Fig.1 : Map of Ghana showing the location of Coconut Grove
1My research in coastal Ghana over the past fifteen years has concentrated on the transformations that occurred within the context of expanding European trade and the entrance of West Africa into the emerging Atlantic system (DeCorse 1987 a, 1987b, 1988, 1989, 1992a, 1992b, 1998, 2001). The focus of much of this work has been at the site of Elmina and other African settlements that were directly associated with the European trade entrepots on the coast founded between the late fifteenth and early nineteenth centuries. This ongoing research has now been greatly expanded and elaborated on by several other researchers, including Ed Carr (2001), Greg Cook (DeCorse et al 2000), Sam Spiers, and Gérard Chouin (this volume). This work, collectively subsumed under the Central Region Project, has provided a great deal of insight into the European trade and the distinctive characteristics of the African polities that emerged, or at least flourished, during the period of the Atlantic trade.
2Since 1993, my research has increasingly focused at the regional level and on earlier sites that span the era of pre- and post-European contact. Documentary records indicate that when Europeans arrived in the late fifteenth century the inhabitants of the central Ghanaian coast were linguistically Akan, linguistic divisions on the coast having been largely consistent with their current distribution (DeCorse 2001:18-19 ; Hair 1967). The objective of my current research has been to place the coastal Akan sites associated with European trade posts in broader context and in that way better assess the nature of pre-European contact society and the transformations that subsequently occurred. A systematic survey of sites was undertaken of the coast and lagoonal hinterlands to the east and west of Elmina during 1993, 1995, 1997, and 1998 (DeCorse 2001 ; DeCorse et al 2000). As a result of this survey, some two dozen sites likely pre-dating European contact were located.
3Data from coastal sites dating from the first millennium to AD 1700, including site distribution and artifact assemblages, challenge assumptions that have, at times, been made concerning the nature of pre-European contact society, technology and trade. They also illustrate the transformations that occurred during the post-European contact period. Prior to the seventeenth century, small fishing and farming villages were scattered along the coast and the adjacent hinterland (DeCorse 2000:18-20). Larger population centers were still located in the interior forest and northern savanna. This pattern would dramatically change in the following centuries‑; population growth along the coastal margin and concomitant changes in the sociopolitical structures of coastal polities characterize the post-European contact period.
4Evidence for the pre-European contact occupation of the coast is represented by surface scatters and midden deposits along low rises just above the shoreline or adjacent to lagoons and, in some locales, hill top sites. At Elmina and Eguafo, both settlements that have substantial post-European contact period occupations, pre-fifteenth century material has been recovered from the bottom of excavation units. The deposits are small, and provide limited insight into settlement size and artifact inventory. In some instances, however, scatters of surface material, mostly consisting of ceramics, with occasional fragments of iron, quartz flakes and stone beads extend over large areas, suggesting significant pre-fifteenth century coastal occupation. At Brenu Akyinim and Coconut Grove, for example, deposits extend for almost a kilometer along the shore. While testament to the importance of coastal resources during the pre-European contact period, this material likely represents small shifting settlement over long periods of time rather than single, continuous settlement occupations.
5The dating of these early sites is difficult. In the absence of suitable charcoal samples, several ceramic sherds have been dated by thermoluminescence. These samples were obtained from the lower extent of the cultural deposits encountered at Elmina, Eguafo, and Brenu Achyinim and thus may represent the earliest occupation dates. The dates obtained include A.D. 1112+276 and 1109+146 for Brenu‑; AD 1238+77 and 1325+67 for Eguafo‑; and A.D. 887+104 and 1012+124 for Elmina. Unfortunately, the ceramics from these sites were extremely fragmentary and eroded, and consequently provided limited indication of vessel form. No features were uncovered.
6A large assemblage of ceramics and other artifacts from a clearly delineated, well dated, stratigraphic feature were recovered from the Coconut Grove Site located on a low rise above the shoreline, four kilometers west of Elmina (fig. 1 p. 44 and 2 p.53). The site is characterized by a large deflated midden deposit of ceramics, with smaller quantities of stone beads, iron slag and quartz debitage. Presently, three AMS dates on charcoal samples and a thermoluminescent date on one of the associated ceramic sherds are available from a sealed feature. The three radiometric dates obtained from within the feature yielded calibrated ages of AD560 to 670, AD580 to 680, and AD 680 to 880, while the provisional, associated thermoluminesence date is AD 584 +/- 597 (James Feathers, personal communication 11 Feburary, 2002). When additional radiocarbon dates are obtained these will provide one of the earliest, well dated assemblages available for coastal West Africa.
7The later occupation range of some of these coastal sites is provided by European trade materials, including occasional pieces of Portuguese earthenwares, early seventeenth century pipes, and a pre-seventeenth century Portuguese coin recovered at Eguafo (Samuel Spiers, personal communication January 2003). Conspicuously absent are later Dutch and English trade materials that characterize eighteenth and nineteenth century settlements.
8Beginning in the seventeenth century, coastal trade entrepots such as Elmina and Cape Coast expanded in size and importance. This growth likely does not solely represent natural population increase, but rather augmentation by immigration from other parts of the coast, hinterland and interior forest (DeCorse 2001:31-36). In the case of Elmina, the African settlement adjacent to the Portuguese Castelo de São Jorge da Mina, the small, fifteenth century settlement had likely been subservient to the coastal polities of Eguafo or Fetu. Following the establishment of Sáo Jorge, the African town emerged as the center of a small, independent state, which eventually expanded control over smaller villages and hamlets. The population of Elmina expanded to between fifteen and twenty thousand inhabitants by the early nineteenth century.
9The artifact inventories of sites dating from the pre-European contact period through the early post-contact era are dramatically different from those of the later seventeenth, eighteenth and nineteenth centuries. Aside from the paucity of European trade materials, the earlier sites are characterized by stone beads, occasional fragments of slag and iron, lithics, and ceramics. The pottery is characteristically fragmentary and eroded, and is quite distinct in form and decoration from later ceramics (DeCorse 2000:116-123). Classic black smudged, carinated vessels common on interior Akan sites only appear in the nineteenth century (fig. 3 and 4). These data complement work undertaken by other researchers in adjacent parts of the coast and in other areas of southern Ghana, including work by Kiyaga-Mulindwa (1982) in the Birim Valley, Bellis (1987) at Twifo Heman, and Ozanne (1962) in Accra and Shai to the east that document similar changes in the post-fifteenth century period.
10A small quartz tool industry appears to continue until at least the end of the first millennium AD. At the Coconut Grove Site, surface material included quartz flakes, angular shatter and microcores. A limited quartz component, consisting of a small numbers of quartz flakes and two polished stone artifacts, was also represented in the first millennium AD excavation contexts.
Fig.3 : Miscellaneous first millenium AD ceramics from Coconut Grove
Fig.4 : 19th century ceramics from Elmira, note black carinated vessels on the right
11Iron was also produced on the Ghanaian coast prior to European contact. The earliest dates for iron production come from the Coconut Grove Site, where small amounts of slag and iron fragments were recovered from surface contexts, as well as from the well dated, seventh-eighth century AD stratigraphic context. Possible furnace bases were also identified at Coconut Grove. These surface features had fired reddened walls 13 to 14 centimetres thick and an overall diameter of 1.20 metres. Unfortunately, permission could not be obtained to excavate these, and their construction, age, and associations with other material at the site could not be determined. The lack of iron slag or tuyeres in clear association with the features may, in fact, suggest that they are not smelting furnaces, but the bases of fish smoking ovens. If, in fact, they represent furnace bases they would appear to be slightly larger in size to a natural draft furnace with an aperture near the base for tapping the slag excavated at Cape Coast, which is of uncertain age (Penfold 1972‑; cf. Pole and Posnansky 1973). Although not securely dated, the Cape Coast furnace is likely pre-seventeenth century.
12Slag occurs on many Central Region sites, including deposits extending almost a meter deep in slag mounds located north of Dutch Komenda (Gerard Chouin, personal communication January 2003). The majority of the coastal evidence, however, consists of isolated fragments or relatively small mounds that are not evidence for substantial production. Limited evidence for iron production or, at least, iron smithing is referred to in sixteenth and seventeenth century documentary accounts of the coast (Bosman 1967:128‑; Müller 1673:254‑; Towerson in Blake 1942:378). The limited post-seventeenth century archaeological evidence and documentary references to iron smelting may suggest that what coastal production there was soon ceased in the face of European imports.
13There is also evidence for gold working. A single gold bead was excavated from the Coconut Grove Site (DeCorse 2001:127). Although from a secure stratigraphic context, it was not found within the clearly delineated, first millennium AD stratigraphic feature previously discussed and associated charcoal samples associated with the bead have not yet been dated. The, hollow, spherical bead is made from cold hammered gold foil. Techniques such as cold hammering, cutting and bending may have been more characteristic of earlier production than the more elaborate cire perdue (lost-wax) casting of gold and brass objects that is described by writers beginning in the late-seventeenth century (e.g. Barbot 1992:494, 527, Plate 43).
14Coastal resources such as fishing and salt production, which are well-documented in nineteenth and twentieth century records, have been cited as important components of a pre-European contact trade network that linked the coastal Akan with the forest hinterland (e.g. Bech and Hyland 1978:21‑; editors in de Marees 1987:188 fn 1‑; Van Dantzig 1980:3). Indeed, some have suggested that these items were sought by Mande traders, who made up a separate quarter of traders in fifteenth century Elmina. This picture is not born out by archaeological and historical research. While there is no question that marine fishing using canoes played a significant role in coastal subsistence, large-scale marine fishing, continuity with ethnographically documented technologies, and a fifteenth century interior trade in smoked fish cannot be taken for granted. In fact, faunal remains and mollusks (particularly Arca senilis) recovered indicate a heavy and continuing reliance on lagoonal resources throughout much of the post-European contact period (DeCorse 2001:104-115 ).
15There are documentary references for salt production and trade in the Central Region in the post-European contact period. However, the extent of production and of long distance interior trade needs to be carefully examined. While salt production using large lagoonal salt pans are typical of the Central Region today, this is a twentieth century phenomena (DeCorse 2001:140-142). Evidence for early production, for the type of manufacturing methods employed, and for a pre-seventeenth century long distance salt trade is sparse. Prior to the twentieth century, production in the Central Region was carried out by boiling sea water in pots (e.g. Müller 1673:244). Production was likely aimed at local consumption and trade with the immediate hinterland, rather than trade with the more distant interior. The extent of salt production and trade at Elmina and adjacent areas of the central Ghanaian coast likely expanded substantially during the post-European contact period. Salt produced by boiling was considered of high quality by European traders and references to a salt trade at Elmina are found in seventeenth through twentieth century documentary references and in oral histories (e.g. Bosman 1967‑; Daaku and Van Dantzig 1966‑; Fynn 1971).
16Evidence for increased craft production characterizes the post-seventeenth century period, including increasing evidence for specialization in pottery making, salt production, and metal working (DeCorse 2001:116-144). Although these industries predate European arrival on the coast they increased in importance. Gold smithing, for example, as well as casting using imported brass or bronze clearly predates the arrival of the Europeans in the forest hinterland. However, evidence for coastal production prior to the seventeenth century is limited. Both documentary and archaeological evidence for production, as well as the diversity and sophistication of the techniques represented expanded with the growth of larger urban concentrations like Elmina and Cape Coast.
17In addition to casting, the local fabrication of gold dust pans (twi, famfa), spoons (twi, atere), and scales (twi, nsenia koraa and nsenia dua) using reused European sheet brass vessels also expanded. The production of forowa had begun by the early eighteenth century. Forowa are small sheet brass containers decorated with repoussé, punch-work, and incised designs of geometric patterns and animals. Although they have at times been considered a nineteenth century product of the coastal Akan, a forowa recovered from a well dated, mid-eighteenth century archaeological context at Elmina indicates earlier production. These items, as well as a wide variety of bracelets, rings, and necklaces were locally fashioned from imported European metal.
18Another early craft was likely stone bead production. Stone beads are common on sites pre-dating European contact. Final assessment of their origins must await further study, but it is possible that some of them were produced locally. Significantly, Elmina is noted for the polishing of the enigmatic cori or aggrey beads in sixteenth century Portuguese sources. The local polishing and production of glass beads appears by the early seventeenth century (DeCorse 1989b, 2001:135-138).
19In addition to the preceding industries, other crafts such as ivory carving became centered in the coastal towns (DeCorse 2001:138-140), while trade in an increasingly diverse range of commodities emerged as a mainstay of the coastal towns (DeCorse 2001:142-174). When the Europeans arrived on the coast they found a ready demand for many items—beads, brassware and cloth. They also brought new items, both to explore new markets and cater to increasingly specific local demands. The coastal settlements were the frontier of opportunity through which these goods passed to the interior.
20The archaeological evidence from the central Ghanaian coast indicates change in settlement patterns and craft production in the centuries following European contact. A pattern of small, dispersed settlements characterized the coast from AD 500 to 1700. In the following centuries, coastal settlements associated with European trading enclaves increased in size and became production centers. These changes in settlement size and artifact inventory only hint at the sociopolitical transformations that occurred. Coastal towns, such as Elmina and Cape Coast, emerged as independent states. While Akan in inspiration, the coastal states developed distinctive aspects of kingship and political organization (DeCorse 2001:38-47). Based on these data it is possible to say that there were substantial changes in coastal Ghana as a consequence of the expanding Atlantic world.
Fig.2 : Coconut Grove excavation units