1Asante, lying in what was once dense forest in the centre of modern Ghana, is one of the best known of indigenous African states. With a rich and colorful culture and an elaborate and sophisticated political system, it has been much studied and several major books have surveyed its history and culture (Fynn 1971 ; Wilks 1989 ; McCaskie 1995). From the historically attested beginning of the Asante State at the time of Osei Tutu (A.D. 1700), the sequence of events is reasonably certain and details are known of political and economic exchanges with European powers on the coast of Ghana. Europeans were in touch with Asante certainly from the early years of Osei Tutu, when we know that the Dutch envoy David Nyendael paid a visit to Kumase. Unfortunately he died very soon after his return to Elmina and apparently left little or no record of this visit.
2There is little information on other early European visitors, though it is likely that there had been occasional visits during the eighteenth century, probably mainly from the Dutch-held fort at Elmina. A Dutch envoy, W. Huydecoper, had been in Kumase from May 1816 to April 1817, just before the arrival of T. Bowdich who – as part of a British embassy – was in Kumase in 1817 and wrote an extremely valuable description (Bowdich 1819), the first to be published. Dupuis, another British emissary, followed him in 1820. From then on a number of Europeans concerned with trade or missionary activities came to Kumase. One was the Dutch envoy Simons, who came to Kumase in 1832 and was the first European to mention the sacred site of Asantemanso. The British Wesleyan missionary T. B. Freeman was in Kumase in 1841 and mentions other visitors, including the Dutch agent, J. Huydecoper, a son of W. Huydecoper. J. Huydecoper arrived in Kumase in 1837 and stayed until 1842, when Pell, whose arrival was noted by Freeman (1844 :175), succeeded him. The British Governor of the Gold Coast, Winniet, visited in 1848. The Basel missionaries Ramseyer and his wife and Kühne, captured by the Asantes at Anum on the river Volta, remained there for several years. They left Kumase in the wake of the British campaign of 1874, which did much damage to Kumase town, including the destruction of the Aban, a stone building erected by Asantehene Osei Bonsu in 1822. In the final years of Asante independence, the British invaded once again ; and in 1896 the Asantehene, Prempe I, was exiled, and a British administrator appointed. A fort, which still stands, was built by the British and was subjected to a siege in 1900, in the last military activity, the Yaa Asantewaa war, named after the queen mother of Edweso.
3The few European visitors, except Bowdich and Dupuis, do not contribute much to our understanding of Asante civilization, and were not much concerned with its beginnings. European interest was largely with trade, mostly slaves and gold, in exchange for European goods, of which guns were an important element and which made possible the victories of Asante over neighboring territories. The vital years when Asante was developing its control over what was to become the Asante federation are scarcely known, though local traditions tell us something of the shadowy personages of Oti Akenten and Obiri Yeboa, the first to purchase the land in Kwaman which became Kumase. These stories do not go back further than the early seventeenth century, and it is very difficult to know how, not only Asante, but also other Akan states of the forest, developed their distinctive culture and political systems.
4A considerable collection of local oral traditions has been made as part of a recent research project, by recording interviews with inhabitants of the villages where archaeological work was carried out. It is notable that all emphasize the period of Osei Tutu and the foundation of the state, or the period immediately preceding it. It is this period to which Asantes return time and again ; and the important and dramatic events of the end of the seventeenth century, including the story of the arrival of the Golden Stool and the role of the priest Okomfo Anokye, are those that concern and are described by the present inhabitants. The Golden Stool, a copy in gold, or wood overlaid with gold, of the wooden stools which are not only objects of domestic use but also symbols of chiefly authority, is regarded as containing the soul of the Asante nation, and is still an object of much veneration. The priest Okomfo Anokye is remembered in oral tradition. He was a contemporary of the first Asantehene, Osei Tutu, and an important personality in the events accompanying the foundation of the Asante State.
5Wilks, in two brilliant papers (1993a, 1993b), has made suggestions as to how the Asante state developed from a number of small hunter-gatherer communities who gradually mastered the special kind of agriculture necessary in forest conditions. His reconstruction is highly plausible, but the time he allows for these developments seems to be somewhat short, and his suggestion that agriculture developed very late in the West African forest is much at variance with what is known of agricultural and political changes in other parts of the world with similar ecological conditions. In these areas, the major shift from a hunter-gatherer to an agricultural economy happened much earlier than the approximately fifteenth century A.D. date that Wilks suggests for Asante. Vansina (1985 ; 1990), in discussing events in the forest of the northern Congo, has suggested a date many centuries earlier.
6Collections of oral traditions do not greatly help in understanding these developments, though Wilks has made very skillful use of them. It is clear that, as with most oral traditions from Africa, these stories do not usually go much farther back than the sixteenth or perhaps fifteenth centuries, and this may have influenced Wilks to date the key development of a shift to agriculture to that period. It is certainly this development which made possible larger and more stable residential settlements, which ultimately became towns ; and led to the various Akan states of which a group came together under Kumase leadership to form the Asante state or federation.
7It was to investigate the archaeological past, the only certain evidence for the development of settlements and the beginning of the Asante State, that the Asante archaeological project was started in 1986. The intention was to excavate a sample of settlements at which oral traditions suggested evidence for early occupation might be found. It was hoped to obtain an idea of the size and date of such settlements, to find examples of the material culture of the time, and to develop a sequence of artefacts that could be dated and thus used as dating evidence for other sites. It was realized from the beginning that it was improbable material other than pottery would be found, in view of the destructive nature of the local environment on all objects made of organic materials.
8The investigation aimed to establish a chronology for the beginning of permanent settlement in the area, and to estimate the size of early settlements so as to judge the probability that the inhabitants had passed beyond the stage of obtaining their livelihood from collecting and hunting for food resources. It can be assumed that the early Stone Age inhabitants of the forest – and they certainly existed, as the work of Oliver Davies (1967) has shown – were surviving by using this mode of subsistence.
9The key question of when the change to agriculture came about is much more difficult to establish unless fragments of cultivated crops can be found. It is also not normally possible to say whether the people who have left archaeological remains were ancestral to the present-day inhabitants. If, however, sites are found at places which have been continually inhabited from early times to the present day, and if the evidence of the pottery suggests a generally continuous tradition, then it is reasonable to suggest that the same people have lived at these sites since the earliest dates that can be obtained. Conversely, if a marked change is seen in the pottery, then an explanation must be sought, although often it cannot be found. The usual solution to the problem is to assume a change in the culture, and perhaps, what is unknowable, in language and 'ethnicity'. These are difficult problems which archaeology is not generally able to answer.
10To find archaeological sites likely to provide information relevant to the early days of Asante and its predecessors in an area of heavy vegetation is a difficult matter, and the normal methods of field investigation are unlikely to be successful. Although most of the dense forest which used to cover Asante and which gave so much trouble to British invaders is now gone, the vegetation is still thick enough to obstruct easy observation of ancient settlements. Because of this, information from oral traditions was used to identify sites relevant to the origins of Asante. Traditions say quite firmly that the Asante people came from a hole in the ground at the place known as Asantemanso, though some versions claim that it was only the Aduana abusua (matrilineage) that originated there. This tradition and the stories, which tell of it, have a number of fascinating details of how various animals came out of the hole. One version says :
It is believed that on one Nkyidwo (Monday night), ten members of the Aduana family emerged from the sacred ground at Asantemanso. They were followed by other members of the various Asante clans, a leopard, frog and a dog. The ten people were (ten names are given).
When Nananom Aduana got out of the ground, they found the area thick with forest and started to scatter.....Nana Adu gathered them together....and led them to build a hut for shelter.....That day was Wednesday.
Tragedy struck....a tree fell and killed Nana Adu Manu.....In commemoration of this tragedy, the occupants of the Numapau stool (the Essumejahene) never travel on Wednesday....The rest of the people built the hut and settled at present-day Essumeja.....The dog who accompanied them...produced fire, the frog provided water and Aberewa Musu miraculously planted Ode (bush yam) and got it harvested within a few hours for meals.
11The story is somewhat contradicted by another which places the beginnings of the people further south, in Adanse – the place where "God created the world", and where the art of building was first developed, as the name of Adanse suggests (Twi edan, house). It may be that at least some branches of the Akan first developed coherent form in Adanse, and that it was at Asantemanso that the Asante first became a consciously organized group, the ancestors of those who later moved north to Kumase (originally known as Asaman). The tradition shows at least three generations living at Asantemanso before Kwabia Amanfi, elder brother of Oti Akenten, found Asantemanso too restricted for his needs and sought other land, first at Kokofu, then at Kwaman (Kumase) (Wilks 1993b : 100-101).
12The site now famous for the “sacred forest”, in which the hole whence the ancestors came is located and still known, remains an important national shrine connected with many observances. Regular ceremonies are held to pour libations and offer prayers ; and in the grove in the forest, seven small pots represent the seven mmusua (plural of abusua) of which the Akan people are comprised. Many other stories are also associated with this place - it was, perhaps still is, a place of sanctuary, where those condemned to death could achieve safety, though they were subsequently not able to leave the settlement and had to live out their lives as servants of the chiefly family of Asantemanso. Amongst other traditions associated with the place is that by which the Asantehene and the Kokofuhene must cover their eyes when passing by the sacred forest.
13Asantemanso was visited and described in some detail by Rattray (1923 :123ff), and it was his description, together with the local tradition, that made it possible to identify the site. It was, however, only in the second season (1987) that the mounds described by Rattray were identified, somewhat concealed in a grove of cocoa trees. Archaeological excavations at the site, carried out for three seasons, have shown without doubt that there was early occupation there.
14Led by Rattray's description, the first season was devoted to a general survey of the area in and adjacent to the present settlement to try to establish the area in which the earlier town was situated. No areas of massive early settlement were found, but a few trial trenches produced potsherds somewhat different from modern ones, as well as material for radiocarbon dating.
15It was only when radiocarbon dates were available that it became clear there had been early occupation at the site, and that further work should be carried out. Two further seasons provided the evidence that there had been a settlement of considerable size at Asantemanso dating back to at least the ninth century A.D. The earliest radiocarbon dates are in B.C. range, but these seem very early ; it is perhaps more reasonable to see in those from the ninth to tenth centuries A.D. a more reasonable estimate for the beginning of occupation at this sacred place.
16Certainly there is good evidence of considerable human occupation from that date (Table 1). It is possible to show how the town shifted at different times in the past, though it was always in the same general area and close to the location of the sacred forest. The full extent of the town in any of the past centuries cannot be determined with certainty, but the evidence for human occupation is clear. From the excavation it can be shown that the site covered an area of approximately 1,300 square meters trending east west along a low ridge. The town went through several periods of growth and dispersal. A period of florescence was reached between AD 1200 and 1700, after which it declined in size and importance. The tradition says that the town had seventy-seven streets but this story is told of many early Akan towns, and should only be used to suggest that it was a place of some significance. Occupation has continued down to the present day ; and though the present tiny village of ten or twenty inhabitants is only a shadow of what it must have been in the past, there has certainly been continuous occupation. Wilks (1993a : 71-72) is doubtful of the importance of Asantemanso and suggests that it may only have been a ritual centre rather than a permanent urban one. The archaeological remains with large quantities of ceramics support the excavator's view that it was a sizable town.
17The excavation at Asantemanso showed human occupation for many more years than conventional views had suggested (Fig. 1). Further, the existence of a sizable settlement strongly suggested an agricultural food supply. It is improbable that a town of the size that can be assumed for the earliest known settlement could have survived on the basis of gathering wild forest produce, together with the collecting of snails and the hunting of animals, most of them the small creatures that we now know as the only wild inhabitants of the forest.
Figure 1 : Asantemanso site map.
18There is a difficulty in making firm statements about the existence or not of agriculture. It is probable that the main crops would have been roots - yams of various sorts predominating. The considerable range of vegetable foodstuffs now available would not have been present in the early years of occupation at Asantemanso, since so many of the crops now grown in abundance are from the New World. They would not have been available prior to, at earliest, the beginning of the sixteenth century, and probably about one hundred years later. Of these, maize, now a widely grown crop, and cassava were imports from the Americas, along with such fruits and vegetables as tomato, pineapple, and pawpaw. These were all late introductions, though it is not known when they first appeared. Bowdich in 1817 records oranges, pawpaw and pineapples as being on sale in Kumase market. Plantain, the favourite carbohydrate food of the modern Asante, is rather a mystery, though it has been assumed that it was introduced along with banana in the first millennium A.D. ; but its name in Twi, borodee, 'European yam' implies association with Europeans. It is not known for certain even if sorghum and pennisetum, both native African crops, could have been grown in forest conditions - it seems unlikely, and we must assume that various species of yam and colocasia were the main diet together with various forest fruits no longer in use. Traces of root crops are not easy to identify in the archaeological record, and although it is not impossible to find traces of yam, none have so far been identified. Thus, there is no firm evidence for the basic diet of early inhabitants of the area.
19It is not easy to develop a picture of what the town was like in early days. Mud houses of the traditional type leave very little trace, though an occasional floor was found. The main evidence for human activity was, as already suggested the presence of much pottery. The discovery of pottery makes it quite certain that there was human activity at Asantemanso from the early date already indicated, though it is difficult to use this evidence to throw much further light on the nature of life. Iron smelting was practiced from some time in the first millennium A.D., and though smelting furnaces have not been found, fragments of slag are good evidence for the presence of iron smelters in the vicinity.
20The pottery, though still in a preliminary stage of analysis, shows two distinct periods. The very earliest levels yielded examples of what has come to be known as 'earthworks' pottery (Fig. 2).
Figure 2 : Open jar (OJ1), Asantemanso
21This pottery, named from the early earthworks sites in the Birim valley, was described by Oliver Davies (1967 : 287-290), who called it “Entrenchments” pottery ; by Kiyaga-Mulindwa (1978) ; and by Bellis (1972) in his excavation at Twifo Heman. This pottery, with its distinctive characteristics marked by elaborate impressed designs and folded rims, is widely spread throughout Akan territory (for a distribution map and a valuable commentary, see Bellis 1987 : 36-50). Its existence in the early levels suggests some change in either the people living at Asantemanso and other sites in the area or, for reasons which we cannot know, a considerable change in the pottery tradition. Many more distributional studies need to be made before much can be said with confidence about the early inhabitants, before about A.D.1000, who made this pottery. Bellis would date the floruit of this pottery much later - between A.D. 1500-1600.
22The second, and later, group of pottery is characterized by wide flaring rims, mostly bowl forms and rather simple incised designs. The shapes and the frequency of black burnished slips suggest that this pottery is ancestral to modern Akan wares, and it may be appropriate to designate it 'Akan ware'.
23Very few imported artefacts have been found at Asantemanso, though one area (site 87/2) produced a few pieces – imported ceramics, clay European pipes, gunflints and some glass – dating from the eighteenth century A.D. These illustrate that by then, as might be expected, the Asante were in touch with Europeans on the coast and there was some trade interchange. The amount of imported material is very small, but this need not be a surprise. It is likely that by the eighteenth century Asantemanso had lost much of its importance, and may not have been in a position to acquire many of the luxury goods which were coming into Asante, at least to Kumase, by soon after the traditional date of 1699 for the establishment of the Asante kingdom.
24Two small villages, both associated with stories of important early personages in Asante history, were investigated to provide information on sites which were assumed to be smaller and less important than Asantemanso, to see if they were of equal age and if the material remains from sites of less prestige were similar to those from a sacred site which had at one time been of political importance. Both were near to Kokofu, known to have been an important place in the early history of the Asante State. One of these, Anyinam, was the birthplace of Osei Tutu, the founder of the Asante kingdom. The story of his birth at this place is told with considerable detail, and there is a shrine in the village where libations are poured to his memory.
25Excavations at this place found that it was not as old as Asantemanso, and the two radiocarbon dates obtained are in the fifteenth and eighteenth centuries. The village is still occupied, and today it is considerably larger than Asantemanso. Survey and small tests in the surrounding neighbourhood did not show any evidence of occupation beyond the area of the present village, so excavation was confined to several large mounds, assumed to be middens, immediately on its edge. Examination of these mounds produced materials from the last two to three hundred years, including (in addition to local ceramics similar to those from later levels at Asantemanso) objects of European origin : beer and gin bottles, European ceramics, kaolin tobacco pipes and trade beads. These show the existence of European trade contacts from at least the early eighteenth century to the present day.
26It is noticeable that there are more European imports here than at Asantemanso, and a greater number of tobacco pipes, not only the European kaolin pipes, but also the much more elaborate locally designed and made ones (Vivian, this volume). The presence of a number of local pipes is of interest. These objects, presumably not in use until after the introduction of the habit of smoking tobacco, have been a matter of discussion for a long time in Ghana archaeology. That they were for smoking is sufficiently clear, but there has been some doubt as to whether or not they can antedate the introduction of tobacco and pipe smoking by Europeans – certainly not before 1492 and probably somewhat later. The range of patterns and variation in the decoration of the locally made pipes, which seem to owe nothing in design to the European clay (kaolin) ones, enabled Ozanne (1962 ; 1964) to develop a chronological sequence for them. He saw European pipes as being the forerunners of the local ones, and with some skill proposed a chronology – which, however, is now in need of re-examination, since the discovery of so many more examples, often from stratified sites.
27The local pipes are so much more elaborate in design than the European that it implies they may not have been copied, as Ozanne (1962) suggests, from the imports. It may be that the practice of smoking and the design of the pipes came from the north. Examples of not very dissimilar pipes have been found along the Niger (Daget and Liger 1962), in the Sudan (Balfour Paul 1951), and in various places in the Eastern Mediterranean.
28The suggestion has been made that Africans were smoking other materials, such as hemp, in pipes prior to European contact. Since pipes are not found in Asante in levels dated prior to about the beginning of the seventeenth century, it seems that this is not so for at least this part of Africa. It is also not certain that smoking spread from the coast inland. It is known that the practice came to the Mediterranean and to North Africa very soon after tobacco was introduced to Europe and it may have spread to the inland parts of Ghana from that direction.
29Vivian (1992), analyzing the numbers of local and imported pipe fragments, has observed that pieces of kaolin pipes are all in late levels at the three sites where they were found, whereas the pipes in the early levels are all indigenous. Vivian suggests that this may be due to a conscious choice and that the Asante were showing a strong preference for pipes of local manufacture, but it is also possible that pipes of northern styles had reached Asante before the European kaolin ones arrived.
30The second small village at which excavations were undertaken was Esiease – this also is connected with the Osei Tutu story and is claimed as the home village of his grandfather. The local tradition says that the people, who belong to the Agona abusua, originally came from Denkyira, the state which in pre-Asante days dominated the area and whose defeat at the battle of Feyiase led to the triumph of Asante. The present-day village elders and the queen mother said that the modern location of the village, on a recently-made road, was the third settlement in the area.
31The areas occupied in the earlier settlements were well known to local informants and we were taken to both of them. The first settlement is covered in bush with a few small cleared plots for crop growing. This area yielded only small scatters of pottery together with several locally made pipes of a type, which Ozanne dated to between 1655 and 1690. The area of the second settlement was also examined and found to consist of a number of low mounds of collapsed houses and several middens. The site was abandoned only within living memory – perhaps about fifty years ago.
32Excavation of this site produced results similar to those from Anyinam, with a selection of similar trade items. The presence of these items throughout all levels shows that the second site was only occupied during the last few hundred years. Carbon dates give a range from the sixteenth to the nineteenth centuries.
33The final site investigated was Adansemanso (Fig. 3).
Figure 3 : Adansemanso site map.
34This site is further south and is in the region of Adanse, of which it was the former capital and seat of the chief, the Adansehene. It is listed in many traditional sources as one of the five original towns of the Akan people, and should, therefore, be as early as Asantemanso. Since Adansemanso was the only one of the five apart from Asantemanso which is easily located on the ground, it was chosen for excavation so as to provide a comparison with Asantemanso both chronologically and in the mainly ceramic material that would be found. Tradition says that the site was abandoned when the inhabitants moved to the nearby villages of Ahinsan and Dompoase. The present capital of Adanse is at Fomena, still further south, where it has been since at least the time of T. B. Freeman, who visited it in 1839.
35There are now no buildings on the site, which is covered with bush ; but some areas have been cleared for farming, and this clearance much facilitated examination of the site. Excavation was carried out here briefly at the end of the 1991 season, enabling a few mounds to be examined and a general survey carried out. Evidence for collapsed houses can be clearly seen in the presence of mounds, some of which are so aligned as to indicate streets. Excavation of a number of these mounds confirmed that they were houses.
36A final season at the site in 1995 placed emphasis on examining these linear mounds, which are approximately 5 m across and in one case 100 m long. It can be confirmed from the several floors discovered that these were domestic buildings. From this work, it can now be seen that Adansemanso was a large town. The exact size still remains to be determined but it seems to be the largest ancient town yet known in the forests of central Ghana. A special interest of this site is that very few foreign imports were found, and these only in the upper levels. The same is true of tobacco pipes, both imported and local, which strengthens the tradition that the site was largely abandoned before the end of the sixteenth century.
37The indigenous material found - mostly pottery, as usual - has still to be compared with that from the other sites ; but first studies show that it is not significantly different from that found in pre-seventeenth century levels at Asantemanso. Amongst the small items found were two brass weights of the well-known type of 'gold weight' common in collections of objects from Ghana, so called because they were used for weighing gold. They have very seldom been found in excavations.
38This site is also the only one of those investigated to show significant direct evidence for iron working. The occurrence of large quantities of slag at Adansemanso confirms traditions reported by Anquandah (1982 : 68), who says :
Oral traditions state that before the coming of the Europeans many Akan farming towns in Adanse built iron furnaces and used local iron called atwetweboo1 in the smelting of iron. Adanse towns such as Dompoase, Akrofuom, Bodwesanwo and Akrokerri specialized in the making of iron shields and weapons and agricultural implements. Adanse Edubiase was traditionally the 'Birmingham of old Adanse.
39This is valuable information, and Adanse-Edubiase is certainly the site of Adansemanso, since Edubiase is a very small village close to the site. Anquandah also says (1982 :40) :
Iron technology... still maintains a foothold, however precariously in a number of Ghanaian towns and villages such as Wawasi, Bamparase and Fumisua in Asante.
40It is not clear whether by "iron technology" Anquandah means smelting or forging or both. It has not been possible to visit these places to see if, some years later, the technology is still being continued.
41Interestingly, Bowdich (1819 :312) says of the Asante, "they have no idea of making iron from ore". It seems improbable that he was right, though it may have been true of Kumase town. It is also possible, though unlikely, that imported iron was already sufficient for local needs. Bowdich (1819 : 334) also says that though iron was available from Dagwumba (Dagomba), European iron was preferred.
42One important result of the four years of investigation in Asante has been to show that some settlements are very much older than previous writers had suggested. The general tendency has been to suggest that the development of large settlements in the forest was a rather late phenomenon, and perhaps only occurred after the stimulus of trade contacts with Europeans on the coast in the sixteenth century. Our recent investigations have shown that Asantemanso and Adansemanso considerably pre-date this, and that other factors must be found for the development of towns and villages from as early as the ninth century A.D. One of the reasons for the cautious and late dating of urban development has been the rather scanty traces these settlements show on the surface. It has been argued that continuous settlement would have caused a greater build-up of occupation debris, so as to form mounds similar to those of the Near East. This was a matter to which some attention was paid, and it became clear that lack of massive build-up at one place did not necessarily mean that the site had not been occupied for a long time. As the plan showing the various centres of occupation at Asantemanso makes clear, there seems to have been over the centuries a number of rather small shifts of location – the recent history of Esiease shows this from oral sources supported by archaeological finds, whilst the shifts of urban emphasis at Asantemanso are shown archaeologically. The reasons for these slight shifts, which inhibited the build-up of large mounds, was considered carefully ; Vivian developed the interesting and convincing suggestion that erosion, clearly seen in many modern villages, caused the shifts in areas of concentration. His paper (Vivian 1991) presents the arguments and the facts to support them. The nature of soil accumulation and the effect of heavy rainfall have also played a part.
43In addition to areas of domestic occupation, several sites at Anyinam, Esiease and Adansemanso were connected with the cult of the dead. These areas, which were places where the dead were commemorated, were found to contain a variety of ceramic materials ; there were many pottery vessels, a number of objects whose purpose could not be certainly understood, and many examples of the well known terracotta heads which have been studied on several occasions by art historians, notably by Preston (1968) and Sieber (1972, 1973).
44The finding of these asensie sites, as they are known in twi, was of considerable interest ; this was only the third time, to our knowledge, that any of them had been examined archaeologically. Previously Davies (1977), followed by Calvocoressi (unpublished), had worked at Ahinsan not far from Adansemanso, where a large number of heads were found ; and Bellis (1972) found many fine examples at Twifo Heman. Their role in funerary ceremonies has been well described, particularly by Bellis (1982), and need not be repeated here. Vivian (1992) has given a description of those found in the recent excavations.
45It is noticeable how the originally sacred and commemorative role of these assemblages has now been largely forgotten - in one case, at Esiease, a modern house had been built over one. The local people had no hesitation in helping us find these sites and had no reservations about allowing us to excavate and investigate what had once been sacred places.
46One of the conclusions that can now be drawn, especially for rural Asante, is that during the years after we might first expect European trade goods to show up, they made a very slight impact at the sites that we excavated. It is certain that during the eighteenth century there must have been massive imports of various both useful and luxury items into Kumase to serve the purposes of the upper classes. Although there is not much direct evidence for the seventeenth or eighteenth centuries, by the time of Bowdich's visit (1817) there was a rich display of European goods, though his description of Kumase market (1819 :324) does not reflect this, and the range of goods was rather limited. Bowdich lists rum, pipes (though he does not say whether local or imported), beads, looking glasses, silk, cotton cloth, powder and small pillows. It is worth noting that T. B. Freeman (1844 :56), in April 1839, specifically mentions 'native tobacco pipes of very neat manufacture' (1844 : 56). Both Bowdich and Freeman give accounts, which suggest that the main items of import, certainly guns, were not put on sale to the public. Many of the more exotic items, no doubt, went straight into the hands of the upper classes. Very little of this imported material is reflected in the villages we investigated, and it seems that the rural population did not have the resources or perhaps the inclination to acquire many foreign items. Some of the items which may have come to the small villages will not have survived - particularly cloth, but the small number of imported tobacco pipes, ceramics, glass and even gun flints strongly suggests that villagers remained almost entirely dependent on local products until well into the twentieth century.