Internal Meanings: Computed Tomography Scanning of Koma Figurines from Ghana

Abstract

Since the 1980s art historians and archaeologists have been aware of the terracotta figurines from Koma Land in northern Ghana (Kröger 1988; Anquandah 1987, 1998). The pioneering excavation and publications by James Anquandah (Anquandah and van Ham 1985; Anquandah 1987, 1998) established their provenance, and unprovenanced figurines from illegal excavations have subsequently increased known numbers. The dominant focus in publication of the Koma Land corpus has been upon what the figurines depict externally (e.g., Anquandah 1987, 1998; Kankpeyeng and Nkumbaan 2008, 2009; Insoll and Kankpeyeng 2014; Insoll in press a). Following the successful trial use of lower resolution Computed Tomography black scanning which produced black-and-white images of five figurines in May 2010 (Insoll, Kankpeyeng, and Nkumbaan 2012:31–32), a further sample of eight terracotta figurines was CT scanned and color images produced in 2013. These are the focus here. All the figurines were from archaeological excavations at Yikpabongo in Koma Land, and the CT scanning indicated that all eight had deliberately made cavities running from their surface into the body of the figurine. This suggests that the importance of some of the figurines was potentially greater than their external appearance and that part of their significance might have been derived from their internal meanings as well.This paper reports on the renewed research in Koma Land that led to the retrieval of the figurines, and on the scanned figurines themselves. Why the cavities were made is unknown, but various possibilities are explored. This is considered with reference to the Koma figurines and through wider comparison with other archaeological terracotta figurines from West Africa that have evidence for cavities.Koma Land spans the borders of three modern administrative regions of Ghana—Upper West, Upper East, and Northern—where it covers an area of approximately 100 km2 and is drained by tributaries of the White Volta such as the rivers Sisili and Kulpawn (Anquandah 1998:21; Kankpeyeng et al. 2013:477) (Fig. 1). The name “Koma” is derived from that of one of the current ethnolinguistic groups in the region. The contemporary Koma are not connected with the makers of the terracotta figurines (Anquandah 1987:177), having settled in the region c. 130 years ago (Kröger and Saibu 2010:1). The disconnection between present populations and the figurine makers is also indicated by the name given the figurines by the Koma, kronkronbali, meaning “progeny of the ancestors” or “olden days children” (Anquandah 1998:13, 15).The Koma Land figurines were first recorded by anthropologist Franz Kröger, working in neighboring Bulsaland, who was told about them by a missionary (cf. Kröger 1988:132). Kröger reported the discovery of the figurines to the Archaeology Department at the University of Ghana, and James Anquandah completed the first excavation in Koma Land in 1985 (Kröger 1988:132; Anquandah and van Ham 1985; Anquandah 1998). During Anquandah's research, 105 mounds were identified, ranging in size from 35 to 4 m in diameter, and four were excavated in the village of Yikpabongo (Anquandah and van Ham 1985; Anquandah 1998:76; Kankpeyeng et al. 2013:479). Anquandah (1987, 2003) interpreted the mounds as burial mounds where the personal possessions of the deceased, including terracotta figurines, were deposited. Parallels were drawn with the “Senegambian and Malian megalithic tombs” (Anquandah 1987:179) and initially the mounds were dated by Thermoluminescence (TL) dating techniques to between the fifteenth and seventeenth centuries ad.1 The chronology was subsequently extended to the period c. ad 1200 to 1800 (Anquandah 1998:82).Anquandah (1987, 1998) put the framework of our understanding of Koma Land in place. The figurines attest social complexity (Kankpeyeng and Nkumbaan 2009:201) and were made by sedentary farmers living in village communities who also produced pottery and iron, as both slag and finished iron objects such as knives and bracelets indicate (Insoll, Kankpeyeng, and Nkumbaan 2012:27–28). Faunal remains from cattle, sheep, and fowl indicate animal rearing (Anquandah 1998:92). The recovery of numerous grinding stones (Anquandah 1998:93–94; Kankpeyeng, Nkumbaan, and Insoll 2011:211) demonstrates plant processing, and some must have been used for foodstuffs such as millet (Insoll, Kankpeyeng, and Nkumbaan 2012:27).2Although stylistically unique and unconnected with other terracotta figurine traditions such as those of the Inland Niger Delta (cf. McIntosh and Keech McIntosh 1979; McIntosh 1989), the Koma figurines were not produced in a vacuum. Horse- and camel-rider figurines attest awareness of long-distance trade by the inhabitants of Koma Land (Fig. 2) (Kankpeyeng and Nkumbaan 2009:201). It is also indicated by the frequent depiction of cowry shells modelled on figurines as items of adornment (e.g. Anquandah 1998:48, 2003:139). These were items obtained by trade (Lovejoy 1985:669; Insoll 2003:251–52), and Cypraea moneta and Cypraea annulus cowry shells and glass beads, other trade items, have also been recovered from excavation in Yikpabongo (Anquandah 1998:77; Insoll, Kankpeyeng, and Nkumbaan 2012:28).With the termination of Anquandah's research in the 1980s, Koma Land became the focus of looters’ attention, with many figurines illegally removed and decontextualized (Kröger 1988:142). Efforts by the Ghana Museums and Monuments Board (GMMB) over the last fifteen years have radically reduced the scale of looting as the local community has become sensitized to the cultural value of the figurines (e.g., Kankpeyeng and DeCorse 2004). A significant factor in reducing looting and increasing public understanding of heritage in Koma Land has been renewed archaeological research by a team from the GMMB and the University of Ghana that has taken place in the area under the direction of Benjamin Kankpeyeng since 2006. The primary focus of this research has also been upon Yikpabongo. Timothy Insoll was involved in the excavations in 2010 and 2011 with the remit of assisting in the interpretation of figurine function. Sharon Fraser completed the subsequent CT scanning.The figurines were recovered from two mound sites in Yikpabongo, given the codes YK07/08 and YK10-3/YK11 (YK = Yikpabongo, followed by the year of fieldwork [cf. Kankpeyeng and Nkumbaan 2009; Kankpeyeng et al. 2013]).YK07/08. YK07/08 is a low mound measuring 10 m east to west and 8 m north to south (N10.14449°, W001.33562°). It was excavated using arbitrary levels of 10–20 cm depth where natural levels were not readily recognizable and a 1-m grid reference recording system. Sterile deposits were reached at a maximum depth of 170 cm. Ceramic figurines, potsherds and other artifacts (e.g., iron bracelets and utilitarian artifacts such as knives, grinding stones, small numbers of glass beads and cowry shells) were frequently recovered in repeat patterns superimposed on each other, “from the upper level of the excavated unit to the bottom” (Kankpeyeng and Nkumbaan 2009:196). These patterns were formed of “contextual spreads of local potsherds above or in association with figurines” (Kankpeyeng and Nkumbaan 2009:196), and other items such as pieces of muscovite-biotiteschist, quartz querns, and ceramic discs made from potsherds chipped into a circular shape (Kankpeyeng and Nkumbaan 2008:99). The latter were probably stoppers from gourds or horn containers,3 which had disintegrated over time (Insoll, Kankpeyeng, and Nkumbaan 2012:36). The artifact concentrations had been laid in pits dug into the natural lateritic gravel. The site has been dated by four TL samples to between the mid-tenth and early fourteenth centuries ad (cf. Kankpeyeng and Nkumbaan 2009:198; Kankpeyeng, Nkumbaan and Insoll 2011:209).4YK10-3/YK11. YK10-3/YK11 is another low mound measuring approximately 18 m east to west and 15 m north to south (N10.14480°, Woo1.34052°). It was excavated using an L-shaped grid measuring 10 m × 5 m × 10 m × 5 m, extended in 2011 to form an overall grid of 10 m × 10 m (Kankpeyeng et al. 2013: 481). The same excavation methodology as for YK07/08 was employed. The mound stratigraphy was simple with an inconsistent depth of archaeological material of between 20 and 30 cm depth overlaid by a thin layer of modern dust and rubbish (c. 1–3 cm), and below this by sterile or nearly sterile soil (c. 10 cm) before the natural red gravel filled deposits were reached at a depth of between 40–50 cm from the surface (Insoll, Kankpeyeng, and Nkumbaan 2012:28).Similar contextual arrangements of artifacts to those described for YK07/08 were recorded with recurrent patterning of spherical stone (quartz and granite) grinding stone, lower querns, and pottery disks overlying the layer containing figurines, and the figurines’ being “sometimes ‘nested’ within arrangements of potsherds and associated with what are possibly libations structures” (Insoll and Kankpeyeng 2014:34–35) (Fig. 3). The latter are clay structures made of daub (low-fired clay) and arranged in a circular pattern interwoven with potsherds (Insoll, Kankpeyeng, and Nkumbaan 2012:36). YK10-3/YK11 was dated by radiocarbon to the eleventh to twelfth centuries ad (Kankpeyeng, Nkumbaan, and Insoll 2011:209).5Human and animal remains were largely absent from both mounds (Kankpeyeng et al. 2013:483). Notable exceptions were provided by a single human skull recorded in YK07/08 (Kankpeyeng and Nkumbaan 2008:97–99), and an arrangement of human remains from YK10-3/YK11. The skull in YK07/08 seemed to have been placed facing the ground and no other human remains were present (Kankpeyeng, Nkumbaan, and Insoll 2011:210). The skeletal remains in YK10-3/YK11 consisted of a fragmentary human skull placed facing into the earth, with fragments of human long bones southeast and southwest of the skull. A human jawbone was also recorded along with a separate pile of twenty-seven human teeth, the latter east of the skull. These teeth were from two individuals, a younger adult of about 20 years age, two of whose nineteen teeth present had been filed, and the other eighteen teeth were from an older adult (Insoll, Kankpeyeng, and Nkumbaan 2012:38).6 In both YK08/09 and YK10-3/YK11 it is clear that the human remains had been deliberately selected and arranged.Although these selected human remains were found, it is apparent that the figurine-filled mounds, contrary to Anquandah's (1987:179; 1998:82–83; 2003:138–39) interpretation, are not burial mounds. Reexamination of Anquandah's (1998:87) data likewise indicates that the human remains he found were fragmentary, demonstrating that the mounds were not for primary burial purposes but instead loci for processes of fragmentation, selection, curation, and deposition (Insoll in press a). Complete burials have only been recorded below floors in house mounds, as with one in mound YK10-4 at Yikpabongo (Kankpeyeng, Nkumbaan, and Insoll 2011:209).7 This is a different type of site, being much larger in size than the figurine mounds, at 91 m east-west and 28 m north-south, and composed of the residue of domestic occupation. The house mound was radiocarbon dated to between the mid sixth and mid seventh centuries ad8 and did not yield any figurines (Asamoah-Mensah 2013:165; Kankpeyeng et al. 2013:492).Based on the contextual evidence and artifacts from the figurine mounds, it has been suggested that the mounds and their contents might have been shrines that served multiple functions in relation to healing, protection, fertility, ancestral veneration, and possibly witchcraft exorcising (Kankpeyeng and Nkumbaan 2008:101, 2009:201; Kankpeyeng, Nkumbaan, and Insoll 2011:209). Whether the figurines and associated material culture were utilized for ritual purposes prior to their deposition in the mounds is unknown. However, the repeat contextual arrangement of the figurines also suggests that “we are viewing the insitu residue of ritual action” (Insoll and Kankpeyeng 2014:35), and what this might have related to is considered below.The renewed research has also allowed reappraisal of the Koma Land chronology. The C14 and TL dates from Yikpabongo allied with two C14 dates obtained from the site of Tando-Fagusa, 25 km southeast of Yikpabongo in Koma Land (Kankpeyeng and Nkumbaan 2009:200),9 suggest the society that produced the figurines can be dated to between the sixth/seventh and thirteenth/fourteenth centuries ad. Why the production of figurines ended and Koma Land was depopulated is not known. Various possible reasons have been suggested, including climatic change, disease, migration, or slave raiding (Insoll, Kankpeyeng, and Nkumbaan 2012:28).Recovered from YK07/08 were 923 figurines and figurine fragments. Of these, 55 were complete and 868 were fragments. From mound YK10-3/YK11, 251 figurines and figurine fragments were recovered. This assemblage comprised 238 fragments, 6 largely complete and 7 complete figurines (Insoll, Kankpeyeng, and Nkumbaan 2012:29). The figurines were made either as solid objects, as with all the examples scanned, or from different parts. Arms and legs were occasionally modelled separately and then attached using joints of a ball or pin and socket type (Insoll et al. 2013:13) (Fig. 4). Orange clay was typically used, containing large particles of quartz. Only one figurine had a noticeably different grey clay fabric (YK08-AB9-L7). Traces of a pinkish or red outer slip are sometimes evident. The skill involved in producing the figurines suggests it was a specialist occupation, but whether this was by the same people who made the pots that were found, and if males, females, or both, produced figurines, is unknownIt is not the purpose of this paper to provide a typology of the Koma Land figurines, a couple of attempts at which have been made (e.g., Anquandah 1987:177, 1988:125–29, 2003:140; Kröger 1988:136). A revised typology is currently under construction (Kankpeyeng in preparation) and at present contains more than forty forms (Kankpeyeng et al. 2013:494). The broad figurine groups identified are human (stylized and realistic), animals and birds (mythical and actual), combined human and animal forms, cone forms (anthropomorphic and nonanthropomorphic), and objects modelled in clay such as gourds (Insoll et al. 2013), or stools (e.g., Anquandah 2003:140).Various interpretations have also been previously ascribed the It has been suggested that the figurines might (Anquandah or (Kröger 1988:136). figurines have been suggested as (Anquandah The of the research suggest that the figurines, the mounds, served purposes within an overall framework of healing, and protection, and this is further below.The CT in was connected to the of figurines in the University of prior to the of the at the for and time the sample size to eight All figurines to the were on in and the eight were on the of three as this was a it was to to cone and human figurines and not animal as these form part of a subsequent of all the figurines had surface some cavities were and different of figurines within the human and cone and and a from a CT scanning was in the at the University of CT have a and the being scanned is figurine was placed on a on a sample and was producing These were then The all the images and a that can then be further in other that are present in the figurines be as to the clay which in of grey on the of the were then in on their that natural the clay was formed into a figurine or within the clay were and cavities which were deliberately by the of the figurines, such as in the of through or were red and was used to where deliberately made were filled with these allowed the of the to be of the figurine cavities by CT scanning are in the is no significant in the of cavities on or figurine type for the only that not have an form which has a single All the and the figurine with a have cavities These cavities are by a clay either a cowry or (Kröger (Fig. figurines that the or such as the with and but as with the have multiple cavities from such as and The and are the frequently and on The cone figurines any meaning was ascribed the depth of the cavities is unknown. All the cavities are within the depth of cm and cm. The is into the figurine (Fig. It that into the clay it was was the used to the were used is unknown, on the forms can be The maximum is cm. The of some of the cavities suggests the of a or a might have been used to some cavities (e.g., on or on or or might have been used for other more cavities (e.g., on It is possible that some of the cavities were made by the or to be (Insoll et Why the cavities might have been made is considered below.The of cavities has been recorded within other West archaeological figurines, but this has been by and CT scanning to have been only used (Insoll in press a). an in the of CT scanning to terracotta figurines from the Inland Niger Delta of these are unprovenanced the is of The CT indicated through the of a to a with clay of the a had placed within the The of the cavities was interpreted as examples of figurines from northern also have evidence for These are (Insoll in press a). three human figurines from the and burial sites (c. seventeenth had circular either in the of the or upper (cf. and suggests that it is possible that these cavities were into of the figurines as having or of some form (Insoll in press connected with what as or the have also been recorded on some of the figurines from the sites in Ghana dated to between the and centuries ad (Insoll in press a). A for these cavities is by that the if at the of the and if from the or However, the of other in the region of some figurines for the of suggests that the other cavities have served purposes (Insoll in press figurines from sometimes also have cavities (Insoll in press a). to cavities in the of the of some and suggests that have been the of to the with increased have been or such as and than from interpretation (Insoll in press are also apparent on some of the figurines from (Insoll in press dated to between c. and ad These can be from the or (e.g., or the or of the that these the clay and but whether this is to all the figurine cavities is (Insoll in press for also to a circular on the of one figurine as has also internal cavities in figurines and related objects (Insoll in press a). has described indicated internal cavities in the and of objects made by the of were with These cavities such as the of animals selected of such as their or also understanding of a (Insoll in press by that a the of the running to a or which be through at one and with of the These examples of figurines and their cavities from the Koma Land examples and as such are not but indicate that the of cavities was not unique to the Koma Land figurine with the cavities on the figurines from the Inland Niger or the the cavities on the Koma figurines were made for a However, it not that were made for reasons (Insoll in press a). all the figurines have it is that were made to to as has also been suggested for some cavities on figurines from the Inland Niger Delta the cavities on the Koma Land figurines were made to it might be that all figurines have The cavities were also not made to the of separately modelled body parts. These but being in the form of a as indicated in 4 and the cavities are also in the to such a reasons can be it is not clear for what purpose the cavities were made and used, but various possibilities The cavities have been utilized for the of libations of as (Insoll and Kankpeyeng the cone figurines have been into the ground (Insoll and Kankpeyeng or into a small mound of (Kröger figurines with cavities (e.g., might have been used but being into the The cavities might also have been with of is in using to if might have been or into the It is their the that the cavities had such as or in them as a form of or the cavities might have been not (Insoll and Kankpeyeng but the material by from on is of in about the figurines might have been the was as from outer to and was between and and deliberately to the as on some of the figurines, it is to of these deliberately produced cavities as more than an or (Insoll and Kankpeyeng if some of the figurines were considered as then the of the cavities might have been significant that were as and (Insoll, Kankpeyeng, and Nkumbaan The cavities also attest the that the figurines were considered as having (Insoll in press a). have been of as objects and as such were by the and communities that made and them (Insoll, Kankpeyeng and Nkumbaan a single interpretation for the cavities on the Koma Land figurines that or is not possible to the healing, and of the This is potentially by other of the figurine assemblage that to reference such The Sisili was an area of (Anquandah the recurrent of modelled on the figurines with disease, of with such as an from (Fig. might also to healing, and It has also been suggested that some of the figurines, have been considered and used for purposes (Kankpeyeng and Nkumbaan of the cavities suggests that some of the Koma Land figurines had an importance what depict and both the internal and external figurine form potentially be significant (Insoll in press This the that were (Kankpeyeng, Nkumbaan, and Insoll but the from which their interpretation it is not possible to if were described by as and or were which were to a or This on the cavities suggest that the figurines, than or to use a made by in relation to figurines, and objects with the figurines suggests that an and a to the of for and with the through and ritual The cavities also have served to the and possibilities of that the figurines and as such of Insoll in press a). the interpretations for what these internal meanings might have been are currently it is that the of of figurine CT and and residue greater understanding of these and