the archaeology of the eastern sahel, sudan: preliminary results

The African Archaeological Review, 2 (1984), pp. 173-188

The archaeology of the Eastern Sudan: preliminary results

Sahel,

R O D O L F O F A T T O V I C H , A N T H O N Y E. M A R K S and A B B A S M O H A M M E D - A L I

Abstract

This paper reports the combined preliminary results of the Italian Mission to Kassala and the joint University of Khartoum/Southern Methodist University Butana Project. Both groups have been carrying out extensive survey and test excavations in the Southern Atbai of the East Central Sudan, between the Atbara and Gash Rivers. This work has led to the recognition of a previously unknown culture area and ceramic tradition which spans about 5000 years and, at its peak, probably covered about 100,000 sq. km east of the Nile. This ceramic tradition, the Atbai Tradition, is associated with the development of large villages (over 10 ha) during the 4th millennium bc.

R6sum6

Le pr6sent article rapporte les r6sultats pr61iminaires combin6s de la Mission italienne ~t Kassala et du projet Butana entrepris en commun par les universit6s de Khar toum et Southern Methodist. Ces deux groupes ont r6alis6 des reconnaissances arch6ologiques extensives et des sondages dans le sud de l'Athai, dans la partie centrale orientale du Soudan, entre les rivibres Atbara et Gash. Leur travail a men6 ~ la reconnaissance d'une civilisation et d 'une tradition c6ramique auparavant inconnues qui durbrent quelque 5000 ans et qui couvraient ~t leur apog6e environ 100,000 km 2 ~t l'est du Nil. Cette tradition c&amique, la tradition Atbai, est associ6e au d6veloppement des grands villages (de plus de 10 ha) au cours du 4 e mill6naire avant notre ~re.

The information and interpretations reported in this brief communication result from the combined work and thinking of two separate archaeological projects which found themselves, unplanned, working in adjacent areas of the Eastern Sudan. These projects, the Italian Mission to Kassala Province and the Joint University of Khartoum/Southern Methodist University Butana Archaeological Project, both had as their goals the elucidation of the culture history of the Eastern Sudan, with specific emphasis upon the relationships between environmental change, economic adaptations (including the introduction of food production), and settlement systems, as they related to adjacent areas.

In the case of the Italian Mission, the focus lay in connections with Ethiopia and

174 Rodolfo Fattovich, Anthony E. Marks and Abbas Mohammed-All

particularly with Pre-Axumite and Axumite developments (Fattovich 1980, 1982). For the Butana Archaeological Project it was to test the degree of cultural interaction between the Nile Valley and the Eastern Sudan during the general period of 'Neolithic' development (Marks et at. 1980). In spite of these differing emphases, the first few field seasons demonstrated to both projects that each was working in part of the same large culture area and that cooperation could result in a breadth and depth of coverage that neither project

could accomplish alone. This preliminary note represents the first published fruits of this scientific cooperation. I t

is based on the joint examination of materials from our respective research areas, the

exchange of field personnel, the drawing up of unified typologies, adoption of the same terms for the various cultural units recognized in our work, and the coordination of research strategies. Our combined work, even though it has really just begun, has brought to light a

large amount of new data. These data lead to a clear but quite unexpected conclusion; there was a large area of the Eastern Sudan (Fig. 1) in which developed a distinct ceramic tradition, called here the Atbai Tradition, which arose during the 5th millennium bc and lasted until the 1 st millennium ad. This ceramic tradition, as far as our data indicate, was not

significantly influenced by either Nilotic or Ethiopian developments. By the middle of the 3rd millennium bc it appears to be present in the Central Nile Valley on the west and in the Red Sea Hills on the east, covering an area of over 100,000 sq. km. In addition, our work indicates

that this area saw the growth of large villages, at least in the core area, as early as the middle of the 4th millennium bc. These villages were at least twice the size of most contemporary

villages of pre-Dynastic Egypt! The methodology employed by both projects involved the survey of geographically distinct

regions within the eastern Sahel of the Sudan. The Butana Archaeological Project in two field seasons carried out systematic survey and test excavations in two areas and a brief

reconnaissance survey of an intermediate area. The westernmost area was just 50 km east of the Nile Valley, including the deeply stratified midden and cave complex of Shaqadud, as well as 60 sq. km around that site (Fig. 1). The second area, 320 km to the southeast, centered around the Atbara River at Khasm el Girba and extended eastward toward the Gash Delta. This encompassed an area of about 2000 sq. km, of which about 400 sq. km have so far been

systematically surveyed and sampled. The reconnaissance was done in the Central Butana

near the Maya es Sawad, some 60 km east of Shaqadud. The Italian Mission, over four field seasons, carried out systematic survey and testing in a

ca 1000 sq. km area around and south of Kassala in the southern Gash Delta. Brief reconnaissance surveys and testing were also undertaken at the edges of the northern Gash Delta, some 140 km.north of Kassala, in the Delta proper, as well as in a 100 sq. km area to

the west of the Delta (Fig. 1). While the area east of the Nile Valley in the Central Sudan was basically unknown

archaeologically (except for the important Meroitic ceremonial sites in the Western Butana), some work had been carried out prior to our present studies. As early as 1917, Crowfoot (1928) traveled to Kassa la - - in those days no small fea t - -and recorded a major site there named Mahal Teglinos, which is still the largest known site near the Gash Delta (Durante et al. 1980). Over the next thirty years three different people traveled eastward noting some sites: Kirwan in the 1930s, Sandison in the 1940s, and Delany in the 1950s. None, however, published his results, merely leaving notes and small collections with the Sudan Antiquities

The archaeology of the Eastern Sahel, Sudan: preliminary results 175

SUDAN

'~ ERKOWIT

o ~ ~ r..~-~

t /

! /.

.t

L ~

-.~ el

7~

1 i

I"*~ )

I t 3s" I

RED

SEA

i. I

i

S ~- . ~ I

. i l ~ t ' l J "J

• Baraka

A G ~ A R A q

/i / J

(, i

o so IOO t n I

Figure l M a p of the eastern Sahel, showing the core area and sites ment ioned in the text.

176 Rodolfo Fattovich, Anthony E. Marks and Abbas Mohammed-Ali

Service in Khartoum. More intensive work was undertaken in the 1960s, both on the western fringe of the Butana and along the Atbara River. Aside from the Meroitic studies of Hintze (1960), Otto (1963, 1964) test excavated the 'Neolithic' site of Shaqadud near the Meroitic site of Naga, and Shiner (1971) carried out a one-month survey around the town of Khashm el Girba, on the Atbara River. Otto's work never saw full publication, while Shiner's pioneering efforts were truncated by political events and the significance of his preliminary finds was, although noted (Clark t976; Phillipson 1977), not appreciated. This previous work did provide enough data to indicate that the eastern Sudan held some archaeological potential, and both of our projects owe a debt to this work.

In addition to these efforts, some data are available from individual sites in the eastern Sudan and in Eritrea. Recently, Callow and Wahida (1981) reported a site near Erkowit in the Red Sea Hills and, earlier, Arkell (1954) described material from four mound sites near Agordat in the Baraka Valley, Ethiopia. Although geographically isolated, these sites are significant in light of our recent finds.

The study area in the Sudan, east of the Nile and west of the Ethiopian border, consists mainly of flat, dry grasslands with a scattering of Acacia trees, and is cut from south to north by two main drainage systems: the Atbara River which separates the Butana proper from the Atbai and the Gash River which cuts through the Atbai at the junction between the steppic grasslands and the tree savanna of western Eritrea (Fig. 1). In spite of these major drainages and a few outcrops of rock, forming inselbergs, the overall impression of this huge area is one of unending, featureless steppe. Actually, the area is full of shallow, seasonal drainages and minor topographic ditterences which have significant effects on the environment of their immediate surroundings. On a broad scale, the area is driest on the west--ca 73 mm of rain per year at Atbara--increasing in seasonal rainfall toward the east until it reaches ca 325 mm per year at Kassala. Today, much of the area is farmed seasonally, after the summer rains, but is mainly used as pasture for livestock and camels. During the driest parts of the year most of the population moves to the river valleys.

Using only the data recovered by the two projects in the Southern Atbai, that area between the Atbara River Valley and the eastern edge of the Gash Delta, it has been possible to construct a preliminary culture-historic sequence which can be used to understand the work in the more disparate areas of the eastern Sahel. This core area, of some 1500 sq. km, has seen the recording and sampling of 175 ceramic-bearing sites and test excavations at 17 of these. This work has resulted in the collection of tons of ceramics and lithic materials, faunal remains, and charcoal and macrobotanical samples. Although these are all still actively under study, the broad outline of the Atbai Tradition and the other archaeological units in the area is already becoming clear. On the basis of primarily ceramic studies, in conjunction with a series of radiocarbon dates, we have constructed a tentative matrix of assemblages into a major ceramic tradition, divided into temporal phases (Fig. 2). Within these phases we have been able at times to recognize regional facies or groups of assemblages that are essentially contemporaneous, share a similar level of complexity, but which show some specific differences in ceramic components. It is recognized that there may well be additional facies discovered for each phase as intensive field work expands into the marginally known areas covered by the Atbai Tradition. What we now have is merely a small portion of the details of regional development within the eastern Sahel. Emphasis here, therefore, will be

placed on the core area.


DATES

lO00 ad

0 ad

I000 bc

2000 bc

3000 bc

4000 bc

5000 bc

v)

PHASE

ATBARA

Gergaf . . . . .

Groups by Area

->

STEPPE GASH ?

. . . . Gergaf

Jebel Taka

V

S/K tr

Saroba

Hagiz . . . . -> <- . . . . . Hagiz

Jebel Mokram

Butana - - /t',

site KG28

Malawiya

>

Gash

site KG 14

Amm Adam

RELATED SITES

Ago~dat

Shaq~adud ~ /Cave

Erkowit

Figure 2 Diagram of spatial and temporal arrangement of units within the Atbai Ceramic Tradition. Stars in the left-hand column represent preliminary I~C dates.


On the broadest level, the Atbai Tradition is defined by a time transgressive complex of

technological and stylistic components used in the production of ceramics. Most character- istic is the use of scraping, not only as a means of thinning vessel walls but also as a decorative technique. While the proportional occurrence of this varies from phase to phase and from group to group within each phase, it is always present except in the earliest sites of the Atbai Tradition. In addition, during much of its duration, the tradition can be recognized by a

limited range of decorative motifs which are found on and just below the rim of open bowls. Also, during some phases, there are types and styles of decorative motifs that are simply unknown in other areas. For instance, during the Kassala Phase, the Butana Group has a

small but significant percentage of fine black ware open bowls with interior burnished geometric decorations. In the Gash Group of the same phase there are open bowls with rims decorated with a thin band of parallel, vertical incised or impressed lines which are unknown

in the Butana Group or to the west of the Atbara. Presently it appears that the earliest ceramic sites in the Southern Atbai represent at least

two different regional groups. One site (KG 14), found along the Atbara River near Khashm el Girba, has been dated to the late 6th millennium bc and its pottery appears to be related to the general ~Khartoum Horizon Style' (Hays 1971, 1976). Some other sites, located at the

northern end of the Gash Delta, named the Amm Adam Group (Fattovich and Piperno 1981, 1982), have pottery which is quite distinct from the 'Khar toum Style' but shares a very distinctive knobbed ware with site KG14 near Khashm el Girba. I t is quite possible that

these groups are the base from which the Atbai Tradition arises, since this knobbed ware is found in small quantities in the first, Saroba Phase, of the Atbai Tradition. However, there

are too few sites and, as yet, too little data to permit a secure attribution. In spite of their differences, both regional groups exhibit similar adaptations. Both are

riverine oriented with exploitation of fish, hippo, and mollusks, as well as savanna edge mammals , such as warthog, and a wide range of small-to-large bovids (oribi, duiker, topi,

reedbuck, hartebeeste, buffalo, etc.). While no macrobotanical remains have been recovered, grinding stones are common at KG14 but absent in the Amm Adam Group. The systematic survey of over 800 sq. km of the steppe between the Atbara and the Gash failed to reveal a

single site of either group. The earliest recognized phase of the Atbai Tradition is represented by sites of the

Malawiya Group which occur at numerous localities on the steppe between the Atbara and

the Gash Delta but which have not, as yet, been located along the main drainages themselves. A single radiocarbon date suggests it was present during the 5th millennium bc. Sites are

characterized by an abundance of Pila shell, indicating at least seasonally swampy terrain. The ceramics are mainly a sand-tempered, unburnished buff-colored ware with impressed decorations (Fig. 3). These decorations fall broadly within the Khar toum Horizon Style, but lack the typical wavy line, dotted wavy line, the banded, and other motifs typical of Early

Khar toum (Arkell 1949). I t appears that the Saroba Phase is sufficiently homogeneous that only a single group can

be recognized. Its adaptation was strongly oriented to the hunting of small bovids, warthog and monitor lizard (Varanus niloticus), although other, larger mammals were also taken. Although no macrobotanical material has been recovered, grinding stones are common. Sites tend to be small, with an average area of ca 5000 sq. m.

The Saroba Phase is recognized as the earliest of the Atbai Tradition because its ceramics


Figure 3 Typical Saroba Phase sherds with impressed decoration: 1, 3. mat impressed; 2, 4, 5. zigzag rocker stamped.

continue in small percentages into the Transitional Saroba/Kassala Phase. So far, only a single site of this transition is identified securely. It is located on the steppe and is also

associated with a large concentration of Pila shell. A radiocarbon date places it in the early 4th millennium bc. The transitional nature of the site may be seen in its increased size (15,000 sq. m), the presence of both typical Saroba impressed wares and typical early scraped wares, along with a hard, burnished, vertically rippled ware which is unique to this transitional phase. Adaptation was no different from that of the Saroba Phase, as far as the faunal remains are concerned, and again ground stone is present.

According to our present knowledge, the Kassala Phase is both the most complex and most developed of the Atbai Tradition. On the basis of eight radiocarbon dates, this phase appears

to span about 2500 years, from the middle of the 4th to the end of the 2nd millennium bc. Within this phase, it is possible that two distinct sub-phases exist but more data are needed to define them. The first half of the Kassala Phase is represented by eight sites of the Butana Group and three sites of the Gash Group. The former are located between the western bank of the Atbara River and a point midway between the Atbara and Kassala, while the latter parallel the Gash River, with one near Kassala and two others at the northwestern end of the Gash Delta near Eriba (Fig. 1). The distinction between the groups lies in proportional variations of the scraped wares (Fig. 4:5) and in the presence of some decorative patterns


Figure 4 Typical Kassala Phase sherds from the Butana Group. 1. rim with short, deep incisions; 2, 3. fingernail impressed; 4. restricted orifice bowl sherd with slipped and burnished rim and rocker stamped wolftooth; 5. scraped sherd with pinched rim; 6. as 4 with incised decoration.

which are limited to one or the other group (Fig. 4:1-3). However, a thick scraped ware is ubiquitous and each site has a variable number of thin, well fired, decorated and burnished

wares (Fig. 4:4, 6; Fig. 5:1-3, 5, 6). The Kassala Phase may be generally characterized by large village sites, both along the

major drainages and in the steppe between them. Site size varies but sites range from 45,000 sq. m to the largest which approach 120,000 sq. m and which have over 2 m of in situ deposits. Test excavations in the two largest sites (Mahal Teglinos of the Gash Group and


KG23 of the Butana Group) indicate that, at both, the large village development was in place prior to the introduction of domestic cattle and small livestock (Costantini et al. 1982). In fact, at both sites, domestic animals represent only a minority of the fi~unal remains during this part of the Kassala Phase. The majority are represented by small bovids of the type exploited since the first ceramic bearing occupations of the area. Tile Kassala Phase sites adjacent to the major drainages indicate a reintroduction of fishing and the exploitation of other riverine animals (tortoise, crocodile, hippopotamus). This is also the phase during which a range of polished stone tools appear, including axes and mace heads, as well as a wide range of ground stone of various forms.

The Kassala Phase, both in its complexity and in specific ceramic components, can be seen as far north as Erkowit in the Red Sea Hills and east in the Baraka Valley at Agordat. In addition, the cave at Shaqadud (Marks et al. 1982) contains ceramics strongly linked to the Gash Group. This also would seem to apply to small concentrations of ceramics collected in the Nile Valley near the Third Cataract and referred to as 'Groupe II ' by Geus (t976, 1981). Thus, during the 3rd millennium bc there is good evidence for the presence of Atbai Tradition ceramic assemblages from almost the Red Sea into the Nile Valley.

By the end of the Kassala Phase (in the 1st millennium bc), there appears to be a marked decrease in site size, with most sites below 50,000 sq. m. Faunal remains now indicate a heavy

Figure 5 Typical Kassala Phase sherds from the Butana Group. 1, 3. zoned impressed; 2, 5, 6. interior burnished decoration; 4. black burnished with impressed banded decoration.


dominance of domestic forms and the number and variety of fine wares drops dramatically. However, the basic ceramic patterns remain intact, as do the polished and ground stone

components. Finally, the last phase of the Atbai Tradition in the core area, the Jebel Taka, is

represented by the Hagiz Group which is present both in the steppe and near to both major drainages. Sites are now relatively small, no larger than 20,000 sq. m and they lack any true

depth to their cultural deposits. While fauna is rare, owing to surficial cultural deposits, all material so far found has belonged to small-sized domestic cattle. The Jebel Taka Phase is characterized by a decadent form of scraped ware, very little decorated pottery, and a few sherds which suggest links to the Meroitic and the Pre-Axumite cultures to the west and east. These sherds indicate a date ti~om the end of the tst millennium bc to the middle of the 1 st millennium ad. In addition, a ti~w sherds of Mediterranean amphorae were found at one site

in the southern Gash Delta, indicating that this phase may last almost to the end of the 1 st

millennium ad. Over 50 sites are attributed to theJebel Mokram group, which, with a radiocarbon date in

the 2nd millennium be, appears contemporary with the second half of the Kassala phase. The group seems to fall within the Atbai Tradition but differs significantly in detail from the Butana and the Gash Groups. It is possibte that it represents a seasonal spread of a regional

group, originating outside the core area, perhaps in Eritrea. All sites occur on the steppe and, while they may be as large as 10,000 sq. m in area, they never display any depth ofcuhura l deposit and all seem to represent rather ephemeral camps. The ceramics include a thin,

scraped ware but the most common pottery is a sand tempered, undecorated brown ware. Decorations are characterized by deeply combed surfaces associated with slipped and burnished rims, as well as by shallow incised crossing lines on unburnished surfaces (Fig. 6). There are a number of other elements which are limited to this group, such as lugs and

ceramic strainers. Although ground stone is not common, polished stone bracelets and flat

mace heads are present at most sites. Given the lack of true in situ deposits, faunal material is rare, but when found always refers to

domestic cattle. In spite of site conditions, numerous thick sherds from special hearths con-tain

large amounts of macrobotanical materials. Some of these have been identified as domestic sorghum (Costantini et al. in press), representing the earliest evidence for this indigenous African domestic crop so far found by some 1500 years (Clark and Stemler 1975; Haaland 1981:196). Given the abundance of these remains, however, it is likely that its development as a domesticate took place well prior to its appearance in the Jebel Mokram Group.

Thus, the floral and faunal remains clearly indicate that the Jebel Mokram Group was

characterized by a mixed herding/farming strategy that may well have paralleled that of some local Beja groups today. The absence of larger Jebel Mokram sites along the major drainages may indicate that we are seeing only one end of a t ranshumant seasonal cycle, perhaps comparable to that of the present day Beni Amer (Barbour 1964).

Finally, there are other groups of assemblages which appear to be relatively recent and lack any Atbai Tradition elements. Included are one Christian site at the northwestern end of the Gash Delta, a series of campsites of the Korak Group, which is.of unknown date and characterized by undecorated red slipped chafftempered pottery, as well as a number of sites forming the Gergaf Group, which is characterized by pottery with incised, cross-hatched decorations below rims which are always flat. This pottery type seems to relate to sub-recent


Figure 6 TypicalJebel Mokram sherds. 1-6. incised motifs; 7. punctate.

times, although it may begin before the t5th Century AD. I t may well last until this century, since a few of the sites contained small amounts of glass. These sites are all rather small and are mainly limited to the badlands adjacent to the major drainages. On the steppe, this group is seen only as extremely minor scatters ofsherds which make theJebel Mokram sites seem densely covered. The pattern of site distribution parallels that of the present Hadendowa Beja inhabitants of the Southern Atbai who use a very limited transhumance from the badlands along the Atbara River to the steppe, at most 15 km to the east, after the rainy season. This limited movement permits them to farm the steppe but, at the same time, to water their cattle at the river every day.


Although there are numerous details to be worked out, a general preliminary overview of environmental change can be presented. First, this overview must be limited to the Holocene, since with a minor exception along the west side of the Atbara River, all sediments in the area seem to post-date the Pleistocene. In particular, the steppe between the Atbara River Valley and the Gash Delta appears to have been built up by a series of successive deltaic depositions that formed as the Gash River changed its course from an original East-~¥est direction to its present ESE-WNW flow (Durante et at. 1980). This interpretation is strengthened by the absence of any preceramic assemblages or even isolated artifacts on the steppe or in the present margins of the Gash Delta.

On the most general level, there appears to have been an overall drying trend from the 6th millennium bc until the present day. All sites of the 6th and 5th millennia bc are associated with large numbers of Pila, an amphibious snail which requires seasonal inundation (Adamson et al. 1974). In addition, the presence of reed rat and arboreal monkeys in the pre- Saroba site of KG14 suggests a gallery forest extending along the Atbara, and about 500 mm or more of rain per year. Wetter conditions may also be indicated by the presence of hippo and fish remains in the Amm Adam Group sites in the northern Gash Delta, an area well

north of the recorded historic Gash flow. By the middle of the 4th millennium bc, sites on the steppe are no longer associated with

Pila but continuing seasonal inundation probably occurred in the northern Gash Delta, since Pila is found with sites dating as late as the 3rd millennium. Given the source of the Gash, well to the south in Ethiopia, such seasonal inundation is probably unrelated to local

precipitation. The megafauna from Kassala Phase sites include a wide range of bovids, from duiker to

buffalo. Giraffe and elephant are present but rare. All these forms indicate a humid grassland environment in the core area, on the steppe between the Atbara and the Gash Delta, up to

about the 2nd millennium bc. The presence of Limicolaria at most sites on the steppe and along the Atbara also indicates a level of precipitation above today's average (Arkell 1949). In spite of this, the faunal material from the stratified site ofMahal Teglinos (Kassala Phase andJebel Mokram Group) indicates that the fauna exploited reflected drier conditions than was the case for the fauna exploited at the Amm Adam Group sites (Geraads in press). When considering the megafauna, however, it must be remembered that the big savanna mammals (elephant, giraft~, rhino) were still living in the lowlands of northern and western Eritrea in

the second half of the last century (e.g. Munzinger 1890). Certainly, by the middle of the 2nd millennium bc, the climate was considerably drier,

since in the steppeJebel Mokram sites are often associated with well mounds (Sadr 1983), as if an attempt were being made to maintain the water sources naturally present during earlier periods, By theJ ebel Taka Phase, all occupations of any size are located near the major water courses, with only an occasional sparse scatter ofsherds on the steppe. It would appear that the steppe was already too dry to support long term occupation. The same pattern pertains

for the Gergaf Phase and for today, as well. The apparent temporal association between the disappearance of Pila on the steppe and

the appearance of large villages there probably reflects new economic potentials stimulated by the drying trend. The introduction of cattle and small livestock shortly thereafter can again be viewed as an adaptation to and an exploitation of these relatively new environmen-

tal conditions.

The archaeology o f the Eastern Sahel, Sudan: preliminary results 185

Although the data base from the core area of the eastern Sahel is impressive and becoming more so with each field season, our knowledge of the geographically peripheral zones is much less secure. The isolated sites ofShaqadud, Erkowit, and Agordat provide indications but, as yet, no sure answers. In the case of Shaqadud, however, excavations in the midden dated to the 5th and 6th millennia bc (Marks et al. 1982) indicate that during those millennia the western Butana was part of the general Nilotic ceramic tradition. Only by the middle of the 3rd millennium does the Atbai Tradition make a clear appearance. The lack of stratified deposits at Erkowit and the absence of excavations at Agordat prevent any realistic view of local development. However, it is expected that the area covered by the Atbai Tradition will vary through time in response not only to its own internal dynamics but also in relation to developments around its borders.

These border areas are those least known today but there is some indication of contacts with other areas and cultures even within the core area. There is no question but that the pre- Saroba and Saroba Phases belong, in a general sense, to the trans-Sahel ceramic horizon, represented in the Sudan by the Khartoum Horizon Style (Hays I97I, 1976). Yet it is unlikely that any inhabitant of the core area had direct links with even the Nilotic populations of that time. The extremely wide spread of this ceramic horizon suggests diffusion rather than migration.

There is one element, the knobbed ware of the pre-Saroba Phase sites of KG14 and Amm Adam, Site 1, which apparently lasts for some time in the north of the eastern Sahel and

finally reaches the Nile Valley at Kerma by the 3rd millennium bc (Privati 1977-78, 1981-82), long after its disappearance in the Southern Atbai.

The ripple ware found exclusively in the Transitional Saroba/Kassala Phase and dated to the early 4th millennium bc is suggestive of similar tbrms found in Nubia and Southern Egypt beginning about a thousand years earlier in Badarian contexts. The absence of other Badarian traits, however, would point to either a general diffusion or, perhaps, even independent invention. Similar pottery is found at Geili on the Nile in later context (Caneva 1980).

During the Kassala Phase there are some indications of extra-regional contacts. In the core area, these are limited to a very few pieces of obsidian originating in the highlands of Ethiopia

and a few decorative motifs which are similar to some from Early Kerma and/or the C-Group well to the north. Extra-regional contact was more intensive in the northern and eastern parts of the Atbai Tradition area, at Agordat and Erkowit. In the latter locality, considerable obsidian was found, indicating contact to the south (Callow and Wahida 1981). At Agordat obsidian was probably common, pointing to contact with the south, while axes, palettes, and car-spools are clearly reminiscent of New Kingdom Egyptian forms (Arkell 1954). Thus, by the 2nd millennium bc there was considerable contact between the eastern area of the Atbai Tradition and both Ethiopia proper and, perhaps, Egypt. It is quite possible that this area of northern Eritrea was, as previously suggested, the land of Punt (Kitchen 1971; Posener 1973, 1977), since no Egyptian influences are seen in our core area and none have ever been reported from Ethiopia proper.

In the Jebel Mokram Group a few pieces of obsidian have been found, but again, considering their relative proximity, it is striking how- little evidence there is for specific contact between the Southern Atbai and the Ethiopian highlands. On the other hand, there are general ceramic elements (lug handles, types of fabric, etc.) which might suggest that the


Jebel Mokram Group was part of an ancestral base from which some Pre-Axumite elements arose.

By the Jebel Taka Phase some actual sherds of Pre-Axumite pottery, as welt as some of Meroitic type are present in the Southern Atbai. Yet, given the oft-stated political conflict between Mero~ and Axum over the Eastern Butana and Southern Atbai (e.g. Adams 1977), it is striking how little concrete evidence there is for the presence of material remains of either kingdom.

In spite of all these examples of elements which are or may be indicative of extra-regional contact, it is important to realize just how little it all amounts to in the area which we have studied in detail. Only to the east, in the Baraka Valley do such connections seem potentially

significant and then, only in the 2nd millennium bc. For the greater part of its duration, it appears that the Atbai Tradition was almost wholly autochthonous and that by the 3rd millennium be it had spread so far as to replace the Nilotic Tradition in the Central Nile Valley.

Acknowledgements Both the Butana Archaeological Project and the Italian Mission owe thanks to the Sudanese Antiquities Service which not only gave permission for all our work but aided and encouraged it. The Italian Mission has been sponsored, since 1980, by the Seminario di Studi Africani, Istituto Universitario Orientale, Naples, and has been funded by the following: Consiglio Nazionale delle Ricerche; Ministero della Pubblica Istruzione (fondi per la ricerca scientifica 60%); Ministero degli Affari Esteri. The Butana Archaeological Project has been funded by National Science Foundation Grant BNS 8102649 with additional support from the University of Khartoum, Sudan. This project includes four Co-Principal Investigators: A. Marks, A. Mohammed-All, T. R. Hays, and Y. Elamin.

The faunal identifications for the Italian Mission were undertaken by Dr D. Geraads, Mus~e de l 'Homme, and for the Butana Project by Dr A. Gautier and Mr Joris Peters, University of Ghent. The floral materials reported here were identified by Dr L. Costantini, Museo d'Arte Orientale, Rome.

References Adams, W.J. 1977. Nubia: corridor to Africa.

London. Adamson, D., Clark, J. D. and Williams,

M. A.J. 1974. Barbed bone points from Central Sudan and the age of the 'Early Khartoum' tradition. Nature 249 (5453): 120-23.

Arkell, A.J. 1949. Early Khartoum. London: Oxford University Press.

Arkell, A.J. 1954. Four occupation sites at Agordat. Kush 2:33-62,

Barbour, K. M. 1964. The Republic of the Sudan: a regional geography. London: University of London Press.

Callow, P. and Wahida, G, 1981. Fieldwork in

Northern and Eastern Sudan 1977-1980. ~),ame Akuma 18:34-36.

Caneva, I. 1980. Early Neolithic settlement and later cemetery at Geili, Sudan. Paper presented at the conference on the Origin and Early Development of Food-producing Cultures in North Eastern Africa. Poznan (Sept. 1980).

Clark, J. D. 1976. The domestication process in sub-Saharan Africa with special reference to Ethiopia. In Origine de l'glevage et de la domestication (ed. E. Higgs): pp. 56-115. IX e Congr~s Internationale des Sciences Prfihistoriques et Protohistoriques, Colloque XX, Nice (Sept. 1976).

Clark, J. D. and Stemler, A, 1975. Early

The archaeology o f the Eastern Sahel, Sudan: preliminary results 187

domesticated sorghum from Central Sudan. Nature 254 (5501):588-9t.

Costantini, L., Fattovich, R., Pardini, E. and Piperno, M. 1982. Preliminary report of archaeological investigations at the site of Mahal Teglinos (Kassala), November 1981. Nyame Akuma 21:30-33.

Costantini, L., Fattovich, R., Piperno, M. and Sadr, K. In press. Gash Delta archaeological project: 1982 field season. Nyame Akuma 23.

Crowfoot, J. W. 1928. Some potsherds from Kassala. Journal of Egyptian Archaeology 14:112-16.

Durante, S., Fattovich, R. and Piperno, M. I980. Archaeological survey of the Gash Delta, Kassala Province. A~.ame Akuma 17:64-71.

Fattovich, R. 1980. Materiali per lo studio della ceramica pre-aksumita etiopica. Annali dell'Istituto Universitario Orientale di Napoli, 40 (4), Suppl. 25. Napoli: Istituto Universitario Orientale.

Fattovich, R. 1982. The problem of the Sudanese-Ethiopian contacts in the antiquity: status quaestionis and current trends of research. In Nubian Studies: Proceedings, Symposium for Nubian Studies, Cambridge 1978 (ed. J. M. Plumley): pp. 76-86. London: Arris and Philips.

Fattovich, R. and Piperno, M. 1981. Survey of the Gash Delta--November 1980. ?~Tame Akuma 19:26-30.

Fattovich, R. and Piperno, M. t982. Archaeological researches in the Gash Delta, Kassala Province (1980-1981 Field Seasons). Paper presented at the Vth International Conference on Nubian Studies, Heidelberg (Sept. 1982).

Geraads, D. In press. Faunal remains from archaeological sites in the Gash Delta, Sudan. Nyame Akuma 23.

Geus, F. 1976. Rapport annuel d'activitg 1975-1976. Khartoum: Sudan Antiquities Service, French Archaeological Research Unit.

Geus, F. 1981. Rapport annuel d'activitg 1979-1980. Khartoum: Directorate General for Antiquities and National Museums of the Sudan, French Archaeological Research Unit.

Haaland, R. 1981. Migratory herdsmen and cultivating women. The structure of Neolithic seasonal adaptation in the Khartoum Nile environment. Bergen.

Hays, T. R. 1971. The Sudanese Neolithic: a critical analysis. Ph.D. Thesis, Southern Methodist University, Dallas, Texas.

Hays, T. R. 1976. An examination of the Sudanese Neolithic. Proceedings, Vth Pan- African Congress of Prehistory and Quaternary Studies (eds B. Abebe, J. Chavaillon and J. E. G. Sutton): pp. 85-92. Addis Ababa: Provisional Military Government of Socialist Ethiopia, Ministry of Culture, Sport and Youth Affairs.

Hintze, F. 1960. Vorbericht fiber die Butana--Expedition 1958 des Instituts f'tir -~gyptologie der Humboldt-Universit~it zu Berlin. In Forschen und Wirken, Festschrift aus 150-Jahr-Feier der Humboldt-Universitiit zu Berlin, vol. 3:361-99. Berlin.

Kitchen, K. A. 1971. Punt and how to get there. Orientalia 40:187-207.

Marks, A. E., Ali, A. M., Hays, T. R. and Elamin, Y. 1980. Survey of the Northern Butana. Nyame Akuma 16:30-35.

Marks, A. E., Hays, T. R., Elamin, Y. and Ali, A. M. 1982. Butana archaeological project: interim note. Nyame Akuma 21:38-40.

Munzinger, W. 1890. Studi sull'Africa Orientale. Roma.

Otto, K. H. 1963. Shaqadud: a new Khartoum Neolithic site outside the Nile Valley. Kush 11:108-15.

Otto, K. H. 1964. Khartoum-Neolithikum am Jebel Shaqadud. In Varia Archaeologica. Wilhelm Unverzagt zum 70 Geburtstag dargebracht (ed. P. Grimm): pp. 9-14. Deutsche Akademie der Wissenschaften zu Berlin, Schriften der Sektion Ftir Vor- und Fr/ihgeschichte 16.

Phillipson, D. W. 1977. The laterprehistory of Eastern and Southern Africa. New York: Africana Publishing Company.

Posener, G. 1973. Le pays de Pount. Annuaire du Coll~ge de France 73:369-74.

Posener, G. 1977. L'or de Pount. In ]{gypten und Kush (eds E. Edensfelder, K. H. Priese, W. E. Reineke and S. Wenig): pp. 337-42. Berlin.

Privati, B. 1977-78. La poterie de la ville de Kerma. Genava 26:I29-34.

Privati, B. 1981-82. Nouveaux 616ments pour une classification de la c6ramique du Kerma Ancien. Genava 30:27-36.

Sadr, K. 1983. Prehistoric settlement


patterns--Khashm E1 Girba. Nyame Akuma 22:28--30.

Shiner, J. (ed.) 1971. The prehistory and geology of

Northern Sudan. Technical Report to the National Science Foundation, Grant GS 1192.

the archaeology of the eastern sahel, sudan: preliminary results

Documents