https://orcid.org/0000-0002-6139-6227
ABSTRACT
This paper reports the results of a technological and microwear analysis of 110 pieces of worked bone from the Early Iron Age site of Broederstroom in the Magaliesberg region of South Africa. The results are compared to previous studies of bone tools from the contemporary Later Stone Age sites of Kruger Cave and Jubilee Shelter in an attempt to understand whether the Broederstroom pieces were made locally by the farmers or acquired through trade with hunter-gatherers. Although the bone tool technology at Broederstroom does bear some similarities to the bone material at Kruger Cave, it is quite different from that at Jubilee Shelter. This technological distinction likely indicates that the bone tool industry at Broederstroom is a product of local farmer industry. The use-wear results further point to a range of different activities practised by the farmers living at Broederstroom.
RÉSUMÉ
Cet article rapporte les résultats d’une analyse technologique et micro-usure de 110 morceaux d’os travaillés provenant du site du début de l’âge du fer de Broederstroom dans la région de Magaliesberg en Afrique du Sud. Les résultats sont comparés à des études antérieures sur des outils en os provenant des sites contemporains de l’âge de pierre ultérieur (Later Stone Age) de Kruger Cave et Jubilee Shelter dans le but de comprendre si les pièces de Broederstroom ont été fabriquées localement par les agriculteurs ou acquises grâce au commerce avec des chasseurs-cueilleurs. Bien que la technologie des outils osseux de Broederstroom présente certaines similitudes avec le matériau osseux de Kruger Cave, elle est assez différente de celle de Jubilee Shelter. Cette distinction technologique indique probablement que l’industrie des outils en os à Broederstroom est un produit de l’industrie agricole locale. Les résultats d’usure mettent en outre en évidence une gamme de pratiques d’activités différentes de la part des agriculteurs qui habitaient à Broederstroom.
Introduction
The authorship of bone tools in Early Iron Age (AD 200–900) sites in southern Africa has been a matter of speculation for some time, especially at sites where there is evidence of contact with autochthonous hunter-gatherers (Mason Citation1986; Mazel Citation1989). On the one hand, farming communities with access to iron are thought unlikely to have resorted to an ‘inferior’ material like bone to make tools. The presence of bone points, ostrich eggshell beads and stone tools in farmer sites is explained away as trade items, through mutual exchange relationships that at times existed between the two communities (e.g. Mazel Citation1989; Wadley Citation1996). On the other hand, there exist Early Iron Age sites with tangible evidence for on-site manufacture of bone tools (Whitelaw Citation1994). It is, however, debated whether these items were made by the farmers themselves or by hunter-gatherer ‘clients’ (Sadr Citation1997, Citation2002; Schoeman and Hay Citation2013).
Broederstroom is one such Early Iron Age farmer site in the Magaliesberg region of South Africa () where a number of bone points and other hunter-gatherer paraphernalia have been recovered. The site lies less than 10 km from, and almost within sight of, Jubilee Shelter, a Later Stone Age (LSA) summer aggregation camp that was occupied contemporaneously. Wadley (Citation1996) noted the similarity in stone tools between these two sites but did not comment on the bone tools. Mason (Citation1986, Citation1988) noted a congruence in overall form of the bone tools, as well as a similarity in their incised designs compared to those recovered from Kruger Cave, Mapungubwe and Olifantspoort, and considered the bone tools at Broederstroom to be of hunter-gatherer authorship.
Figure 1. Map of the Magaliesberg Biosphere Reserve, South Africa, adapted after Carruthers (Citation2019), showing the location of sites mentioned in the text. Core areas of the Reserve are shown in dark green, conservancies in light green and the transitional zone in yellow.
The mere presence of a particular item is, however, insufficient to establish authorship (see Mitchell Citation2003). For that we need to understand the manufacturing technology and whether these items were used in similar or different ways. Use-wear analysis on stone scrapers from Jubilee Shelter indicates that they were used primarily to work hides (Binneman Citation1984) and an analogous function has been imputed to the stone scrapers at Broederstroom (Wadley Citation1996). Further north, however, recent studies have demonstrated that stone scrapers changed from being principally hide-working implements to tools used in the manufacture of bone tools (Sherwood and Forssman Citation2023). Microwear analysis of worked bones from several LSA and Middle Iron Age sites has shown that these tools were used for diverse functions (Bradfield Citation2014, 2016; Antonites et al. Citation2016). However, no actual use-wear studies have yet been conducted on any material from Broederstroom.
My aim in this article is therefore to present a close analysis of the worked bone assemblage from Broederstroom and to assess the likelihood that they were made by local farmers or whether they could have entered the site through barter, or by other means, with hunter-gatherers occupying nearby sites such as Jubilee Shelter. I describe and compare the manufacturing techniques and microwear on the bone tools from Broederstroom to those from other sites in the region.
Background to Broederstroom
Discovered in 1973, Broederstroom was excavated intermittently between the late 1970s and the early 1990s (Mason Citation1974, Citation1981; Huffman Citation1993). One of the oldest Iron Age sites in South Africa (AD 350–650), and the most well-preserved at the time of its discovery, Broederstroom has played an important role in debates about the social importance of cattle in early farming societies (cf. Huffman Citation1990, Citation1993; Badenhorst Citation2009) and the extent to which there was material culture transmission between the new arrivals and autochthonous hunter-gatherers (Mason Citation1986, Citation1988; Wadley Citation1996).
About AD 350, the first southwardly moving farmers settled in the Magaliesberg valley (Huffman Citation2007; Carruthers Citation2019; Carruthers and Carruthers Citation2024), an area that had already been occupied for millennia by Stone Age hunter-gatherers at sites like Jubilee Shelter (Wadley Citation1987, Citation1989). Attracted by the temperate climate and rich soils, these migrants brought with them herds of domestic sheep, goats and cattle. Their subsistence, however, was mainly derived from the abundance of nutritional wild plants and game hunting (Mason Citation1981), although circumstantial evidence in the form of several grindstones and three underground dung-lined storage pits hints at the processing of cultivated cereals like sorghum (Sorghum bicolor) and pearl millet (Cenchrus americanus) (Huffman Citation1993). The bin platforms consist of circular arrangements of large stones that were used to support large, elevated baskets in which unthreshed grain may have been stored (Huffman Citation1993, Citation2007). Archaeological excavations revealed the remains of 49 hut floors, 36 grain bin platforms and six furnaces spread over a 25-ha area (Mason Citation1974, Citation1981; Huffman Citation1993). At its height, Broederstroom is thought to have housed 200–300 people.
Broederstroom is one of only three first-millennium AD sites in the Magaliesberg that has evidence of metal-working, the others being Uitkomst and Olifantspoort (Friede Citation1977, Citation1983a). The furnace remains at Broederstroom, which date to the fifth century, are all poorly preserved, like all other Early Iron Age furnaces in South Africa, which makes it impossible to identify their type and whether they were used for smithing or smelting (Friede Citation1983a). Whereas villagers sourced low-grade iron ore locally from sedimentary ferrous outcrops, some copper artefacts have also been found that have been shown to be almost 100% pure (Friede Citation1977). It is suggested that the a copper chain and pair of earrings may have derived from established trade networks with pre-Islamic traders on the Indian Ocean coast, a conclusion tenuously supported by the presence at the site of seashells (Moffett Citation2020), although they are not thought to have reached as far south as Sofala in southern Mozambique until the seventh century AD (Le Roux Citation2010; Sinclair et al. Citation2012). The possibility thus remains open that the copper at Broederstroom was locally mined and worked (Miller Citation1995; Huffman Citation2007).
The ceramics at Broederstroom belong to the Mzonjani facies, which is part of the Kwale branch of the Urewe Tradition (Huffman Citation2007). Thousands of potsherds have been found, representing jars and bowls, some with carinated profiles and everted rims (Mason Citation1974). The pottery is tempered with quartz and grog and was fired at temperatures estimated to have reached 1000°C (Friede Citation1983b). Ceramic figurines of animals, beads and medallions were also found. The latter may represent earlobe ornaments or spindle whorls used for cloth-making.
A large number of stone tools, grooved stones, bone arrowheads and ostrich eggshell beads were recovered from Broederstroom and point to the possibility that its inhabitants interacted and traded with hunter-gatherers visiting Jubilee Shelter or sites further afield such as Kruger Cave (Mason Citation1981; Wadley Citation1996). Mason (Citation1988) further notes the similarity in incised motifs on bone points from Kruger Cave, Broederstroom, Olieboom-spoort, Olifantspoort and several Middle Iron Age sites in the Limpopo Valley. Similar symbiotic relationships exist at some of the latter and hunter-gatherers appear to have manufactured bone tools for trade with farmers there (Sherwood and Forssman Citation2023) and in KwaZulu-Natal where the two communities existed in similar proximity (Mazel Citation1989). To date, except for a basic morphological description (Mason Citation1988), none of the bone tools from Broederstroom have been studied in any depth.
Today, Broederstroom is a declared provincial heritage site situated on land now belonging to the Tswane University of Technology. Virtually nothing of the archaeological site remains visible except scatters of slag, daga and pottery. Baboon activity has displaced most of the large stones that once formed part of the site’s grain bin and hut structures. The university’s plans to develop the property to generate third-stream income may place the Iron Age site at further risk of destruction. To mitigate this undesirable outcome, the site is now the subject of an ongoing remedial and public outreach project by the Magaliesberg Association of Cultural Heritage (Carruthers and Carruthers Citation2024).
Materials and methods
The assemblage described in this study comprises 110 pieces of worked bone recovered from excavations of the Broederstroom Early Iron Age site in the late 1970s through to the early 1980s. There is very little spatial resolution recorded specifically for the bone tools, but they all came from the top 25–30 cm of deposit from Mason’s main excavations, which revealed eight hut structures, three furnace remains and two burials (Mason Citation1981). Most of the pieces conform to a standard bone point morphology, that is they are longer than they are wide, tending to taper towards one or both ends, and are cylindrical to elliptical in cross-section. In some cases, the specimens are simply broken shafts of cortical bone that show signs of scraping or grinding and likely represent unfinished tool blanks. The worked bones are analysed as a single assemblage with no attempt to sub-divide them according to a morphological typology. Nor, since my focus is on the manufacturing technology and use-wear indicators, is it appropriate to create such an artificial distinction, which may be misleading as bone point morphology, at least in southern Africa, does not necessarily bear any relation to function (Bradfield Citation2016a).
A technological and microwear analysis was conducted using a dual microscopy approach that combined low- and high-power reflected light imaging and three-dimensional scanning of topographic surfaces. An Olympus BX51M, capable of various illumination options, including UV, dark- and bright-field, polarised and cross-polarised light and various fluorescent wavelengths, was used to visualise microwear details on the bone surface. Visual data were acquired and processed with Evident software. An Olympus DSX1000 3D scanning microscope was used for stereoscopic analysis of specimens and for three-dimensional scans of topographic features like cut marks. Three-dimensional data acquisitions were analysed using LEXT software. Observations were conducted at magnifications ranging from 20x to 200x, following standard microscopy protocols for bone tool use-wear (Bradfield Citation2015a, Citation2022, Citation2023). Macrofractures were identified using established criteria (Fischer et al. Citation1984; Bradfield Citation2012a), as were taphonomic modifications (Fisher Citation1995; Greenfield Citation1999; Reynard Citation2013; Fernández-Jalvo and Andrews Citation2016). Similarly, I referred to published literature (e.g. d’Errico and Henshilwood Citation2007; Christadou Citation2008; Orłowska Citation2016; Orłowska et al. Citation2022), as well as my own unpublished experimental referents, in respect of manufacturing wear.
Results
The results of the technological and microwear study are presented in and with the raw data given in . Of the 110 modified bones included in this study, only 32 were intact. The rest were either broken in antiquity or during excavation/curation. All breaks occurred when the bone was dry. Macrofractures consist primarily of snap, hinge and feather terminations, with three spin-off fractures recorded (). Spin-off fractures may be considered diagnostic of projectile impact but occur infrequently (Bradfield Citation2023). There is considerable variability in the length of intact bone points, ranging from 10 to 101 mm, with an average of 46.7 mm. This length is considered short compared to similar pieces from elsewhere in southern Africa (cf. Bradfield Citation2011, Citation2014). Many of the Broederstroom intact bones, however, had one or other of their apical ends modified, either through grinding it flat against an abrasive surface or through ring snapping, a technique commonly used to control the fracture propagation when shortening a length of bone. Width and thickness metrics are fairly standard across the sample.
Figure 2. A graphical representation of the morphometric results of this study. The raw data are presented in .
Figure 3. A graphical representation of the microwear results of this study. The raw data are presented in .
Table 1. Raw data table of modified bones from Broederstroom showing: maximum measurements of length, width and thickness; manufacturing technique; distal and proximal modifications; macro-fracture terminations; and the presence of taphonomic modifications, microwear and ancient residues. The microwear column should be understood as ‘consistent with’, rather than as a definitive interpretation (see Bradfield Citation2016b).
Most of the bone tools were manufactured by whittling and scraping, in most cases with a metal blade (). There is some indication that a cryptocrystalline stone tool may have been used in certain cases (). A suitable bone blank was reduced by whittling away protuberant edges of the bone, creating progressively shallower facets (), which might then be further reduced by continued scraping or grinding. In most cases, however, manufacturing did not proceed beyond the faceting stage, leading to the superficial impression that these items may represent unfinished pieces. This is not necessarily the case, as seen at other Iron Age sites where bone points were hafted in this faceted condition (Antonites et al. Citation2016). Be that as it may, there are certainly some examples of bone tools (e.g. #531, #2127, #3003 and #3004) that were clearly abandoned before completion either because the bone broke or for some other unidentified reason. Except in one case (#437), the bones were worked without the application of a lubricant, which can be identified by the presence of horizontal cracking overlying the scraping striations (cf. Orłowska et al. Citation2022). No chattermarks are present on any of the pieces, suggesting that the artisan was skilled in the task and knew how much pressure to apply when scraping. In almost all cases scraping runs parallel to the bone’s long axis.
Figure 4 Broederstroom manufacturing traces: a) experimental referents of the two manufacturing techniques used to fashion bone tools at the site, showing from left to right manufacturing striations that accrue through grinding in one direction against a piece of chert, scraping with a chert blade and scraping with a metal blade; b) examples of manufacturing traces on three bone tools of the site, showing scraping with a metal blade (#90), possible scraping with a retouched lithic blade (#3005) and scraping with a metal blade while the bone was wet (#437); c) macroscopic examples of faceting and scraping; d) macroscopic examples of grinding, ring-snaps and deliberately flattened ends. Note the incision on #1988, which is a failed ring-snap modification.
Grinding occurs on 21% of the sample. Generally speaking, the bone fragments that show signs of grinding are much shorter than those that are scraped, averaging 21.5 mm versus 43.6 mm, and also thinner, averaging 3.6 mm versus 4.9 mm (). Apart from their size, there do not appear to be any other distinctions between fragments that were ground and those that were scraped and in 12 instances both manufacturing techniques occur on the same specimen, with grinding invariably overlaying scraping. Grinding striations consistently run diagonally from the top left to the bottom right when the point is orientated with the distal (tapering) end at the top. This grinding orientation is the most common pattern in southern Africa and indicates manufacture by a right-handed person (Bradfield Citation2015b).
Apical modification is dominated by flattening through grinding against an abrasive surface. Flattening of the distal end of the bone point (i.e. the tapering end) is more prevalent than on the proximal end ( and ). Simple ring-snaps follow and are most prevalent on the proximal end. Whittling follows and may range from fairly crude, as in the case of #3004, to neat, conical finishings as can be seen in artefacts #410 and #1340 (). Incised lines are present as solitary circumferential incisions and in clusters of three on several bone points. shows the profiles of four incised lines. The butchery cut was clearly (although perhaps somewhat surprisingly) made with a stone blade, whereas the ring snap bears closer similarities to what we would expect from a dulled metal blade (cf. Greenfield Citation1999; Greenfield and Horwitz Citation2012; Jurkovičavá et al. Citation2018). The ‘decorative’ incisions appear to be intermediate between the two, although the pointed nadir and shoulder inclination profiles are more consistent with stone blades. All the circumferential incisions were executed in two or three strokes of the blade rather than in a single application of pressure while rotating the bone.
Figure 5. Three-dimensional profile scans of various cut marks found on the Broederstroom worked bone assemblage. The top right is a cut mark from a piece of butchered bone, the top left an incision for a ring-snap and the bottom row two circumferentially engraved lines.
Taphonomic alterations are present on 29% (N = 32) of the sample, while microwear related to use or curation is present on 60% (N = 66) of specimens. Weathering and rodent gnawing are the most common type of taphonomic alteration affecting the Broederstroom bone tools, followed by trampling and sediment abrasion (). Most of the specimens are encrusted with a resinous sediment, which resembles poison (see Bradfield et al. Citation2015; Chaboo et al. Citation2019; Isaksson et al. Citation2023), but which without chemical analysis is impossible to identify with certainty.
Most microwear is indeterminate. This means that a bone tool was not used for a sufficient period of time to develop diagnostic traces or that the palimpsest of different traces renders recognition of a single one impossible. In the case of Broederstroom, most indeterminate identifications are the result of weak polish formation. It should be noted here that all use-wear attributions must be understood as representing the most likely agent responsible for their formation and not as definitive diagnoses, the problem of equifinality being omnipresent in use-wear studies. Cleaning wear occurs on 32% (N = 35) of the bone tools (). This polish presents as clustered stripes of longitudinal striations, equifinal with many other modifying agents, but which diagnostically often overlies the sediment/poison adhering to the bone (). Ethnohistoric examples of poisoned arrowheads (for example those collected from various groups in northern Namibia by Louis Fourie between 1916 and 1926) do not exhibit similar polish, whereas experimental pieces rubbed with a cotton cloth exhibit almost identical traces, both in terms of form and placement (). This cleaning wear must have occurred when the bones were being prepared to have the accession number inscribed on them, as the ink lies above the polish in every instance.
Figure 6. Examples of polish associated with cleaning wear. The top row shows an ethnohistoric poisoned bone arrowhead collected in 1926 with no polish, followed by two experimental examples showing polish that has developed as a result of rubbing with a course cotton dishcloth. The bottom row illustrates comparable polish on three bone tools from Broederstroom. All the micrographs were taken at 100x magnification.
Cleaning wear may in some cases appear similar to trampling damage (: top left) but the former is generally smoother, more rounded and has fainter micro-striations. Other use-wear categories include handling wear, soft skin and leather-working. Handling wear presents in varying degrees of development and does not occur in a consistent location, as does soft skin wear and leather wear, which most often occur at the pointed end. Microwear that is consistent with leather-working is often associated at Broederstroom with horizontally inclined micro-striations that would suggest a twisting motion (). These specimens were likely used as awls, perhaps as tools for clothing manufacture, or even, as in the case of #2502, as an accoutrement to a leather garment. Apart from the cleaning wear, all the other use-wear indicators consistently occur directly on the bone surface, underneath the resinous sediment accretions.
Figure 7. Examples of microwear indicators associated with trampling, handling and leather-working.
Twelve specimens have use-wear or residues that indicate the presence of plant material, in one form or another (; ). Plant-working was identified on a range of bone tool morphologies (ranging from classic squat bone points to minimally modified, broad peg-like pieces), highlighting the multifunctionality of bone tool forms and the inadequacy of relying on morphology to interpret function (Bradfield Citation2016a). In some cases, specimens are only very slightly modified, and these artefacts may be ad hoc tools or bone blanks that were wrapped in plant material during manufacture (#2418, #3004). In this case the plant material may have served to grip the bone during whittling. Five specimens preserve degraded residues of plant tissues (cf. Cooper and Nugent Citation2006). On items #66, #486 and #587 these residues present as numerous strips of generic parenchyma tissue that have been blackened through heat but retain a distinct cellular structure. The location of these tissues on the dorsal and ventral surfaces at the proximal end (that is, away from the pointed working end) and the banded placement of the individual strands suggests that they represent the remains of a plant material that was concentrically wrapped around the bone shaft, perhaps to serve as a handle or grip, or as a ligature onto which a handle would have attached. The blackened nature of the residues and the absence of any indications of heat exposure to the bone surfaces may suggest that the heat was generated locally through friction, possibly during use. All indicators suggest these residues are in situ. Except in one case, the plant residues are not associated with actual plant-working microwear. The plant-working microwear may be further divided into different types, including those consistent with basketry, hafting into a reed (or other Monocotyledon) collar and processing starchy tubers (; cf. Stone Citation2009; Bradfield and Antonites Citation2018: Fig. 5; Desmond Citation2022). They all present as bright areas of flattened micro-topography with well-defined directional striations (). The polish usually intensifies in one direction but does not generally cover the entire tool. In the case of artefact #238, there is well-developed plant microwear over the entire specimen, which, since it is broken, could indicate that this piece is the remains of a handle that was once covered in a plant-fibre sheath.
Figure 8. Examples of bone tools with various plant-working indicators. This selection is intended to show the variety of tool morphologies on which plant-working is implicated. Selected micrographs show polish indicative of wrapping in a fibrous plant material, possibly as a handle or grip (#238); unspecific plant wear located underneath long, blackened strips of parenchyma tissue (#446); and smooth bright polish consistent with basketry production (#532) The micrographs were taken at 100x magnification.
Discussion and conclusion
One of the questions posed by Mason and other early researchers was whether the bone tools found at Early Iron Age sites like Broederstroom represent local produce or items introduced through trade with hunter-gatherers. Mason (Citation1986) himself considered the bone points at Broederstroom to have been of hunter-gatherer authorship but locally manufactured. This hypothesis was supported by the presence of several grooved pieces of sandstone that Mason believed were used for fashioning ostrich eggshell beads and bone points, and of stone scrapers identical to those at Jubilee Shelter, which were brought to Broederstroom as finished products (Wadley Citation1996).
Jubilee Shelter, located on the other side of what is now the Hartbeespoort Dam, would have been the logical point of contact for such exchange networks. Three phases of contact have been recognised this site. The radiocarbon dates of these phases are here presented according to the SHCal20 calibration curve (see Hogg et al. Citation2020), listing only the range of the largest intercept and not the full 95.4% range (Lombard et al. 2022). The earliest phase, from cal. AD 116–364, is characterised by a wide faunal representation, including many large game species. This number decreases dramatically during the second phase of contact (cal. AD 445–640), which coincides with the height of the Broederstroom farmer occupation. Bone points are also scarce at this time, although there are still many stone scrapers. After cal. AD 699, the largescale settlement of the area by farmers led to the impoverishment of the hunter-gatherer material culture at Jubilee Shelter. Overall, the numbers of large game represented in the faunal record decrease through time, concomitant with bone points, of which 150 have been recovered from the site. It is thought that the meat needs of the occupants of nearby Broederstroom clashed with those of the hunter-gatherers and resulted in diminished game availability. The surviving hunter-gatherers would thus have had to alter their hunting strategies to incorporate smaller, trappable prey (Wadley Citation1987, Citation1989, Citation1996).
A study undertaken in 2019 to establish whether people were selecting the same or different animals from which to make their bone tools proved inconclusive in respect of the Jubilee sample (Bradfield et al. Citation2019). At Broederstroom, however, a range of bovids is implicated, including species that are not represented in the much larger unmodified fauna sample. Although that study concluded that mechanical suitability was the primary determinate for species raw material selection, the sample size was small and the unique species alluded to may indicate some selective sampling strategy that may have been culturally determined. There is certainly far more work needed in this direction.
If we compare the technological and microwear results of the Broederstroom bone tools with those of the 118 pieces previously analysed from Jubilee Shelter (Bradfield Citation2012b) and the 36 from Kruger Cave (Bradfield Citation2014) some distinct differences emerge. The Jubilee specimens were all manufactured by grinding against an abrasive surface. There are no examples of scraping or whittling. Most of the Kruger Cave worked bones were diagonally ground, but there are examples of longitudinal scraping with a stone tool. Three bone points from Jubilee had stemmed butts, but these are unlike the faceted butts seen at Broederstroom in that they all have diagonal grinding overlaying the facets, whereas this feature is never the case at Broederstroom. Apical flattening is fairly uncommon at Jubilee compared to some Iron Age collections such as Broederstroom and Mapungubwe (cf. Antonites et al. Citation2016), although it does occur infrequently in contemporaneous levels at Kruger Cave (Bradfield Citation2014). The Jubilee and Kruger Cave bone tools are also far more fragmented than those at Broederstroom, with virtually no pointed tip sections present in the contact-period layers at Jubilee. The macro-fracture profile also differs, with diagnostic impact fractures occurring on 34% of Jubilee specimens and 10% of the Kruger Cave ones, compared to only 3.6% at Broederstroom. The incised markings on the Jubilee worked bones were not measured with three-dimensional scanning technology in my 2012 study, but I recollect that they appeared shallower than those at Broederstroom.
These differences, together with the fact that a metal blade is implicated in the shaping of many Broederstroom specimens, point to a clear technological distinction in the manufacturing of the Broederstroom and Jubilee bone tools, and to a slightly lesser extent to those at Kruger Cave. Additionally, the presence of ad hoc bone tools (bones that have been minimally modified but still used for a short duration, e.g. artefact #157), plus several pieces that have been clearly abandoned before completion, points to the probability that the Broederstroom bone tools were made on-site by the farmers themselves. This scenario would support the late Tom Huffman’s supposition that the ostrich eggshell beads were made by the farmers and not acquired through trade (personal communication cited in Wadley Citation1996). The grooved stones may well have been used in the manufacture of the eggshell beads, but are unlikely to have played a role in bone tool production for the simple reason that the few specimens that were ground had their striations orientated diagonally and not longitudinally as would have resulted had they been used for this purpose.
The faceting of many worked bones at Broederstroom shares certain affinities to those from K2 in the middle Limpopo Valley (Antonites et al. Citation2016), and to a lesser extent to worked bone from Kruger Cave (Bradfield Citation2014, Citation2015b). At both sites, the makers of the tools clearly considered their work finished despite not attaining the smooth cylindrical profile that characterises most bone points from Later Stone Age sites. This expediency is another feature distinguishing the bone tools at Broederstroom from those at Jubilee Shelter. The Kruger Cave specimens are intermediate between the Jubilee and Broederstroom examples. Diagonal grinding is the most common manufacturing method here, but longitudinal grinding is also evident, albeit always underlaying the diagonal striations (Bradfield Citation2014). Manufacturing techniques aside, the bone tools from Broederstroom seem to have been used for much the same tasks as at Jubilee Shelter and Kruger Cave. Putative arrowheads, linkshafts, awls, digging implements and tools used to work plant-based materials are all represented. These names are, of course, derived here from the use-wear indicators and not, as has traditionally been the case, from the gross morphology of the bone implement. This result aligns with what we know about the Early Iron Age economy at the site. On the other hand, the tentative identification of plant fibre handles has thus far been rarely observed in southern Africa.
With regard to the incised lines present on some of the specimens, Mason (Citation1988) drew attention to the similarity of these simple ‘motifs’ to incised bone points at other Iron Age sites further north in the Mapungubwe cultural landscape in the middle Limpopo Valley, as well as to contemporary LSA sites like Kruger Cave, Olieboomspoort and Jubilee Shelter. Twentieth-century Dobe San in the northwest Kalahari marked their arrows in a similar fashion to indicate ownership (Yellen Citation1977; Wiessner Citation1983), while at the Middle Iron Age site of Mapungubwe (AD 1220–1300) the simplicity and ubiquity of single line circumferential incisions suggests a notative device to indicate a place marker for something (Antonites et al. Citation2016). Mason hypothesised that these might be early-stage trance symbols, but I consider this option unlikely as incised decorations on bone objects in southern Africa seldom conform to the entoptic shapes that have been described in the context of trance visions and rock art (Lewis-Williams Citation2004).
In summary, the technological and microwear evidence suggests that the bone tools from Broederstroom were made locally by the farmers. Further such studies are needed at other Early Iron Age sites where bone tools occur in high numbers to ascertain whether the Broederstroom industry is anomalous.
Acknowledgments
I extend my thanks to Vincent and Jane Carruthers of the Magaliesberg Association for Culture and Heritage for inviting me to participate in the Toppieshoek heritage project, of which the Broederstroom archaeological site forms part. My thanks also to Thembi Russell of the University of the Witwatersrand’s Archaeology Collections for facilitating the loan of the bone material.
Additional information
Notes on contributors
Justin Bradfield
Justin Bradfield is an associate professor and co-ordinator of the microwear laboratory at the Palaeo-Research Institute, University of Johannesburg. His primary research focuses on traditional knowledge systems as they are expressed through the medium of organic materials, particularly bone.
References
- Antonites, A., Bradfield, J. and Forssman, T. 2016. “Technological, functional and contextual aspects of the K2 and Mapungubwe bone industries.” African Archaeological Review 33: 437–463. doi: 10.1007/s10437-016-9233-z
- Badenhorst, S. 2009. “Phytoliths and livestock dung at Early Iron Age sites in southern Africa.” South African Archaeological Bulletin 64: 45–50.
- Binneman, J.N.F. 1984. “Mapping and interpreting wear traces on stone implements: a case study from Boomplaas Cave.” In Frontiers: Southern African Archaeology Today, edited by M. Hall, G. Avery, D.M. Avery, M.L. Wilson and A.J.B. Humphreys, 143–151. Oxford: British Archaeological Reports.
- Bradfield, J. 2011. Macrofracture Analysis as a Method for Identifying Bone-tipped Weapons: A Functional Approach to the Recognition of Stone Age Hunting Points. Saarbrücken: Lambert Academic Publishing.
- Bradfield, J. 2012a. “A macrofracture analysis of bone tipped arrows from the Fourie collection of hunter-gatherer arrows from Namibia.” Antiquity 86: 1179–1191. doi: 10.1017/S0003598X0004833X
- Bradfield, J. 2012b. “A comparison of three Later Stone Age bone point assemblages.” South African Archaeological Bulletin 67: 32–43.
- Bradfield, J. 2014. “Pointed bone tool technology in southern Africa.” PhD diss., University of Johannesburg.
- Bradfield, J. 2015a. “Use-trace analysis on bone tools: a brief overview of four methodological approaches.” South African Archaeological Bulletin 70: 3–14.
- Bradfield, J. 2015b. “Pointed bone tool technology in southern Africa: results of use-trace analyses.” Southern African Humanities 27: 1–27.
- Bradfield, J. 2016a. “Bone point functional diversity: a cautionary tale from southern Africa.” In Osseous Projectile Weaponry: Towards an Understanding of Pleistocene Cultural Variability, edited by M. Langley, 31–40. Dordrecht: Springer.
- Bradfield, J. 2016b. “Use-trace epistemology and the logic of inference.” Lithic Technology 41: 293–303. doi: 10.1080/01977261.2016.1254360
- Bradfield, J. 2022. “Scales of analysis: the use of appropriate magnification in use-wear studies.” Archeologia 3: 5–16.
- Bradfield, J. 2023. “Bone tool analysis.” In Encyclopaedia of Archaeology, edited by T. Rehren and E. Nikita, 406–417. New York: Academic Press.
- Bradfield, J. and Antonites, A. 2018. “Bone scapula hoes from the Middle Iron Age, Limpopo Province, South Africa.” Quaternary International 472: 126–134. doi: 10.1016/j.quaint.2017.10.028
- Bradfield, J., Forssman, T., Spindler, L. and Antonites, A. 2019. “Identifying the animal species used to manufacture bone arrowheads in South Africa.” Anthropological and Archaeological Sciences 11: 2419–2434. doi: 10.1007/s12520-018-0688-5
- Bradfield, J., Wadley, L. and Lombard, M. 2015. “Southern African arrow poison recipes, their ingredients and implications for Stone Age archaeology.” Southern African Humanities 27: 29–66.
- Carruthers, V. 2019. Cradle of Life: The Story of the Magaliesberg and the Cradle of Humankind. Johannesburg: Penguin Random House.
- Carruthers, V. and Carruthers, J. 2024. The Heritage Treasures of Toppieshoek. Middleburg: Magaliesberg Association for Culture and Heritage.
- Chaboo, C., Hitchcock, R., Bradfield, J. and Wadley, L. 2019. “Beetle and plant arrow poisons of the San people of southern Africa.” In Toxicology in Antiquity, edited by P. Wexler, 11–72. Dordrecht: Elsevier.
- Christidou, R. 2008. “An application of micro-wear analysis to bone experimentally worked using bronze tools.” Journal of Archaeological Science 35: 733–751. doi: 10.1016/j.jas.2007.05.017
- Cooper, J. and Nugent, S. 2006. “Tools on the surface: residue and use-wear analyses of stone artefacts from Camooweal, northwest Queensland.” In Archaeological Science Under a Microscope: Studies in Residue and Ancient DNA Analysis in Honour of Thomas H. Loy, edited by M. Haslam, G. Robertson, A. Crowther, S. Nugent and L. Kirkwood, 207–227. Canberra: ANU E-Press.
- d’Errico, F. and Henshilwood, C.S. 2007. “Additional evidence for bone technology in the southern African Middle Stone Age.” Journal of Human Evolution 52: 142–163. doi: 10.1016/j.jhevol.2006.08.003
- Desmond, A. 2022. “Bone tool proxy evidence for coiled basketry production in the North African Palaeolithic.” Journal of African Archaeology 20: 156–175. doi: 10.1163/21915784-bja10018
- Fernández-Jalvo, Y. and Andrews, P. 2016. Atlas of Taphonomic Identifications: 1001+ Images of Fossil and Recent Mammal Bone Modification. Dordrecht: Springer.
- Fischer, A., Vemming Jansen, P. and Rasmussen, P. 1984. “Macro and micro wear traces on lithic projectile points: experimental results and prehistoric examples.” Journal of Danish Archaeology 3: 19–46. doi: 10.1080/0108464X.1984.10589910
- Fisher, J. 1995. “Bone surface modifications in zooarchaeology.” Journal of Archaeological Method and Theory 2: 7–68 doi: 10.1007/BF02228434
- Friede, H.M. 1977. “Iron Age metal working in the Magaliesberg area.” Journal of the South African Institute of Mining and Metallurgy 77: 224–232.
- Friede, H.M. 1983a. Typology of Metal-Smelting Furnaces from Iron Age South Africa. Johannesburg: University of the Witwatersrand.
- Friede, H.M. 1983b. Lab Investigations of Transvaal Iron Age Pottery. Johannesburg: University of the Witwatersrand.
- Greenfield, H. 1999. “The origins of metallurgy.” Journal of Archaeological Science 26: 797–808. doi: 10.1006/jasc.1998.0348
- Greenfield, H. and Horwitz, L. 2012. “Reconstructing animal-butchering technology: slicing cut marks from the submerged Pottery Neolithic site of Neve Yam, Israel.” In Bone for Tools — Tools for Bones: The Interplay between Objects and Objectives, edited by K. Seetha and B. Gravina, 53–63. Oxford: Oxbow Books.
- Hogg, A.G., Heaton, T.J., Hua, Q., Palmer, J., Turney, C.S., Southon, J., Bayliss, A., Blackwell, P., Boswijk, G. and Bronk Ramsey, C. 2020. “SHCal20 Southern Hemisphere calibration, 0–55,000 years cal BP.” Radiocarbon 62: 759–778. doi: 10.1017/RDC.2020.59
- Huffman, T.N. 1990. “Broederstroom and the origins of cattle-keeping in southern Africa.” African Studies 49: 1–12. doi: 10.1080/00020189008707724
- Huffman, T.N. 1993. “Broederstroom and the Central Cattle Pattern.” South African Journal of Science 89: 220–226.
- Huffman, T.N. 2007. Handbook to the Iron Age: The Archaeology of Pre-Colonial Farming Communities in Southern Africa. Pietermaritzburg: University of KwaZulu-Natal Press.
- Isaksson, S., Lombard, M., Hogberg, A. and Bradfield, J. 2023. “Potential biomarkers for southern African hunter-gatherer arrow poisons applied to ethno-historical and archaeological samples.” Scientific Reports 13: 11877. doi: 10.1038/s41598-023-38735-0
- Jurkovičová, L., Sázelová, S., Hromadová, B., Polanská, M. and Komoróczy, B. 2018. “Morphoscopical approach in material agent analysis: cut marks produced by stone and metal tools from early Bronze Age site Pashlávky (Czech Republic).” Quaternary International 472: 97–107. doi: 10.1016/j.quaint.2018.01.022
- Le Roux, M. 2010. “Pre-Islamic Arabic trade and religion on the east coast of Africa”. Ekklesiastikos Pharos 92: 88–108.
- Lewis-Williams, J.D. 2004. Discovering Southern African Rock Art. Johannesburg: David Philip.
- Lombard, M., Bradfield, J., Caruana, M.V., Makhubela, T., Dusseldorp, G., Kramers, J. and Wurz, S. 2020. “The southern African Stone Age sequence updated (II).” South African Archaeological Bulletin 77: 172–212.
- Mason, R.J. 1974. “Background to the Transvaal Iron Age — new discoveries at Olifantspoort and Broederstroom.” Journal of the South African Institute of Mining and Metallurgy 74: 211–216.
- Mason, R.J. 1981. “Early Iron Age settlement at Broederstroom 24/73, Transvaal, South Africa”. South African Journal of Science 77: 401–416.
- Mason, R.J. 1986. Origins of Black People of Johannesburg and the Southwestern Central Transvaal AD 350–1880. Johannesburg: University of the Witwatersrand Press.
- Mason, R.J. 1988. Kruger Cave Late Stone Age, Magaliesberg. Johannesburg: University of the Witwatersrand Press.
- Mazel, A.D. 1989. “People making history.: the last ten thousand years of hunter-gatherer communities in the Thukela Basin.” Natal Museum Journal of Humanities 1: 96–159.
- Miller, D.E. 1995. “2000 years of indigenous mining and metallurgy in southern Africa — a review.” South African Journal of Geology 98: 232–238.
- Mitchell, P.J. 2003. “Anyone for hxaro? Thoughts on the theory and practice of exchange in southern African Later Stone Age archaeology.” In Researching Africa’s Past: New Perspectives from British Archaeology, edited by P.J. Mitchell, A. Haour and J.H. Hobart, 35–43. Oxford: School of Archaeology.
- Moffett, A. 2020. “The archaeology of metal production in southern Africa.” Oxford Research Encyclopaedia of African History https://doi.org/10.1093/acrefore/9780190277734.013.602 Site accessed 12 February 2024.
- Orłowska, J. 2016. “Reading osseous artefacts — an application of micro-wear analysis to experimentally worked bone materials.” In Close to the Bone: Current Studies in Bone Technologies, edited by S. Vitezović, 236–247. Belgrade: Institute of Archaeology.
- Orłowska, J., Ćwiek, M. and Osipowicz, G. 2022. “Was it ground? A closer look at various prehistoric bone grinding techniques — An experimental and traceological study.” Journal of Archaeological Science: Reports 46: 103675.
- Reynard, J. 2013. “Trampling in coastal sites: an experimental study on the effects of shell on bone in coastal sediment.” Quaternary International 330: 156–170. doi: 10.1016/j.quaint.2013.11.007
- Sadr, K. 1997. “Kalahari archaeology and the Bushman debate.” Current Anthropology 38: 104–112. doi: 10.1086/204590
- Sadr, K. 2002. “Encapsulated Bushmen in the archaeology of Thamaga.” In Ethnicity, Hunter-Gatherers, and the ‘Other’: Association or Assimilation in Africa, edited by S. Kent, 28–47. Washington: Smithsonian Institution Press.
- Schoeman, M.H. and Hay, M. 2013. Mapungubwe Reconsidered: Exploring Beyond the Rise and Decline of the Mapungubwe State. Pretoria: Mapungubwe Institute for Strategic Reflection.
- Sherwood, N. and Forssman, T. 2023. “Macro use-wear identifiers on lithic scrapers and behavioural shifts at Little Muck Shelter, SLCA.” Journal of Archaeological Science: Reports 49: 104034.
- Sinclair, P.J.J., Ekblom, A. and Wood, M. 2012. “Trade and society on the southeast African coast in the later first millennium AD: the case of Chibuene.” Antiquity 86: 723–737. doi: 10.1017/S0003598X00047876
- Stone, E. 2009. “Wear on Magdalenian bone tools: a new methodology for studying evidence of fibre industries.” In North European Symposium for Archaeological Textiles X, edited by E.B. Andersson Strand, M. Gleba, U. Mannering, C. Munholt and M. Ringaard, 225–232. Oxford: Oxbow Books.
- Wadley, L. 1987. Later Stone Age Hunters and Gatherers of the Southern Transvaal. Oxford: British Archaeological Reports.
- Wadley, L. 1989. “Legacies from the Later Stone Age.” South African Archaeological Society Goodwin Series 6: 42–53.
- Wadley, L. 1996. “Changes in the social relations of precolonial hunter–gatherers after agropastoralist contact: an example from the Magaliesberg, South Africa.” Journal of Anthropological Archaeology 15: 205–217. doi: 10.1006/jaar.1996.0008
- Whitelaw, G. 1994. “KwaGandaganda: settlement patterns in the Natal Early Iron Age.” Natal Museum Journal of Humanities 6: 1–64.
- Wiessner, P. 1983. “Style and social information in Kalahari San projectile points.” American Antiquity 48: 253–276. doi: 10.2307/280450
- Yellen, J.E. 1977. Archaeological Approaches to the Present. New York: Academic Press.