Archaeology and history: contribution to our understanding of medieval Nubia

Derek A. Welsby

@Pharaoh

Published in 

Egypt

 · 15 Dec 2023

Contents

• The Geographical and Chronological Limits
• From the Meroitic Empire to the Medieval Kingdoms
• Continuity or Change: History versus Archaeology
• Archaeological and Historical Chronologies
• Scientific Dating Techniques to the Rescue?
• Archaeology versus History, a Case Study
• The Suitability of Archaeological Data for our Understanding of the Past
• The Archaeology Study of the Nubian Kingdoms Today
• The Archaeological Study of the Medieval Nubian Kingdoms, the Future
• Acknowledgements
• References

This paper continues the survey of archaeology's contribution to the history of Nubia and concentrates on the "Christian" period.

The Geographical and Chronological Limits

The various definitions of Nubia are well known to the assembled audience. Although in the north the border of Nubia can be seen to coincide with the border of the northernmost of the three kingdoms which flourished in the "Christian" period, the southern limit of my survey must extend far to the south of what would usually be termed Nubia. It must extend to the southern border of the Kingdom of Alwa, but we are still very unsure where that actually lay. It was almost certainly over 300 km to the south of Khartoum as the presence of a church at JebelSaqadi Crawford and Addison, 1951 associated with a few sherds of typically Alwan pottery of types recovered from the site of the capital of the kingdom at Soba, indicates. It may, however, lay further south still. In the north, Nubia extends little beyond the banks of the Nile and it is rarely more than a few kilometres wide. In Southern Nubia, where the greater annual rainfall makes nomadic and even sedentary occupation away from the river possible, the eastern and western limits of the Nubian cultural zone, the limits of control particularly of the King residing at Soba, and the limits of the Nubia I wish to discuss here, are very imprecise. Hence it is not possible to define at present the geographical area covered by the term "Christian" Nubia.

The term «Christian Nubia» is itself far from ideal. It appears to conflate a religious ideology with a geographical and linguistic area and a political system. Nubian history is a continueum, and the worship of the god of the Christians by the people of that region is an episode not only in the history of Nubia but also in the history of the three Nubian kingdoms that are the real subject of this paper. It is clear from the Byzantine Church historians of the mid 6th century that a stable political system was in existence along the middle Nile with three autonomous kingdoms, before the official adoption of Christianity by their rulers. There are indications of hostility between them, the detour of Longinus into the Eastern Desert on his journey from Nobatia to Alwa may reflect political rather than religious animosity. Also the activities of Longinus in Alwa were said to be the result of a specific request by the Alwan King to the King of Nobatia, suggesting friendly contacts between the north and south of Nubia. One can appreciate why the King of Makuria would have been suspicious of the friendliness of his northern and southern neighbours, both potential enemies.

Well before the final demise of Christianity on the Middle Nile the designation, at least of the Kingdom of Makuria, as a Christian kingdom is not apposite. Bill Adams prefers to call the period of the breakdown of the kingdom, the feudal period, neither Christian nor Islamic. As early the first or second decade of the 14th century a muslim ruler sat on the throne at Old Dongola Adams, 1977: 528.

From the Meroitic Empire to the Medieval Kingdoms

The reasons for the fall of Meroe, which is traditionally dated to the mid 4th century AD, and the emergence of the Nubian kingdoms in its aftermath cannot at present be provided by archaeology (now see Török, 1988, pp. 209 suiv.), but that is not to say that archaeology does not contribute to our understanding of the history of this most difficult period. The excavations by Emery and Kirwan at Ballana and Qustol 1938) indicated clearly that a new centre of political power was in existence in Lower Nubia, in an area that had hitherto rarely enjoyed autonomy, an area that had never been the power base of a regime, being always ruled either from Egypt to the north or by the Kushites from Napata or Meroe far to the south.

Archaeological research uncovered the Ballana and Qustol tumuli with their exceptionally rich grave goods. It provided evidence for the great temporal power of the "primary" occupants, people whose prestige after death allowed them to consign lesser mortals to their fate to accompany them into the next world. These burials must surely be amongst the least ambiguous of burial types containing the mortal remains of kings and queens with crowns still in place on their heads. This has enabled us to state the fact of the presence of a major political power in the area. It does not help to indicate how that political power came to be consolidated and why it developed in that particular area.

The recovery of so many articles of Byzantine and Egyptian manufacture does indicate that the rulers who were buried at Ballana and Qustol enjoyed the support of the Empire to the north, a support perhaps both financial and moral (cf. Kirwan, 1987: 124­127) and this may go some way to explain the apparent paradox between the desire to retain "the kingly traditions of Ancient Kush while abandoning or suppressing everything associated with priestly traditions" Adams, 1977: 415. Yet the rulers of Lower Nubia who were buried at Ballana and Qustol where not at that time converted to Christianity, but were still treated with great tolerance by the Byzantines. The Nubians were allowed access to the Temple of Isis at Philae, the temple uniquely being exempt from the strict application of the Edict of Theodosius I (AD 390) which ordered the closure of all pagan temples throughout the Empire. It was only in the time of Justinian, on the eve of the official introduction of Christianity into the region, that the temple was finally closed. That the closure appears to have met with no protest, in stark contrast to the state of affairs in the 5^th century, presumably indicates that the cult was already in a state of terminal decline.

Reisner, the first archaeologist to formulate a cultural sequence for Nubia based on the archaeological material, a sequence which still holds good today, was forced to name the cultural assemblage of the immediately post-Meroitic period, the X-Group, highlighting the "prehistoric" nature of our evidence for the culture even though it lies in an area which had been included in historical writings since the time of Herodotus, and millenia before by the Ancient Egyptians.

Recent work by our French colleagues at el­Hobagi, following up a small scale excavation by the late Neville Chittick, appears to have found the southern Nubian equivalent of the Ballana and Qustolnecropoli. As in the north the associated habitation settlement has not been located, but at least we have a glimpse of the rulers of what had been the heartlands of the Meroitic Empire. The siting of the necropoli at Ballana/Qustol and el-Hobagi some distance from the earlier centres of political power, in the north at QasrIbrim where the Meroiticpestos or governor resided and in the south at Meroe itself, indicates that the new rulers of Nubia, of which there may well have been more than two, were not related to the old political order and did not need the status which would have accrued by occupying the old centres of power.

Other major necropoli, perhaps the burial grounds of more or less independent chiefdoms have been located at Genial and Firka (see T örök, 1988, p. 71) while burials, perhaps of members of the ruling hierarchy, but of lesser status, have been noted at Tanqasi and Tabo. At Tanqasi and Tabo large tumuli were found, but the poor grave goods indicated that the occupants of the tombs were not in the same class as the kings of the post-Meroitic period. The grave goods do suggest the possibility that Tanqasi was associated with the southern Nubian cultural zone, Tabo with the northern.

Although the post-Meroitic royal necropoli date to the period between the end of the Meroitic Empire and the first mention of the Nubian kingdoms in the historical sources, how they relate to the three kingdoms of Nobatia, Makuria and Alwa, first attested in the mid 6th century, is just as obscure as how they relate to the collapsing Meroitic Empire of the mid fourth century. The three Nubian Kingdoms have in the past largely been recognised by the evidence for their Christianity, and it is this evidence for Christianity that is generally taken as the characteristic feature of the Nubian kingdoms' archaeological signature. This approach will tend to obscure the evidence for the kingdoms before the arrival of Christianity.

The apparent lack of continuity between post- Meroitic Nubia and Nubia under the Kings of Nobatia, Makuria and Alwa may well be an artificial creation. The obsession of archaeologists with assigning human activity over time to distinct periods does not allow us to recognise the often extremely gradual process of the change, but only serves to highlight the differences between the situation before and after that change.

Continuity or Change: History versus Archaeology

We have here a classic problem in archaeology and archaeology's contribution to history. To what extent can political changes, the changes with which the historian and our historical sources are often most interested, be expected to be reflected in the archaeological record. These changes are much more likely to directly influence the lives of a relatively small and elite group in society, most notably the elite of the eclipsed society and the new elite of the emerging political group. Archaeological data is, in the normal course of events, statistically biased towards the recovery of the mundane rather than the exceptional and necessarily rare structures and artefacts directly associated with the upper echelons of society actively engaged in the new political system.

To some extent the nature of the archaeological research in Nubia, when it has been able to follow its own course, rather than being directed by the special rescue forces generated by the Aswan dams, has tended to concentrate on the special, the impressive and grandiose monuments of Nubia at all periods. Excavation of major Egyptian monuments, of Napatan and Meroitic pyramids, of the royal city of Meroe and of major complexes at Musawwarates Sufra, Naga and Wad benNaga, has been followed by excavation in the churches and cathedral at Faras, the churches at Old Dongola and the churches and palatial structures at Soba.

Not only are these sites of major political and/or religious importance in their heyday, but they are also the most likely sites to provide us with information from which to write political history. The reconstruction of the list of rulers of Napata and Meroe is a classic example of the use of data derived from archaeology, together with the archaeological technique familiar to the pre-historian, of seriation, being used to form an historical framework. However, we have not been so fortunate in the Medieval period in being able to produce a similar tour de force.

Christianity like most other major religions, both ancient and modern, holds out the promise of life after death, but it contrasts markedly with ancient Egyptian religion, the religion of the Meroites, and the religion of the kings buried at Ballana and Qustol, in that it is assumed that, at least in theory, everyone will be of one status in the afterlife and that the physical trappings of ones wordly wealth and status will not be necessary. There was thus no need to bury ones possessions with the dead nor to provide provisions for the journey to the other world, nor to boast of ones prowess in this life. Christian burials hence generally contain little or nothing in the way of artefacts (cf. Zurawski, 1986: 414) and none of the boastful inscriptions like those which graced earlier tombs. Also in Nubia they would appear not to have been housed in grandiose funerary monuments, the pyramid of the Meroites and the massive tumuli of the post-Meroitic period have no successors for the rulers of Nubia who adhered to the Christian faith. There are some special tomb monuments, as at Faras where the tomb of the bishop Joannes was a cuboid structure capped by a small dome, and at Soba where a circular red brick structure 4 m in diameter has been found, but these are very much the exceptions and they do not compare favourably with the grandiose monuments of earlier periods. At Faras the tomb just mentioned housed an ecclesiastical dignitary and the same may be the case at Soba. Of the rulers of Medieval Nubia, however, no burial have been found suggesting that their graves were not given special prominence. The recently discovered tombstone of King David of Alwa Jakobielski, 1991: 274-276, who ruled from AD 999 to 1015, that is in the heyday of the southern Nubian kingdom, gives no hint that he was the ruler of the rich and powerful kingdom as recorded by the Arab writers.

Archaeological and Historical Chronologies

Archaeological data and history are often on parallel tracks and it is the establishment of links between the two disciplines that is one of the major tasks of the archaeologist in the historical period. The greatest problem facing the archaeologist of the Nubian kingdoms in this task is the absence of an absolute chronology.

Very few artefacts to be found in Nubia at this period are intrinsically dated, and those that are, are rare and imported. In other advanced cultures coinage is often the most abundant and most useful category of material in this respect. Even when the coins are not dated to a specific year they can often be tied down to a short date range of several years, although this is not invariably the case (note the numismatic problems facing archaeologists of Axum Munro­Hay, 1989: 330). The absence of any native coinage places the archaeologist of Nubia at a very serious disadvantage. Finds of coins in the area are so unusual as to be of little value. In other parts of the world where coins are found it is their abundance which makes them a useful chronological indicator. Within the Kingdom of Alwa not a single coin has been recovered from an archaeological context.

Christian tombstones do often give the date of death which, when they are of major political or religious figures is of use in our quest to write history, they contain little else of interest. There is not usually any indication of where the deceased lived, what was his or her occupation, or even how old they were. If found within a church they do provide a terminus ante quem for the building of the structure of which they are a part if they are not being reused, but nothing more.

Building inscriptions are exceedingly rare. There are the two well known examples at Faras Jakobielski, 1972: 37, built into the wall of the palatial structure close to the cathedral. They record the foundation of a church in the eleventh year of the reign of Merkurios, when Markos was eparch, in the year 423 of the era of Diocletian, by the bishop of Pachoras, Paulos.

Although they are explicit in recording the construction of "this holy place was established as worthy for service in it of the Catholic and Apostolic Church " they were not built into the wall of the cathedral. Initially it was assumed that they related to the construction of the cathedral, but owing to their position in an adjacent building this interpretation is open to question. However, other evidence now conclusively shows that the cathedral excavated by the Polish Mission in the 1960's was that built by Paulos (pers. comm. Dr. W. Godlewski), and is presumably the structure referred to in the inscriptions. Why they were not built into the fabric of the cathedral is unclear.

Also from Faras is the foundation inscription from the Church on the South Slope of the Kom, carved on a door lintel in Coptic (ibid., p. 111). It gives the name of the reigning king and the years of his rule, the names of his parents and the name of the bishop who was responsible for the construction, along with the date of the construction, 646 in the era of Diocletian, i.e. AD 930.Another dates from well towards the end of the Christian period when in AD 1317 an inscription was set up in what may have been the throne hall (audience room) of the Kings of Dongola when it was converted into a mosque.

The rarity of building inscriptions largely denies us one of the most reliable sources of dating of major monuments. It also denies us any clear connection between historically attested and dated personages, the structures they built, or were built in their reigns, and the artefacts used in those buildings, and of equal importance the artefacts sealed beneath their floors.

The only intrinsically dated artefact from the recent excavations at Soba, apart from the tombstone already mentioned, was a glass jeton or weight of the Fatimid caliph Al-Zahir, dateable to the period AD 1020-35 Allason Jones, 1991: 143, which was probably over 200 years old when it was buried among the rubbish material used to fill the ground floor rooms of the mud brick palatial structure excavated in 1986. Its value, from a chronological point of view is thus severely limited.

As Adams pointed out in his summing up of the section on "Nubian Christianity: North and South", in the proceedings of the Uppsala conference Adams, 1987, it is the quest for an absolute chronology which so differentiates the historian's approach to a society to that of the pre-historian or the art-historian. Without a fixed and detailed chronology it is impossible to allocate events into their correct sequence. The problem can occur at all levels of the archaeological process. Even within a single building, if there is no direct stratigraphical relationship between two adjacent rooms for instance, it is frequently impossible to establish the contemporeinity or otherwise of activities, the evidence for which is found in the two discrete areas. Where two parts of the same site are excavated, as at Soba where the recent excavations on mound B lie over 1 km from the church on mound C excavated by Peter Shinnie, direct stratigraphical links are extremely unlikely and contemporeinity of activities found in the archaeological record cannot be proven. On the next level up attempts to establish the relationship of events at different sites within the same kingdom or across political boundaries is an impossible task in most cases. These are archaeological problems faced by archaeologists of all periods and cultures, the difficulties of dovetailing archaeological data into an historical framework is of identical type.

There are exceptions where major events can affect in the same way a number of different localities and allows us to glimpse the true chronological relationship between different areas. The fire which destroyed the western end of the palatial structure on mound B at Soba allows us to recognise the same moment of disaster over the whole 400 sq. m. of the structure damaged at that time. One can envisage scenarios where events will leave evidence in the archaeological record over wide areas of a site, e.g. major destruction deposits caused by serious flooding, as were found at Meinarti Adams, 1965: 149, and as may be found by future archaeologists excavating in the Dongola reach in levels associated with the great flood of 1988,or destruction by fire whether started by accident or by localised enemy activities, and on a larger scale by the evidence for invasion. In the present state of our knowledge large scale activities are not recognised archaeologically as having left their mark on the Nubian kingdoms although we may expect to find evidence of the Arab invasions which penetrated into Nubia as far south as Old Dongola in AD 642 and 651-2. However characterising "destruction by the enemy" is itself no easy matter.

No other single events are recorded in the historical sources which may have left an unambiguous signature in the archaeological record. There were other military incursions into Nubia, but the scale of these will make their identification difficult. The Nubian kingdoms also do not appear to have fallen to the sword in one cataclismic event, but to have been gradually infiltrated over a considerable period of time from the north, south and east so that it may even have been difficult for a contemporary observer to have been sure when the Nubian kingdoms ceased to exist. By all accounts they gradually fell apart, central authority breaking down allowing the rise of petty chiefs who held sway over restricted areas. The break-up of the Nubian kingdoms in the period 12th-16th centuries may serve as a model for the collapse of the Meroitic state 1000 years before.

Scientific Dating Techniques to the Rescue?

Scientific dating techniques at first sight appear to provide a solution to some of the problems of bridging the gap between archaeological and historical data, but in reality, at present, they leave much to be desired. To the pre-historian, who is only able to appreciate general trends and where the changes taking place over centuries if not millennia are of interest, the imprecise dating techniques such as radiocarbon dating are of value. In the historical framework of the Nubian kingdoms the degree of precision is rarely sufficient to make the technique useful. Archaeomagnetism is a possible source of data, particularly useful in providing dates for the last firing of a pottery kiln, of a hearth, or to date the destruction by fire of a mud brick building where the temperatures attained were sufficient to bake the green bricks red. However, the data derived from the measurement of the fired structures has to be related to a data-base derived in part from the association of material already dated by historical evidence to which the archaeomagnetic data can be compared. The necessary data base has not yet been prepared for the Sudan and the risk of being involved in circular arguments is ever present.

Thermoluminescence is potentially of great value, allowing the dating of the manufacture of pottery vessels. Pottery as the most common artefact found on sites in the period of the Nubian kingdoms, has many advantages as a means of attaining an understanding of the chronology of features and deposits uncovered by archaeology. Even in the present state of pottery research, where we are without a tight chronology for the periods of manufacture of the pottery types, not withstanding the detailed researches by Bill Adams and others, pottery bears the onus of providing us with our site chronologies to a large degree.

Of equal value is the ability to date fired brick by thermoluminescence. This may, in the future, allow the construction of particular buildings to be closely dated. In many cases, as at Soba, it is clear that the bricks were specifically made for use in the construction of a building. Buildings A, B and C beneath the western end of mound B were each constructed of bricks of a particular type. Building A was constructed from rectangular bricks 44-48 x 18-27 cm in size, building B from bricks 28 cm square and 14x28 cm, and building C from bricks with dimen­sions of 32 x 18 cm. However, the degree of precision attainable at present is usually only ±7% of the age of the sample, i.e. for a red brick building constructed in AD 650 the date range from TL would bracket the period AD 556-744 and cannot be expected under ideal circumstances to be more accurate than to within ±5% of the age (data from the Durham T.L. Dating Service, England). One must await the refining of the technique.

The other scientific technique which holds out hope for the future is dendrochronology, but again the basic data to allow samples of timber to be placed in their correct relative position in time is not yet available. Palm, the most widely used timber for building purposes at all periods of Nubian history is not, on account of its structure, suitable for dendrochronological dating. Also the very dry climate throughout much of the region, coupled with the relative scarcity of hardwoods, will have resulted in timber surviving for long periods and being reused whenever the opportunity permitted.

Archaeology versus History, a Case Study

It is salutary to look at the recent excavations at Soba East, with which the writer is most familiar, as a case study to highlight the problems of using archaeological data to write history and conversely, the contribution of historical evidence to the interpretation of the archaeological data. The Kingdom of Alwa effectively falls within what has been defined as proto-history, so infrequently do the historical sources refer to the Kingdom. We have extremely few dates provided by the historical sources. AD 580, the arrival of the missionary Longinus is the first, the second, and incidentally the last, is the fall of Soba prior to AD 1504 at the hands of the Funj and the Abdallab Arabs, recorded in the Funj Chronicle. Between these two dated events there are references to Alwa which show it to have been a large and flourishing state in the 9th and 10th centuries. There are indications of close contacts with the Makurian royal family to the north, for the presence of Muslim traders in the capital and information about the topography of the city of Soba and its vicinity, about its agricultural produce and about its customs, i.e. that they were the same as those of the Makurians. What, however, is the value of this information.

The date of AD 580 has been taken as the date of the introduction of Christianity into the Kingdom but Longinus met Christians at the capital city of Alwa who had come from Axum. AD 580 cannot be confidently used as a terminus post quemfor any Christian artefacts or buildings in the kingdom, just as the dates given by the same ecclesiastical historians for the arrival of Christianity in the northern kingdoms were probably not the dates of the earliest introduction of Christianity into those areas.

The excavations at Meroe by John Garstang in the 1909-1910 season found some evidence for Christianity in the Temple of Isis and it has been suggested that the temple may have been reused as a church Kirwan, 1989: 300. A similar reuse of a Meroitic temple as a church in the area which may be pre­sumed to have lain under the control of the Kings of Alwa has been noted at Musawwarates-Sufra where pottery of identical type to that found in the Christian levels at Soba, was recovered from Temple IILA Török, 1974, Abb. 10. The discovery of Christian graves, one with a pendant in the form of a cross, a little to the east Hintze, 1962: 195-196, gives further support to this suggestion. One might suggest that such reuse of temples as places of Christian worship is more likely to be an early phenomenon and there is certainly little evidence for the continuing importance of Meroe and Musawwarat into the Christian period. Whether the temple-churches date to before AD 580, however, cannot be ascertained and it should be noted that the pottery types from Temple IRA are found at Soba in a number of contexts dateable to throughout the Christian period, but are rarely associated with the early Christian Soba ware.

Apart from the initial impetus to the birth of Christianity in Alwa given by Longinus' visit, we have no information on the mechanisms for the spread of the religion from the king and his entourage to the mass of the population. This may have been a slow process and the bulk of the population of Alwa may have remained pagan well into the "Christian" period. The relationship of the Christian burials found in post-Meroitic cemeteries, the one type differentiated from the other by the burial customs, is usually taken as sequential, the Christian burials being assumed to be later than the post-Meroitic ones. This, however, is rather a simplistic view and it should not automatically be assumed that the differences in burial practice, at least in the earlier phases of the growth of Christianity in Alwa, reflect a chronological progression. Within the Kingdom of Alwa, the evidence for the post-Meroitic and the Christian periods is mutually exclusive. The post-Meroitic inhabitants of the area are only recognisable by their graves, the Christian inhabitants largely by major religious monuments and habitation sites.

In time the influence of Christianity percolated down through society and into many facets of life. At Soba the earliest manifestation of Christian influence, to be seen in quantity in the archaeological record is the presence of Soba ware, the highly distinctive and often profusely decorated fine pottery whose decoration is derived from the motifs to be found in the mural art used to decorate Christian churches as exemplified by the cathedral at Faras. Prior to the production of Soba ware there is no reason to expect Christianity to be visible in the archaeological record away from the church and the cemetery. Christian burials can be recognised, but the absence of grave goods, one of their most characteristic features, divorces them from the artefacts which we so desperately need to date.

During the hey-day of the Kingdom of Alwa the historical sources are largely silent. They do speak of the power, wealth and prosperity of the kingdom and here, where general aspects of the culture rather than specific events are being discussed, archaeology and history are compatible, although the evidence is not always taken at face value.

The case of Soba is a classic example of this. IbnSelim el Aswani, who may have visited Soba towards the end of the 10^th century, wrote of the, "fine buildings and large monasteries, churches rich with gold and gardens" (in Vantini, 1975: 613). This is a clear statement by a well respected historical source. Archaeologists have, however, in the past, observing the sad ruins of Soba - uninspiring mounds of red and mud brick broken only by the granite columns of a church close by the river - been rather sceptical of IbnSelim's observations. The results of the excavations by Peter Shinnie at Soba from 1950-1952 Shinnie, 1961 were not interpreted as offering support to IbnSelim, even though Shinnie found what he took to be the remains of a palatial structure and recovered a quantity of fine Islamic glass, high value imports.

The recent excavations by the British Institute in Eastern Africa, which concentrated on an area at the western end of mound B, circa 80 m from the excavations of Shinnie, were located so as to allow the examination of a red brick spread, a category of surface remains defined as a result of the BIEA's first season of survey and trial trenching in the winter of 1981-1982 directed by Mr. C. M. Daniels. The discovery of three churches, two of cathedral size by Nubian standards Welsby and Daniels, 1991, was unexpected but immediately forced a change of attitude to the status of Soba, and of the Kingdom of Alwa, among the medieval Nubian kingdoms. It cast serious doubts on recent interpretations both of the poverty of Soba and of the veracity of IbnSelim and other Arab writers who refer to the southernmost of the Nubian kingdoms. The palatial structure located immediately to the east of the churches added further confirmation of the presence of substantial "public" buildings at Soba. In hindsight the building partly excavated by Shinnie on mound B can be accepted as probably another substantial, palatial building, as Shinnie himself had suggested.

The complex of structures on mound B, with at least seven red brick structures (churches?) and mud brick buildings covering an area of 3 hectares in the centre of the city, suggests the presence of a monumental quarter of massive proportions. The excavations of Somers Clarke Clarke, 1912: 34-38 and Shinnie Shinnie, 1961: 25-27 indicate that there was another cathedralsized church on mound C towards the river. At this stage in our knowledge of the Soba site we already have evidence for the presence of three of the largest churches known in Nubia and the poten­tial for the location of others is considerable.

In the later phases of the medieval kingdoms the historical sources hint at the growth of an independent state, the Kingdom of A1 Abwab, on the northern frontier of Alwa, presumably between it and the Kingdom of Makuria. It is assumed that this state had broken away from Alwa and that its presence suggests that the power of the rulers of Alwa was on the wane. However, we have no idea of the political organisation of Alwa, whether the state was a loose confederation of petty rulers under the suzerainity of the king of Alwa, or whether the king ruled directly through his appointed officials. The archaeology of Alwa is almost totally confined to the archaeology of Soba, the capital of the kingdom at least from the 9th century when A1 Yaqubi records that it was the residence of the Alwan kings (in Vantini, 1975: 71). We cannot thus tell whether there were distinctive features of the Alwan cultural assemblage that are wanting in the Kingdom of A1 Abwab. In any event if there were differences it is extremely unlikely that these could be related directly to the political allegience of the area.

The fall of the Kingdom of Soba, presumably the same as the Kingdom of Alwa, is recorded in the Funj Chronicle, a source only comitted to writing probably late in the 19^th century (in Vantini, 1975: 786-787). The reliability of the information in the Chronicle is thus suspect. It seems clear from the archaeological evidence that the churches on mound B were abandoned well before the conquest of the city at some date prior to 1504. This may suggest that, before the fall of Soba to the Abdallab Arabs and the Funj, it was already a Muslim city. The Chronicle states that the city was turned into a Funj metropolis. This is at variance both with the archaeological data and with the statement of David Reubeni who visited the site in 1523 and noted "the town of Soba, which is in ruins; there they have wooden dwellings" (in Vantini, 1975: 751). The distinctive Funj pottery is an extremely rare site find at Soba and this casts doubt on the authenticity of the Funj Chronicle.

Conversely there is a hint in the archaeological record of a major event which is not recorded by the historical sources. The southern church was the latest of the three to be constructed at the western end of mound B and it included the characteristic range of five rooms at the east of the building. Whatever the function of the rooms in the eastern range, their number was reduced to three at the end of phase la in the northern church and during phase I in the southern church. The presence of the range of five in the southern church, therefore, strongly suggests that it was constructed during phase la in the northern church.

Although building C was clearly designed and built as a church, and we have every reason to suspect that it was completed and presumably functioned for a time as a place of Christian worship, it was abandoned within the Christian period, at least partly demolished, and was given over to intensive domestic activities. Evidence for a similar hiatus in the use of the northern church was also found. The extensive post-Christian robbing of the central church may have removed similar evidence from that building.

It is extremely difficult to suggest why, during the currency of the Christian state, churches would be deconsecrated and demolished. The implication is that Soba ceased to be Christian at some date in the medieval period. Was there a period during which the Christian kings were ousted from power? This is, of course, pure speculation, but some major event must have been responsible for the features observed in the archaeological record. The rebuilding of both the northern and southern churches to a similar plan indicates that whatever the problem had been it was relatively short lived, the Church and presumably the Christian state recovering and surviving for centuries thereafter.

Here history is desperately needed to come to the aid of archaeology. However, in the nature of things, the possibility of new historical data emerging is remote whereas, by contrast, new archaeological data, often it must be admitted not of great worth, is brought to light with every stroke of the trowel.

The Suitability of Archaeological Data for our Understanding of the Past

The discussion above has illustrated some of the difficulties facing the archaeologist and the historian when an attempt is made to relate the data from the two disciplines. This is not necessarily a deficiency in the science of archaeology, but is frequently a failing of the scholar who asks the wrong questions from his data. Although archaeology rarely provides information suitable for the reconstruction of political history it is an invaluable source of information on other branches of historical research, social and economic history, art history and the history of architecture among them.

The recovery of artefacts containing the written word is the most direct means by which archaeology contributes to historical research. Owing to the especially favourable climatic conditions, particularly in the far north of Nubia, papyri and parchment can survive intact and the spectacular finds of such material from QasrIbrim, itself a major political centre in the medieval period, holds out the hope for the recovery of further significant historical documents which may make a fundamental contribution to our understanding of Nubian history. QasrIbrim is now a unique site in Northern Nubia, having survived relatively unscathed the flooding of the Nile Valley behind the Sadd el Aalli. Further to the south the climatic conditions steadily become less favourable to the preservation of organic materials and the threat from the hungry termite is ever present.

The Archaeology Study of the Nubian Kingdoms Today

At present a number of archaeological missions are actively involved in the study of Medieval Nubia. The long term excavations by the Polish Mission at Old Dongola, a project now of 26 years duration, seems set to continue and has contributed a considerable amount to our understanding, particularly of Nubian ecclesiastical architecture. The recent work in the industrial quarter at Old Dongola is particularly welcome and the results of the pottery kiln excavations should go some way to allowing us to begin a detailed study of the pottery of Central Nubia. Only a few kilometres to the north of Old Dongola the Canadian Mission, after two seasons of survey over a 80 km stretch of the east bank of the Nile in the Dongola Reach from 1984-1986 Grzymski, 1987, is excavat­ing a site of uncertain character at Hambukol which should give us some idea of the character of the capital city and of its immediate environs.

Moreover the Canadian Project is designed "to study in depth the cultural development of a stratified habitation site" Grzymski, 1988: 2 in an area where there is evidence for Meroitic and earlier occupation as well as the obvious Christian monuments in the vicinity.

A campaign of three seasons of excavation at Soba is also in progress during which an attempt is being made to excavate a number of areas of differing character and date within the city to provide a more balanced appraisal of the development, layout and history of the site. Of particular interest is the evidence, recovered during the initial surveying and trial trenching of the site from 1981-1983, for the city having attained its greatest extent in the early Medieval period, at a time when the highly distinctive polychromatic painted pottery known as Soba ware was in use. In the present state of our knowledge this occupations appears to be associated with timber structures of post-hole construction. The nature of the structures, which were found in profusion under the churches at the western end of mound B, remains unclear. They may, however, reflect a tradition of timber architecture in marked contrast to the more familiar stone, red and mud brick buildings of the Meroitic and Medieval phases of occupation in Central Sudan. If the fine quality of Soba ware reflects its high cost of purchase in the market place, its presence may indicate that the timber buildings were not the dwellings of the poorest strata in Alwan society.

The work of the French Mission at el Hobagi has already been mentioned. Given the extent of the Nubian kingdoms and the paucity of archaeological research within them, particularly in Central and Southern Nubia, the gaps in our knowledge are immense.

What are the outstanding problems facing the student of Medieval Nubia on which light can be shed by archaeological research? Obviously there are many and the relative importance one would assign to them depends largely on the personal interest of the scholar concerned. In Northern Nubia much of Nobatia had already disappeared beneath the waters of Lake N asser apart from the QasrIbrim, and archaeological research is confined to post-excavation studies. In Central and Southern Nubia it is the disparity between the number of sites which we may assume 'With confidence to have been occupied in the Medieval period and the very small number of sites, which have actually been examined archaeologically, which makes any conclusions that we draw from the data available extremely tentative. The data base at present is very small and the relation between the data from different sites, often a considerable distance apart, is problematical.

The Archaeological Study of the Medieval Nubian Kingdoms, the Future

What then of the future direction of research into the archaeology of Medieval Nubia. One feature stands out and demands attention. This is the absolute necessity to publish the results of archaeological research, quickly and in detail. Interim reports are of interest, but are not in any way a substitute for the final report where all the evidence from the site is aired at length.

Publication is increasingly a problem. It is very expensive and funding bodies, which may be attracted by the "glamour" of excavation, are often reluctant to make grants to the less spectacular post-excavation projects which are of considerably longer duration than the field projects with which they are concerned. There is also, it would seem, all too often the fear of putting pen to paper and producing the "definitive" report while the work is still in progress. However, With major sites, where work may be expected to continue for decades, the final reports cannot be delayed indefinitely. Any archaeological report where less than the "whole" site has been excavated, is only an interim statement, but if the work is well done and the material is published in detail, although the conclusions and interpretations may be modified by subsequent work, the value of the publication will remain for ever.

There was some discussion, during the concluding session of the Uppsala conference, of the relative merits of rescue archaeology versus research excavations. Ideally of course both types of work are needed, but there is rarely enough money for both, this is the case in Britain and elsewhere just as much as in Sudan. Many of the research excavations in Sudan are also, to some extent, necessitated either by a direct and imminent threat or by the necessity to demonstrate the continuing need for the preservation of the sites already under the control of the Department of Antiquities. However, it is rather difficult to justify pure research excavations in the face of the widespread threat to archaeological sites along the banks of the Nile from the expansion of agriculture and from the encroachment of housing, and once again the potentially devastating destruction of another large tract of the Nile Valley, in the Fourth Cataract area looms before us. By concentrating on the sites well known to us, and which are protected to some extent, we risk the destruction for ever of sites, the existence of which we are not even aware. The contrast between our knowledge of the human geography of Northern Nubia at all periods and that of Central and Southern Nubia the former benefiting from the detailed surveys associated with the construction of the Aswan dams, serves to indicate just how much information there was and to some extent there must still be available.

Ideally, survey/rescue archaeology should go hand in hand with research excavations. The research excavations will then provide the dated sequences of artefacts which are so essential to allow us to recognise the significance of the sites located through survey work and revealed by excavations, the siting of which was decided by factors beyond the archaeologist's control.

Today the disparity in our knowledge of Medieval Nubia mirrors the thrust of archaeological research hence the heavy weighting towards the very north of Nubia. As has often been pointed out the discussion of Nubia, North and South is the discussion of the archaeological work in those areas rather than of the differences in settlement patterns in the Medieval period. If we fail to take up the challenge posed by the threat to the archaeological heritage in Central and Southern Nubia we risk the disparity in our knowledge of Nubia, North and South, at all periods, being with us and succeeding generations for ever.

Acknowledgements

The writer would like to thank L. Allason-Jones, D. Edwards and J. Maddox for reading earlier versions of this text and making a number of helpful comment. However, the views expressed are those of the writer alone.

References

• Adams W.Y. 1965, 'Sudan Antiquities Service Excavations at Meinarti, 1963-1964', in: Kush XIII, pp. 148-176.
• Adams W.Y. 1977, Nubia: Corridor to Africa, Princeton.
• Adams W.Y1987, 'Three perspectives on the Past: The Historian, the Art Historian, and the Prehistorian' (Comments on Session II), in: Nubian Culture, pp. 285-291.
• Allason-Jones L.1991, 'Small objects from the western end of mound B,' in: Welsby and Daniels (eds.), Archaeological Research at a Medieval Capital on the Blue Nile (BIEA), Monograph series, N° 12, London: 126-157.
• Clarke S. 1912, Christian Antiquities in the Nile Valley, Oxford.
• Crawford O.G.S. - Addison F. 1951, Abu Geili and Saqadi and Dar el Mek, London.
• Emery W.B. - Kirwan L.P. 1938, The Royal Tombs at Ballana and Qustol, Cairo.
• Grzymski K. 1987, Archaeological Reconnaissance in Upper Nubia, SSEA Publications, XIV, Toronto.
• Grzymski K. 1988, 'Canadian Expedition to Nubia', in: Bulletin VIII (2), pp. 1-5.
• Hintze F. 1962, Preliminary Report on the Excavations at Musawwarat, in: Kush X: 170-202.
• Jakobielski S. 1972, Faras IIL A History of the Bishophric of Pachoras, War­szawa.
• Jakobielski S. 1991, 'The Inscriptions, Ostraca and Graffiti', in: Welsby and Daniels(eds.), Archaeological Research at a Medieval Capital on the Blue Nile (BIEA), Monograph series, N° 12, London: 274-296.
• Kirwan L.P. 1987, 'The Birth of Christian Nubia: Some Archaeological Problems', in: Revista degli Studi Orientali, LVIII, fasc. I­IV, 1984.
• Kirwan L.P. 1989, Meroe, Soba and the Kingdom of Alwa, in: Meroitica 10: 299-304.
• Munro-Hay S. 1989, Excavations at Axum, British Institute in Eastern Africa (BIEA), Monograph series N° 10, London.
• Shinnie P.L. 1961, Excavations at Soba, in: SASOP 3, Khartoum.
• Török L. 1974, 'Ein Christianisiertes Tempelgebaude in Musawwarat es Sufra (Sudan)', in: Acta Archaeologica Academiae Scienta­rum Hungaricae 26: 71-103.
• Vantini G. 1975, Oriental Sources concerning Nubia, Heidelberg and War­saw.
• Welsby D.A. - Daniels C.M. 1991 (eds.), SOBA: Archaeological Research at a Medieval Capital on the Blue Nile (BIEA), Monograph series, N° 12, London.
• Zurawski B. 1986, 'Bishop's Tombs in Faras', in: Nubische Studien: 413-418