruction of biographies. Overall, this leads to a more comprehensive understanding of the cultural landscape by also taking into account overlapping concepts such as biography of landscape (Samuels, 1979), temporality of the landscape (Ingold, 1993) and landscapes as life worlds (Kolen & Renes, 2015). Another approach with relevance for some aspects in our study of New Kingdom Nubia is the notion of the political landscape (A. T. Smith, 2003) which is also addressed in the recent overview of archaeological landscape studies in relation to urban experience (M. L. Smith, 2014). For Sudan, earlier strands of landscape archaeology in the region of the Middle Nile ((Grzymski, 2004) and the recent summary (Edwards, 2021)) and Eastern Sudan (Fattovich et al., 1988–1989; Manzo, 2017) were also taken into account (see also (Fattovich, 2003) on a later case study in Ethiopia).

Within the spatial framework of a contact space, we consider the significance of materiality (Miller, 2005; Maran & Stockhammer, 2012) and understand objects as active partakers of culture and intercultural contact (but see also Howley, 2018, p. 23; Appadurai, 1986; Steel, 2018). A related approach we also take into account is the cultural biography of objects (Gosden & Marshall, 1999), see also (Hoskins, 1998; Koptyoff, 1986). In particular, the recently developed concept of objectscapes offers several advantages (Pitts & Versluys, 2021). As Lemos has convincingly demonstrated, phenomena traceable within the material culture of cemeteries in New Kingdom Nubia offer insights into various social realities reflecting lived experiences and local choices (Lemos, 2020; Lemos, 2024). Objects in motion actively shaped a new set of material culture with a distinctive local character in New Kingdom Nubia and are, therefore, a means by which to characterise specific contact spaces (Lemos & Budka, 2021).

2. Materials and Methods

A starting point for the project was a survey conducted in the region during the 1970s (discussed in more detail below). Many of the sites identified were relocated and recorded by foot and car surveys over several seasons of the MUAFS project (2018–2023) when a number of additional sites were also identified and recorded. This formed the basis of our GIS database of Bronze Age archaeological sites in the MUAFS concession area which we use to consider the distribution of sites, particularly in relation to the local landscape and significant geological changes (for example shifts in the flow of the Nile or the drying up of river channels, forming so called palaeochannels). Combined with remote sensing approaches, this initial work was crucial to developing and designing fieldwork plans in Sudan, in particular for the establishment of excavation strategies.

Survey work and drone flights were carried out in the 2021, 2022 and 2023 DiverseNile field seasons. Excavation work focused on (1) domestic sites in the Attab West district, which were confirmed by survey work as being associated with both Classic Kerma pottery and 18th Dynasty Egyptian-style pottery and which are therefore material evidence for Bronze age contact spaces; (2) cemetery sites in Ginis East attributed in the 1970s survey to the Bronze Age for which magnetometry surveys were conducted in 2019. Excavations by the DiverseNile project took place in 2022 and 2023 as single-context excavations and follow the recording system introduced by the ERC funded AcrossBorders project for Sudan (Budka, 2020b, pp. 69–70), which includes calculating georeferenced othroprojections and Digital Surface Models (DSM) for every stratigraphic unit where possible. Our fieldwork in Sudan is supported by the National Corporation of Antiquities and Museums (NCAM), in particular our inspector Huda Magzoub, as well as the expertise of a team of local workmen. The initial ceramic analysis is conducted on the spot by Julia Budka while the necessary specialists for other materials, scientific analyses and dating are typically conducted by a member of the DiverseNile team in Munich or outsourced. Unfortunately, detailed analyses of some of the material, including osteological work is still pending due to the ongoing war in Sudan.

While the focus of the DiverseNile project is on Bronze Age sites, earlier and later evidence from the MUAFS concession area is also discussed as this provides indicators for the continuity, or lack of continuity, of significant aspects in the region, such as landscape use, technological changes, infrastructure, material culture, and potential cultural shifts. Furthermore, some of the sites recorded in the 1970s were misattributed to the Egyptian New Kingdom and were found to be later in date when investigated by DiverseNile. In the following, we first discuss the use of historical data and remote sensing by the project, before turning more specifically to the study of cemetery and settlement sites.

2.1. Historical Data

A large part of the research area being studied by the DiverseNile project was surveyed in the 1970s (Figure 2) under the direction of André Vila on behalf of the French section of antiquities in Sudan (Section Française de la Direction des Antiquités du Soudan [SFDAS]) and the National Corporation for Antiquities and Museums in Sudan (NCAM) (Vila, 1975). The survey was published shortly after with all volumes now available online through the SFDAS website. This is an invaluable resource for the project as the survey provides a summary of numerous archaeological sites in the region and was minimally invasive where possible (Vila, 1975). All sites were recorded on individual record cards and also pin-pointed on aerial photographs provided by the Sudan Survey Department.

The survey was invaluable in identifying potential sites of interest for further investigation, in particular excavation. However, rather than simply use this existing data to relocate relevant archaeological sites, a key objective of the DiverseNile project is to integrate the results of the 1970s survey into our ongoing research (for more details see (Ward, 2024)). This integration is based on existing discussions on the re-use of archaeological data (e.g., Casarotto, 2022; Dallas, 2015; Faniel et al., 2013; Huggett, 2018; Nicholson et al., 2023; Wylie, 2017) and the idiosyncrasies of the DiverseNile project and research area. As such we not only re-use data from the 1970s survey but also re-assess, integrate, and alter them where necessary, alongside the creation of new and complementary data from ongoing research and fieldwork. This has included re-dating some of the sites based on both survey and excavation work, in particular a number of sites identified as New Kingdom by Vila and his team are actually later in date. Therefore, a consequence of using data collected in the 1970s as a ‘starting point’ means that some of the sites investigated do not strictly fall within the chronological remit of the project. This is inevitable in all use of historical and legacy data.

This work also enables us to reflect on our own data collection and sharing practices based on existing principles of data re-use and sharing (e.g., FAIR principles (Wilkinson et al., 2016; De Haas & van Leusen, 2020; Nicholson et al., 2023) and the importance of meta- and para-data (Hart, 2016; White et al., 2023). For the reidentification and survey of sites we retain the original 1970s designation (numbered based on a grid-square location, e.g., 2-S-54). However, to avoid confusion between work conducted by the DiverseNile project and the 1970s survey we provide new site codes for any sites for which we do anything more than a straightforward survey—most obviously excavation but also non-invasive fieldwork such as magnetometry surveys (e.g., we record our excavation of 2-S-54 as AtW 002)—and use an independent numbering system for any ‘new’ sites uncovered (e.g., MUAFS 059).

The information provided by the published results of the survey is invaluable, but unfortunately due to the ongoing civil war in Sudan we do not have access to any of the archival sources from the 1970s. This includes more extensive documentation and material collected during the survey which as far as we know was stored in the Sudan Nation-al Museum in Khartoum. Unfortunately, recent reports of damage and looting at the Museum (Bashir, 2024; Adams, 2024; Salih, 2024; Salih & Burke, 2023) may mean some of this material will never be recovered. Although the extent of the loss is still unknown, and somewhat trivial compared to the ongoing suffering and loss of life in Sudan (United Nations, 2024). Nevertheless, the published results allow us to assess key factors such as the type, scale, location, and date of the sites. This information which was extremely beneficial when we were able to conduct fieldwork in Sudan has become indispensable due to the current impossibility of accessing sites in the country, particularly when combined with remote sensing techniques.

2.2. Remote Sensing

The foot and car survey of the MUAFS concession area was supplemented by drone recordings with a DJI Phantom 4 aircraft and DJI Go and Drone Deploy applications on an iPad (8th generation). The initial orthophotos were calculated from the drone imagery using Agisoft Metashape Pro 2.0.0 and since 2023 Drone2Map (Esri) 2023.1. In the field, geodata was collected using a Trimble Catalyst GPS receiver and a TDC 6000 data collector. We aimed for a horizontal and vertical accuracy of less than 5 cm for all points collected with the Trimble Catalyst, although accuracy could vary due to factors such as wind speed and number of satellites. Ground control points, recorded with the Trimble Catalyst receiver, were used to georeference and to improve the accuracy of all drone models. Most of our drone survey was conducted in Attab West and Ginis West (for the location of these districts see Figure 5 along a palaeochannel (see Woodward et al., 2017). This included detailed, low-flying (30 m) mapping of two sites we excavated in 2023, Attab West 001 (AtW 001) and Attab West 002 (AtW 002) (Budka et al., 2023). During the foot survey on the west bank in Attab West and Ginis West, we mapped approximately 300 dry-stone features as significant features within the cultural landscape. Depending on their size, five to 30 GPS points were recorded for each feature using the Trimble Catalyst Receiver. Special attention was taken to capture the highest and lowest point of each feature, as well as the overall extent. Another focus was on documenting the palaeochannels in the research area which extend west from the Nile into the desert in order to understand terrain patterns and changes over time. Here, historic aerial photographs are also very helpful (see Vila, 1977a, Figure 3; Spencer et al., 2012).

As fieldwork was delayed in 2020 due to the COVID-19 pandemic and has been on hiatus since 2023 due to the war in Sudan, we have shifted our focus to more comprehensive remote sensing of the area. Thanks to a cooperation with the German Aerospace Center (DLR), we have carried out initial geological modelling of the research area based on satellite data. We are also making use of SAR (Synthetic Aperture Radar) data, which are already being used extensively for archaeological prospection (e.g., Linck et al., 2013), including in Sudan (Karberg, 2021). The high-resolution SAR images from the German TerraSAR-X satellite have a penetration depth of around 30 cm. As some of the archaeological remains in the MUAFS research concession area are only slightly covered by sand and visible on the surface, we expect excellent results for remote mapping of sites. This is especially true in combination with the available DEM (Digital Elevation Model) data from the TanDEM-X satellite (for the use of TanDEM-X data in Sudan archaeology (see Eger & Karberg, 2020)) and our own drone imagery for large areas of the concession.

2.3. Cemetery Sites

The types of cemetery sites in the research area vary in size and date, from isolated graves to more extensive cemeteries many of which were described by Vila and his team (Miller, 2005; Vila, 1976a, 1976b, 1977a, 1977b, 1979). This includes typical large Kerma tumulus cemeteries in Kosha East and Ferka East which were identified by Laurence Kirwan in the 1930s (Kirwan, 1939, pp. 19, 27) but have unfortunately now been looted or destroyed. Yet more cemeteries were recorded by Vila and his team on both the east and west banks of the MUAFS concession area.

In addition to relocating, mapping, and surveying the cemeteries, the DiverseNile project focused on the excavation of Kerma cemetery GiE 003 (numbered 2-T-39 by Vila and the 1970s team) in 2022 and 2023. Five trenches yielded a total of 58 graves dating from the Middle (c.2050–1750 BC) to Classic Kerma (c. 1750–1500 BC) period, all unfortunately looted (Figure 3). The post-excavation analysis of GiE 003 employed a comprehensive data analysis method, translating extensive burial data into detailed visual representations developed for the project to investigate potential social and chronological stratification within the cemetery. Each grave was assigned a type based on the typology developed by Vila for his excavations at the Kerma cemetery at Ukma West (Vila, 1987, Figure 20) in the Batn el-Haggar region (Figure 1). Together with the nearby site of Akasha (Maystre, 1980), Ukma is the closest parallel for GiE 003 as a provincial Kerma cemetery (see Budka, 2022). Vila’s typology is based on 228 graves, which he divided into three types according to the basic shape of the grave-pit (round, oval and rectangular), all of which are also present in GiE 003.

To systematically analyse patterns in GiE 003, data for each grave were meticulously catalogued in Excel tables, capturing as much detail as possible regarding burial characteristics, size, depth, and associated artefacts. These data were then visualised using Adobe Illustrator (see Figure 6 in the Results section), with each feature represented by a longitudinal section symbol. Colour fills were applied to these symbols, differentiating between grave types and visually highlighting trends across the cemetery. This visual approach allowed us to identify distinct burial types, each of which reflected variations in burial customs over time and potentially between social strata of the community.

2.4. Settlement Sites

The Bronze Age settlements in the MUAFS concession area include sites of a domestic nature—including campsites and sites with standing architecture—as well as sites more likely linked to agriculture and flood defenses. Many of these sites were identified and described during the 1970s survey (see Vila, 1976a, 1976b, 1977a, 1977b, 1979). As part of the DiverseNile project we have re-located, surveyed, and excavated a number of these settlement sites in order to understand them on a micro- and macro-level. We have also identified a number of sites not previously surveyed in the 1970s (e.g., MUAFS 059, Figure 4). Many sites in the research area are either small, or based on limited or scattered remains (e.g., small concentrations of pot sherds and stone tools) without any standing architecture which can make them difficult to identify on the ground (Figure 4).

Through the assessment of the material culture, architecture, and location of these settlement sites, a major objective of the project is also to assess their intercultural nature (Budka, 2020a). Many show evidence of colonial Egyptian influence, local Nubian influence and connections to cultures further afield. Whether Egyptian or Nubian, major local sites such as Sai Island would have been dependent on these smaller outposts to access resources such as gold. Many of the sites in the MUAFS research area would likely have been linked to expeditions into the desert, possibly as supply points for gold exploitation.

Through a bottom-up methodology based on the Contact Space Biography we draw on broader models of trade and mercantilism on the local-scale (e.g., Rathje & Sabloff, 1976; Moreno García, 2021) as well as comparative approaches with other regions centred on the extraction of local resources such as gold (e.g., Mills, 2011). We also consider similar sites in the surrounding region to place our research in the wider context of archaeology in Northern Sudan, in particular around Amara West (e.g., site 2-R-18 (Stevens & Garnett, 2017)) and beyond.

Regions such as the MUAFS research area are key contact areas in understanding the interactions between Egypt and other states to the south such as Kerma. As such, these small settlements and outposts are likely to have been used by a range of people, although may have been managed by local populations. Bottom-up approaches which look at small non-elite sites are more likely to reveal the realities of cultural interactions and influence on the region through the adoption, assimilation, and rejection of particular cultural traits and material culture (Stockhammer & Athanassov, 2018), especially compared to elite settlement sites and towns which are more heavily influenced, at least outwardly or in appearance, by the supposed dominant power in the region. This is of course nothing new in archaeology, but the study of Nubia and northern Sudan has long been constrained by elite and imperial models of interaction based on Egyptian influence and at times occupation of the region (van Pelt, 2013; Lemos & Budka, 2021; Budka, 2019; Carrano et al., 2009). These approaches are outdated, and likely influenced by the British colonial period in the region during which such models were conceived, but continue to have an implicit impact on much ongoing research [see also the concept of sublimation (Kratimenos, in press)].

3. Results

The following summarises the results of some of the work achieved by the DiverseNile project so far for the cemetery and settlement sites, as well as the study of the landscape and riverscape of the MUAFS concession area. The locations of specific sites discussed are shown on the following figure (Figure 5).

3.1. Cemetery Sites

Overall, we have identified 19 Kerma, eight Egyptian New Kingdom (two of which can be more specifically dated to the Ramesside period) and two probable Bronze Age cemetery sites in the concession area, 12 of which are newly recorded sites not documented by the 1970s survey. The following presents the results from the excavation of cemeteries GiE 003 and GiE 002.

3.1.1. GiE 003

58 graves and 3 offering pits were excavated in GiE 003 across five trenches excavated in 2022 and 2023 (Figure 6). All of the graves appear to have been looted, most likely in antiquity. As such there are only a few in-situ burials, with a number of individuals identified in contracted positions, and some of the grave goods remain. The main fill is a silty-sand matrix, a mixture of the original fill and subsequent wind-blown sand in the area, especially in the looter’s pits. The superstructures are destroyed but the cuts of the graves are clearly visible below the topsoil of the site.

The cemetery includes considerable variations in burial style, a feature it shares with other Kerma cemeteries such as those at Akasha (Maystre, 1980), Ukma (Vila, 1987), and Sai (Budka, 2022; Budka et al., 2023). The dimensions of the grave-pits in GiE 003 range from 2.2 m × 1.3 m to 1.2 m × 0.8 m with depths of between 0.3 m and 1 m, comparable to similar burials at Ferka and Akasha (Maystre, 1980; Budka, 2022). Based on the grave goods, the circular graves in the southernmost trench, Trench 2, can be dated to the Middle Kerma period (c.2050–1750BC) and as such represent the earliest phase of the cemetery excavated so far. All of them can be classified as the reversed conical pit type, I-A3, as defined by Vila for Ukma West (Vila, 1987, p. 20, Figure 20.3).

Most of the graves in GiE 003 are typical of the Classic Kerma, and can be classed as Vila’s Type III (rectangular graves) used for Ukma West. The most common variants at GiE 003 are III-A1—rectangular with sharp corners and walls sloping inwards—and III-B—sub-rectangular cut, which were both assigned to 16 graves. 11 of the III-A1 and 15 of the III-B graves at GiE 003 excavated so far have small shallow trenches cut into the head and foot of the base which in some cases hold the feet of funerary beds (based on in situ evidence for Features 10, 20 and 61), and in other cases were used for grave goods. Such trenches and the use of funerary beds are well-known from other Classic Kerma cemeteries, both to the north (e.g., Allaqi and Shellal, see (Gratien, 1978, pp. 110–112, 115–166, Figures 30 and 32) and to the south of the research area (e.g., at Soleb, Gratien, 1978, p. 42, Figure 8b). The use of trenches is also attested to at another site in the MUAFS concession area in Ferka East (e.g., 3-G-19 (Vila, 1976a, p. 39, Figure 15A,C)) and nearby on Sai Island (e.g., Tomb 27 (Gratien, 1985, Figure 216)). Evidence from Ukma shows that sometimes four circular holes, one for each foot of the bed, were sometimes used instead of trenches (see Vila, 1987).

Feature 7 only has a trench dug out in the western end rather than two. Such a design is also known from Ukma (e.g., Tombs 9 and 27 (Vila, 1987, pp. 46, 60, Figures 52, 68)). There, the remains of a contracted burial on a bed frame testifies that graves with just one rather than two trenches were also used with funerary beds. At Ukma, the depression was used for ceramic grave goods and an animal offering was placed on the bottom surface of the pit at the end of the bed frame. In 3-G-19 at Ferka East, one trench in the east is combined with two circular depressions for the feet of a bed in a rectangular Classic Kerma grave (Vila, 1976a, p. 39, Figure 15c). This testifies to both the variability of the design in the base of the pits as well as the seemingly multi-functional use of these trenches.

Although most of the Classic Kerma graves are east-west oriented and rectangular or sub-rectangular in shape, Trench 1 also included Vila’s Type II (oval) graves datable to Classic Kerma times based on the grave goods. Feature 3, can be classified as Vila’s II-A2 variant (oval with walls sloping outwards) but has an unusual small niche at the bottom of the pit (Vila, 1976a) which differs from all the other graves excavated so far (Figure 3). Two circular burial pits with the remains of an infant and a sub-adult individual respectively were also excavated in this trench (Features 14 and 17). GiE 003 therefore shows the contemporaneous use of various shapes of pits (circular, oval, and rectangular) attested to in other Nubian cemeteries (see Vila, 1987; Cohen, 1992).

Kerma funerary practices and identities are traceable in GiE 003 through the ceramics as well as the primary grave goods. Strong Kerma cultural expressions are noticeable for the Classic Kerma burials through the position of the interments on wooden beds, animal offerings and Black Topped fine ware. However, some of the ceramics are not typical of the Kerma culture but are more closely associated with the Pan-Grave horizon, and, as such, may reflect hybrid elements to the funerary practices. The discovery of Feature 50 (Figure 7), a large Pan-Grave style burial in Trench 5 in 2023 provided clear evidence for the combined occurrence of Kerma style and Pan-Grave style material culture in one burial (Budka, 2024a, pp. 22–24), reflecting findings from the Fourth Cataract region (Emberling & Williams, 2010, pp. 27, 29, 33). Hybrid elements seem to be especially characteristic of late Pan-Grave burials (Bourriau, 1981; Gatto, 2014, p. 25; de Souza, 2019) which could suggest that Feature 50 is closer in date to the Classic, rather than Middle, Kerma period (Figure 7).

Unlike other Kerma cemeteries, GiE 003 does not contain evidence of bucrania (ox skulls or horns), which are often associated with higher-status graves at elite sites such as Kerma and Sai Island (for a discussion see (Budka, 2022, p. 57), for Kerma see also (Chaix et al., 2012; Dubosson, 2018; Dubosson & Honegger, 2023)). This hints, together with the material culture found, at complex cultural expressions within the Kerma state, which likely reflect both local preferences and wider interactions (see Walsh, 2022). Overall, evidence from the GiE 003 cemetery most likely reflects broad trends in cultural diversity in the surrounding area and illustrates the contact space nature of the Attab—Ferka region.

3.1.2. GiE 002

Another cemetery in Ginis, GiE 002, (Vila’s site 2-T-13 (Vila, 1977a, p. 48)) was excavated in 2022, after a magnetometry survey in 2019 (Figure 8). Vila and his team dated this small cemetery with largely dismantled stone superstructures to the New Kingdom, but the single pit burial excavated in the 1970s, in particular the ceramic evidence suggests that this date is too early (Vila, 1977a, pp. 47–48, Figure 16). The cemetery more likely dates to between 900 and 300BC during the Pre-Napatan or Napatan period.

We opened a total of eight trenches and one small extension and only discovered two graves of the pit burial type described by Vila. All of the associated ceramics postdate the New Kingdom. In Trench 1, one heavily looted grave was identified in the eastern part of the trench which contained the disturbed and displaced remains of at least two adult individuals (Feature 1). Since Vila and his team found the remains of seven individuals in a similar burial pit (Vila, 1977a, p. 48, Figure 16), it is conceivable that there were once more individuals in the heavily looted pit we excavated. Feature 2, in Trench 4, proved to be more interesting and slightly different to the graves described by Vila. This includes the presence of mud-bricks and a much deeper grave-cut with a niche cut into the side of the pit. The grave was used in at least two different periods: the Pre-Napatan/Napatan era and the Medieval period, with possible use in between. The oldest burial was discovered in an extended position in the southern niche of the grave. The burial was partly disturbed by probable looting in antiquity before the Medieval reuse but is otherwise complete. Based on the remains of mud-bricks which would once have walled-off the niche and the grave goods—including pottery and amulets—this grave has close parallels to the nearby cemetery of Missimina, thought to be the single funerary site for this period in the Dal-Abri region (Vila, 1980). GiE 002 proved to be more complex than presumed by Vila and his team. The cemetery does not only contain simple pit burials but also graves with side chambers sealed with mud-bricks. The date proposed by Vila needs to be corrected: this small cemetery was mostly used in the 10th–8th centuries BC, with later use in the Medieval period.

Burials dating more definitively to the Egyptian New Kingdom are attested in the concession areas as dome/cleft tombs—all looted and therefore partly dated with uncertainty, but there are similar 18th Dynasty tombs around Faras West, (Nordström, 2014, p. 157))—and tumuli/unclear stone structures with underground chambers. The most spectacular pharaonic tomb is 3-P-50, excavated by Vila in Ginis West and dated to the Ramesside period (Vila, 1977a, pp. 145–159) (see also Lemos & Budka, 2021, p. 413; Lemos, 2022, p. 26, Figure 3). This unusual tomb type finds a parallel in Amara West with the newly excavated tomb G244, a funerary monument with a Nubian-style burial mound superstructure and an Egyptian-style substructure (Binder, 2017, pp. 599–606). At both Ginis West and Amara West, the tombs clearly reflect local cultural entanglement with the presence of both Nubian and Egyptian cultural traits in the architecture and material culture. It is worth stressing that no elite mud-brick pyramid tombs with rock-cut chambers, commonly found at urban sites of the New Kingdom in Nubia, are known from the Attab–Ferka region.

3.2. Settlement Sites

Overall, we have been able to either identify 39 Kerma and 12 New Kingdom settlement sites (of which five can be dated more specifically to the 18th Egyptian Dynasty and one to the following period) in the MUAFS concession area, 25 of which are newly recorded and were not been documented in the 1970s survey. The size and distribution of the majority of the Bronze Age settlement sites, especially those on the west bank of the Nile, suggest many are most likely small outpost or campsites. Their main purpose would probably have been linked to expeditions into the desert, providing shelter and resupply opportunities. The results from excavating two of the small settlement sites, AtW 001 and AtW 002 support this interpretation. Conversely, other ‘New Kingdom’ sites relocated by the DiverseNile project were actually later than the dates provided by Vila and his team.

3.2.1. AtW 001

AtW 001 is a small site on the west bank of the Nile (Figure 9). It was at first unclear if this site was identified by Vila and his team, as the only site identified nearby in the 1970s, 2-T-62, was described as a series of dismantled structures, hearths and concentration of pottery on a series of small mounds. One rectangular structure was excavated at the time and is quite clearly not the same as the mound excavated in 2022 and 2023, as AtW 001 is slightly further south-west. However, it seems probable that AtW 001 is one of the other mounds identified during the 1970s.

AtW 001 was first partially excavated in 2022 by the DiverseNile team before more extensive excavation in 2023 (see (Budka et al., 2023) for a more detailed report on the results of the excavation). These excavations revealed that the site was composed primarily of a large midden which accentuated the natural mound where the site is located. This took the form of several layers of, in places, relatively compact debris (up to 0.75 m high overall) composed of a mixture of ceramic, animal bone, and mud-brick remains in a sandy-silt matrix. Distinguishing between different layers of the midden was tricky and it was partially excavated in arbitrary spits where it was not possible to identify distinct stratigraphic units.

While there was potential evidence for storage on the site, in the form of small pits, it seems most likely this was a refuse area for the surrounding area; although, a number of postholes may suggest that a campsite was erected near the south-east corner of the mound. The relationship with the midden is unclear as this area was not covered in the same amount of debris, what material there was likely fell from the mound. Given the lower level of the post-holes they most likely pre-date the creation of the midden. Beneath the main part of the midden there were ephemeral remains of mud-bricks, although the preservation was too poor to determine distinct structures. These could have been used to delineate different parts of the site, rather than representing formal structures, or may have been part of an earlier occupation phase, possibly coinciding with the post-holes. The ceramic evidence dates the site to the early-to-mid Egyptian 18th Dynasty and also includes typical Kerma pottery.

Based on location and archaeological evidence, as well as the multicultural nature of the site, it was likely linked to market or supply activities in the area. This is corroborated by the material culture excavated, in particular the pottery which contained mostly a mixture of Egyptian and Nubian sherds, some of which were hybrid in nature. There were also Levantine wares which would have been imported. This combined with the large amount of animal bone found at the site suggests that AtW 001 was linked to expeditions into the desert and was most likely used seasonally, which would explain the lack of permanent structures. This is further supported by the geology and calculation of DEMs for the surrounding area, which show that AtW 001 would have been on small mound. Combined with the documentation of a possible palaeochannel running east to west just north of the site it may well have been located on a former, possibly seasonal, island.

3.2.2. AtW 002

AtW 002 (Figure 10), is another small domestic site located further into the desert, on the northern edge of the main palaeochannel, recorded in 1973 as 2-S-54 (Vila, 1977b, p. 85). The site is primarily composed of a rectangular structure made of a mix of mud-brick and dry-stone architecture, with a circular structure to the north. The rectangular structure (6.5 × 3.5 m) was excavated in 2023, although unfortunately the site was looted and destroyed part-way through our excavations. Although, initial excavations suggested the structure was divided into three parts, including an open courtyard in the centre with a hearth. Ceramic and ¹⁴C dating of the rectangular structure indicate it was in use in the Classic Kerma and early Egyptian New Kingdom periods (c.1688–1517BC). The architecture and artefacts found at the site have traits which would typically be associated with both Egyptian and Kerma traditions which makes it likely that the site was used by a mixed community who did not strictly adhere to one or the other. This further demonstrates the problematic labelling of sites in the region based on cultural categories primarily developed for urban centres in the region; arguably problematic in itself but with even less bearing on these more rural communities.

3.2.3. Re-Dated Sites

Other sites which were identified by Vila and his team as dating to the Egyptian New Kingdom in Ginis West actually date to between 750–600BC during the Napatan period. The 1970s survey team misattributed the wheel-made ceramics they found on the surface to the New Kingdom when they are actually more recent, c.8th century BC. These include three sites, 2-T-53, 2-T-57, and 2-T-69 which are composed of small huts and stone walls spread across a relatively large area, located in a region referred to as ‘Sand Hills along River’ (as noted on an 1886 map, see (Woodward et al., 2017, pp. 228–229, Figure 1) and still largely covered by sand.

2-T-53 is a relatively small site 150 m from the Nile with nine structures, three circular and six square (4–5 m) identified and one excavated in the 1970s (Vila, 1977a, p. 114). We further documented 2-T-53 in a foot survey and with drone aerial photographs. Based on this additional documentation, in particular the calculation of orthoprojections and DEMs we were able to locate the nine structures and to identify the rectangular one Vila excavated (Figure 11). There also appear to be more structures to the north as well as stone walls which were not previously documented, as well as at least four, rather than three circular structures. Remote sensing by the DiverseNile project has identified additional structures at a number of other sites previously recorded by Vila and the 1970s survey team. Many of these structures would have been difficult to identify on the ground due to the scattered archaeological evidence in a sandy-dune landscape.

3.2.4. Summary

Based on the evidence from Bronze Ages sites in this geographically peripheral region, using the Contact Space Biography model it seems more productive to consider this region as more of a borderscape in line with the work conducted by Maria Gatto and the Borderscape project (Siegel et al., 2024), see also “https://www.borderscapeproject.org/ (accessed on 23 December 2024)”. The local population may have helped provide various local and nonlocal groups with access to the desert as well as possibly working there themselves. The small sites on the west bank could as such be seen as a series of outposts, providing shelter and possibly supplies. The use of these areas by multiple people would explain the multicultural nature of many of the sites and also tie into models of gold working sites in other parts of the world (e.g., Mills, 2011).

3.3. Landscape and Riverscape

Out of a total of 283 sites in the MUAFS concession area, 58 settlement and cemetery sites could be dated to the Kerma period and 20 to the Egyptian New Kingdom. This includes 37 sites which were previously unrecorded such as MUAFS 003 and MUAFS 059 which provide additional evidence and increase our understanding of the land- and riverscape. The mapping of all Kerma sites (Figure 12) between Attab and Ferka shows a marked clustering of settlement sites in Attab West and Ginis West. The far fewer New Kingdom sites show a similar pattern in Attab and Ginis but no presence north of Kosha (Figure 13). During the New Kingdom, Attab and Ginis would have been in the hinterland of the urban site of Amara West (the same region may also have been the hinterland of Sai Island during the Kerma period).

The mapping of dry-stone walls adds significant new information to the ancient use of the districts of Attab and Ginis (Budka et al., 2023). The mapped features show a clustering in three main areas: along the west bank of the modern Nile, in the main local palaeochannel (the Northern palaeochannel according to Woodward and colleagues) and in a wadi (river valley) running north-east to south-west between AtW 001 and AtW 002. The spatial structuring of the stone walls in the region can be contextualised by recent research. Dalton and colleagues published an analysis of river walls, known as groynes, in the Amara West region to reconstruct the long-term landscape history and practice of hydraulic engineering in the region (Dalton et al., 2023). The authors identified at least six different construction styles of river walls, ranging from dense layered walls to indeterminate stone scatters. In Attab West, we observed a similarly wide variety of building materials, techniques, forms, and orientations in the landscape, including flat and upright stone mounds as well as walls with and without internal fill.

In the MUAFS concession area, the stone walls appear to have been constructed during almost all phases of human occupation in the region, from Kerma through to the New Kingdom, Napatan, Medieval and Post-Medieval periods (see Budka, 2019). For the earlier walls associated with the Kerma sites, the arrangement of these walls along and within the palaeochannels suggests that they may have served as wadi floodwalls. This is of particular importance since Woodward and his team have shown that the environmental conditions of this region have changed significantly over time (Woodward et al., 2017). The New Kingdom was a time of significant environmental change, in particular with water channels in the region drying up (see Woodward et al., 2017). Dalton and colleagues suggest this occurred sometime during the 18th Dynasty, and no later than 1000BC based on the dates of the sites along the channel (Dalton et al., 2023). In the MUAFS research area, the large palaeochannel (aka, the Northern palaeochannel) in the districts of Attab and Ginis is of prime significance and needs to be investigated further. For the time being, the local landscape of sites such as 2-R-18 in the Amara West district and AtW 001, and AtW 002 in the Attab district, during the Classical Kerma and early New Kingdom periods, can be reconstructed as crisscrossed by old Nile channels and the aforementioned sites being located on islands—possibly seasonal islands during the flooding season. Although it has been proposed that after the end of the New Kingdom, the active Nile channel must have fully dried up based on apparent lack of activity along the channel (Dalton et al., 2023, p. 19), the location of the newly identified Napatan sites in the eastern part of the palaeochannel suggests that our understanding of changes in the riverscape is still limited, particularly in terms of absolute dating.

An association of circular structures with the stone walls in Attab and Ginis was already noted by Vila (e.g., for site 2-T-53, see above). While some of these circular structures were most likely huts, others with a small diameter of 1.2–2 m could also be linked to irrigation. This was recently suggested for Vila’s site 2-T-67 (Vila, 1977b, pp. 93–96) which comprises three basins and channels linked to post-flood irrigation by Dalton and colleagues (Dalton et al., 2023). They suggest very convincingly that the site, and in particular the circular features, one of which was surrounded by a stone setting, actually served as wellheads for shadufs (irrigation tool used to lift water). A similar interpretation might also be possible for other circular features, for example at site 2-T-53. Our results from remote sensing show that these circular features tend to be located in low areas and not on mounds like the rectangular structures. This makes a link to post-flooding irrigation very likely and raises questions regarding seasonal Nile channels and the extent of land used for agricultural.

Remote sensing, mapping, and landscape analyses demonstrate significant changes to the landscape and riverscape in the region which clearly had important impacts on local populations. Some of these are very evident in the distribution of sites over time and in attempts to manage environmental and seasonal changes, potentially with flood defences. Further geoarchaeological work including OSL dating is needed to better appreciate these changes and potential adaptations in order to fully integrate the results into our concept of Contact Space Biography.

4. Discussion

The DiverseNile project has adopted the principle of the Contact Space Biography—a bottom-up approach which combines the concept of contact spaces with landscape biographies. The search and exploitation of natural resources such as gold would not have been possible without access to the goods, knowledge, resources and waypoints established by those living in the ‘hinterlands’. The bottom-up approach provided by the Contact Space Biography allows us to fully appreciate the importance of local populations to broader economic and social interests.

The idea of a more decentralised power in the area has already been effectively argued by Walsh (Walsh, 2022) through the idea of ‘marginal communities’ following Emberling’s (Emberling, 2014) model of a decentralised Kerma state. In times of increased Egyptian presence in the region and potential tensions as well as trade with the Kerma state, regions, such as the MUAFS concession area, take on a role not only as contact spaces between different groups—with the inherent influence on social practices and material culture that brings—but also as border zones providing access to more inhospitable areas with little discernment or care as to who that access was being provided to.

The findings from the settlements, cemeteries, and broader work on the local landscape all indicate a highly complex contact space with clear evidence of mutual interaction and influence based on a likely combination of social and cultural norms, personal decisions, shifting demands for natural resources such as gold, changing environmental conditions, and a certain level of pragmatism. Rathje & Sabloff’s concept of ‘mercantile pragmatism’ where, to avoid any potential conflicts between groups, as well as a ‘constant influx of goods […], information and merchants’ the ‘flexibility of cultural norms’ is key to avoiding conflicts (Rathje & Sabloff, 1975) and their port-of-trade model where a key aim is to avoid conflict between opposing cultural groups (Rathje & Sabloff, 1976) are worth considering further for this region. As are the meeting points and trade diaspores suggested for Egypt by Moreno García (Moreno García, 2021, pp. 198–203), that are also appealing to the so-called hinterlands in northern Sudan. Moving forward it seems that more work is needed on considering regions such as the MUAFS concession area as multicultural, or possibly transcultural, decentralised contact and trade area.

Therefore, following these models, it is worth pursuing further the idea of a decentralised economy in the MUAFS research area, one that was based on mercantile pragmatism and while linked to more centralised economies such as the ones on Sai Island and Amara West, themselves linked to broader Egyptian and Kerma states, was not fully dependant on them. This is supported by the cultural fluidity evident in both settlement and cemetery sites in the MUAFS concession area, with a mix of traditionally-local and Egyptian architecture and material culture, as well as numerous examples of ‘hybrid’ ceramics. Evidence from GiE 003 shows a similar intermixing of traditions while certain traits, perhaps those with a greater cultural significance that we cannot fully comprehend based on archaeological evidence alone, remain clearly emphasised. The diversity in burial styles in this cemetery provide significant insights into Kerma mortuary practices and both how these progressed over time and were influenced by social hierarchy.

Feature 50 provides clear evidence that Pan-Grave communities were present in the Attab–Ferka region between c.1800–1700BC (for dating aspects see (de Souza, 2022, p. 189). This is significant in itself, as there was previously no evidence for Pan-Grave material in the Nile Valley between the Second and Fourth cataracts (de Souza, 2019, pp. 82–89) while also providing an interesting example of different cultural and local influences. The presence of aspects associated with the Kerma culture, such as the bedframe and some of the ceramics, combined with a typical Pan-Grave circular grave and grave goods (ceramics and ivory artefacts, see Figure 7) renders Feature 50 a local variant with key differences from Pan-Grave cemeteries further north (for typical Pan-Grave burial assemblages in Upper Egypt and Lower Nubia see (de Souza, 2019)) but very similar to the evidence found at Al-Widay much further south in the Fourth Cataract region (de Souza, 2019). Regional differences between Pan-Grave communities are well attested to in the archaeological record. But the distinction between Pan-Grave and Kerma populations are often less evident than compared to other cultural groups they intermingled with in the Nile Valley (de Souza, 2019, p. 14). This makes it difficult to assess whether the localised variation in graves at GiE 003 is due to the presence of both Pan-Grave and Kerma populations or primarily influenced by one or the other. Although it remains difficult to assess in detail at the current state of work, it is tempting to view the connections between the community buried in a rural Kerma cemetery with nomadic people in relation to desert gold mining and/or wadi working sites close by. These interactions have left clear signs of cultural diversity, with regard to ceramics but also funerary practices. When considering social practices, it seems crucial to take into account the seasonality of gold mining and other activities and to examine what archaeological traces were left by subsistence strategies such as agropastoralism (as opposed to sedentary communities or nomads). The latter also reminds us that communities were dynamic and could both modify their means of subsistence or combine different strategies depending on a number of factors, making a realistic reconstruction for archaeologically tangible groups very difficult. The link to gold working also allows us to consider models of different types of gold working sites used in other parts of the world. For example, Mills (Mills, 2011) suggests a model of five different types of settlement system for gold-mining based mostly on evidence from the gold-rush in north America worth applying to the MUAFS research area:

‘Entrepôts’ described as ‘major supply centers outside the gold extraction region’;

‘Intermediate transfer points and […] supply points’;

‘Central distribution centers’;

‘Secondary distribution points’;

‘Extraction camps’.

Following this model, Sai could be seen as the central distribution centre and sites such as AtW 001, AtW 002 and 2-T-53 as supply (and possibly transfer) points. Extraction camps based on our modern knowledge of gold prospection in the region are likely to have been campsites and leave little evidence in the archaeological record but would have been a key part of cultural, social and economic life in the region (see, for example campsite, 3-P-8, at Kosha (Budka, 2024b, p. 86). The Contact Space Biography approach allows us to properly include such sites into our archaeological assessment of the study area, despite only often leaving ephemeral, if any, traces.

5. Conclusions

The preliminary results of the DiverseNile project show that our bottom-up approach allows us to challenge the well-established categorisation of sites as either ‘Egyptian’ or ‘Nubian’ through the application of Contact Space Biographies and considering evidence of cultural fluidity. This method supports an alternative narrative regarding one part of Kerma and New Kingdom Nubia which could be relevant to general understandings of the Bronze Age Middle Nile. There is a fundamental necessity to include social practices, communities and subsistence strategies within ‘marginal’ regions when considering the complexity of cultural processes and encounters, often addressed solely through discussion of more central sites. Based on settlement sites, the spatial and temporal dimensions of the landscape for the communities living in the Attab West and Ginis West districts can be addressed. The environmental conditions in this landscape markedly changed over time. Woodward and his team (Woodward et al., 2017) have already illustrated that the New Kingdom was a time of significant environmental change with water channels drying up in the region; this was further confirmed recently by Dalton and colleagues (Dalton et al., 2023). In the MUAFS concession, this will be further investigated along the large paleochannel in the districts of Attab and Ginis.

Cemetery GiE 003 can also be used as a case study for identity formation and community-building in a contact space of the Middle Nile. The new evidence from Trench 5 allows us to propose that people associated with the Pan-Grave horizon were present, in contact and in exchange with the local communities and probably part of the social structure of the Attab to Ferka region between c. 1750–1650 BCE (and possibly longer). Whether these people had a specific space in the cemetery to be buried, remains unclear due to the selective state of excavation (but note the general chronological differences in GiE 003 which suggest a south-north development of tombs; see also (Budka et al., 2023, p. 20)).

The sometimes-subtle shifts in social and cultural practices at many of the sites in the MUAFS concession area have multiple causes that we have yet to fully appreciate. However, it is clearly over-simplistic to suggest that these sites can be identified as either Egyptian or Nubian or, indeed, occupied by people who would have identified as either. The ‘cultural’ markers evident at both cemetery and settlement sites would have been just as influenced by pragmatic and personal decisions. Overall, the concept of the Contact Space Biography effectively allows us to combine models of contact spaces, the idiosyncrasies of a changing landscape and the technological and industrial prerogatives of those living in and accessing this region. While we are still refining the concept for the MUAFS research area, a similar approach could be applied to many other regions of the world. In particular in terms of considering the local realities of cultural interactions away from the more carefully cultivated identities of ‘core’ sites.