The Antiquity of Man

   Home   |   About   |   What is Evolution   |   World Archaeology & Palaeoanthropology   |   North African Archaeology   |   Pseudoscience   |   Recommended Readings   |   Reviews   |   Links   |   Search   |   Contact

Lyn Wadley. 1989. Legacies of the Later Stone Age. South African Archaeological Society Goodwin Series 6: 42-53

Abstract: Band aggregation and dispersal is a universal practice among hunter-gatherers. Modern San aggregation camps are characterized by socializing, ritual, formal behaviour and active manufacture of tools and gifts for hxaro exchange. Like their modern counterparts, Stone Age aggregation sites should contain evidence for gift manufacture and formal behaviour. In contrast, dispersal phase camps should contain expediently produced assemblages and few symbolic expressions. The contemporary mid-Holocene assemblages of Cave James and Jubilee Shelter, in the Transvaal [now the province of Gauteng in South Africa], differ considerably in their content. In Jubilee there are high frequencies of standardized, curated artefacts and I interpret this as an aggregation camp assemblage. On the other hand, the expediently produced artefacts at Cave James are interpreted as a dispersal camp assemblage. I suggest, further, that some of the temporal variability in Later Stone Age assemblages can be explained by the aggregation and hxaro models. Shifting aggregation centres and the difficulty of aggregating in harsh environments would have affected both the distribution and character of Stone Age sites.


South African Stone Age archaeology in the 1920s was concerned with the description and classification of stone tools and the interpretation of their spatial distribution. It is a tribute to the research of men like Goodwin and Van Riet Lowe that their stone tool classification and sequence, devised more than fifty years ago, has changed little. They attributed the changes observed in Later Stone Age artefacts to population movements of separate cultural groups, and to different regional distributions of raw materials. They were not, however, completely satisfied with their interpretations, and they believed that the major stumbling block to understanding the Stone Age was the small number of sites excavated and therefore a paucity of empirical data (Goodwin & Van Riet Lowe 1929: 188). This complaint remains the unspoken attitude of many archaeologists - if enough empirical information becomes available to us, the truth will reveal itself.

Goodwin and Van Riet Lowe's successors have not lacked dedication in amassing new empirical data. Frustration with the results from intuitive lithic studies, and a desire to compare assemblages from geographically separate sites, prompted archaeologists to experiment with analytical techniques ranging from statistical descriptions to edgewear analyses, using electron microscopes (Mason 1957; Deacon 1972, 1974, 1982; Binneman 1984). Precise descriptions have indeed been generated and answers have been found to a great many of the "how", "what", "when" and "where" questions. But the frustration is not alleviated because the data themselves do not provide information on the reasons for change either across time or through time. Unfortunately, prehistoric behaviour cannot be explained by describing more sites, stone tools or diets. Artefacts and food debris constitute the data to be explained, not the explanations themselves. The failure of stones to reveal the past, and answer the question "why", is sometimes seen as an intrinsic flaw in the stones, hence the disenchantment that many archaeologists have with lithic studies. But the stones are merely the most tangible symptom of a theoretical ailment. Archaeological problems can only be solved by asking questions within a relevant theoretical framework.

Clark (1959) was the first to suggest that different environmental conditions in southern Africa may have an effect on the character of stone tool assemblages, and this contributed to subsequent local interest in reconstructing environments, diets, settlement patterns, seasonality and prehistoric scheduling (Carter 1970; Deacon 1976; Parkington 1977; Yellen 1977; Klein 1980, 1981; Cable 1984; Mazel 1983; Wadley 1984). Deacon (1976), for example, suggested a model for the diffusion of technology between related linguistic groups, a model that incorporated the concept of technological change as an adjustment to ecological change.

These ecological studies have provided fascinating and informative data and they have become increasingly rigorous through the use of optimization models, such as those described by Winterhalder and Smith (1981). By using such models, the archaeologist can avoid accusations of environmental determinism and can marshall quantities of economic data. These data give information on the forces of production (Godelier 1978: 764) and thereby account for only part of human behaviour. Ecological models cannot inform us of cultural contexts or the role of prehistoric decision-making (Moore 1983: 175), nor can they inform us of social and ideological change in society. For these aspects of hunter-gatherer life we must turn to social models and specifically to social relations of production.

Social relations of production determine the social form of access to resources; they control the means of production and reproduction; they allocate and organize labour processes, and they determine the way in which resources will be redistributed (Godelier 1978: 763). The relations of production do not take the same form in each culture. The dominant social factor may be, for example, kinship, politics or religion (Godelier 1975: 14; 1977: 35) but, whatever it is, the dominant factor will determine rights of ownership, the right to live in an area and make use of its resources, the right to perform specific tasks and to claim the benefits of this work, and it will also determine the parameters for obtaining a spouse.

Archaeologists have always shied away from interpreting social organization in prehistory saying that the physical nature of archaeological data restricts heir interpretation to technology and function. Archaeologists should not, however, constrain their theory to suit the purported inductive limitations of their data. The challenge to archaeologists is to search for Stone Age relations of production and to identify the dominant factor within them.

I construct my own model using San ethnography. In so doing I do not wish to imply that southern Africa Stone Age people are merely fossilized San whose culture can be reconstructed through a game of "ethnographic snap". Nonetheless, whilst it is true that aspects of Stone Age life and modern San life must differ, there is convincing evidence that symbolic as well as technological continuities exist between the ancient and modern hunter-gatherers (Lewis-Williams 1984). A model derived from San ethnography is therefore preferable to any model derived from a western, eurocentric viewpoint. To ignore the wealth of ethnography about the living descendants of Stone Age people in southern Africa is the equivalent of trying to learn a language without opening the dictionary.

The Ethnographic Model

It is generally accepted that the San dominant social relationship is kinship, but I have argued in detail elsewhere (wadley 1986b, 1987) that kinshipitself is dominated by gender relations. The importance of gender relations has also been discussed by Mazel (1988 and this volume).

San gender relations are influenced by, and display their range of variation in, changing band configurations. San bands have fluid membership because of the practice of band dispersal and periodic aggregation.

Among modern San, the dispersal phase of band life is an informal, "private" phase (Silberbauer 1981). when households separate from the larger band and spouses work closely together at subsistence tasks. This is an energy conserving period when ritual, travelling and even subsistence work are mnimize (Lee 1979: 265; Silberbauer 1981: 202, 205, 209, 247). Manufacturing also wanes because gift exchange is rare in the dispersal phase. It would not be practical to store or transport gifts for later use and, apart from this, stock-piling of possessions is not part of the traditional San value system. Some of the social "rules" relating to the "normal" division of labour and formal gender behaviour may relax during dispersal. Men may, for example, help women to gather plant food and firewood even though these tasks are said to be female preserves (Marshall 1969: 97; Lee 1979).

Aggregation, on the other hand, is a "public" phase of band life (Silberbauer 1981) in the sense that kin-related households congregate at this time, making and exchanging gifts, marriage brokering and intensified ritual, such as initiations and trance curing dances (Lee 1979; Tanaka 1980; Silberbauer 1981). Aggregation is central to the maintainence of San social relations because it provides the means through which people gain access to social, religious and economic resources. Although small bands have modest nutritional needs and can sometimes feed themselves in territories with a low biomass, a viable band cannot remain permanently isolated. For social sustenance a hunter-gatherer band must remain close enough to others to allow periodic aggregations.

Although aggregation may last only a few weeks San regard this phase as the social norm and they derive great pleasure from the company of friends and kin. Of course, protracted close contact brings with itthe danger of inter-personal and even group conflict. The trance curing dance is one means of maintaining or restoring social equilibrium but tensions are sometimes averted by avoidance practices and the "tightening" of social rules (Marshall 1969; Lee 1979; Silberbauer 1981; Guenther 1986), particularly gender rules. One example is the insistence that there be a formal seating arrangement at the !Kung family hearth: the man's side is to the right of the observer facing the shelter entrance and the woman's side is to the left (Marshall 1976: 88, 249).

A form of delayed gift exchange, hxaro (Wiessner 1977, 1982, 1983, 1984), takes place in earnest during the aggregation phase and there is consequently a flow of valued artefacts, particularly bead ornaments and arrows.

Explanations for Temporal Variabilityin Later Stone Age Sites

Expanding the model
Since aggregation sites carry the widest range of activities, they seem best for detecting change in traditions. The character of an aggregation site depends on the rigidity with which social rules are applied, the intensity and frequency of ritual performances, and on the amount of tool and gift manufacture carried out there. These variables are, in turn, influenced by factors such as social and environmental stress.

Social rules tend to be rigidly applied as a universal reaction to stress caused by unpredicted short-term scarcity or local scarcity, or even intergroup tensions (Dirks 1980). In the initial stage of stress, people tend to be "hyperactive" and mobilize a broad spectrum of coping mechanisms, such as increased reciprocity, social interaction and particularly ritual, because the victims blame mystical interference. Only if these mechanisms fail will people withdraw into small groups and guard their remaining resources. The Yaps, for example, become concerned with rights and obligations in the wake of a hurricane; they closely observe taboo restrictions and carry out ritual to inhibit further disaster (Laughlin & d'Aquili 1979: 302). even marriage rules can be affected by stress: the Coaiquer Indians of Ecuador have relaxed marriage rules when land and resources are plentiful, but during bad time their preference for marrying parallel and cross-cousins is emphasized (Ehrenrich 1983: 59). San increase trance performance and shamanistic healing when they are stressed: the Nharo, for example, perform trance frequently during drought because they believe that such positive action will alleviate the crisis (Guenther 1986: 173).

Also of relevance here is Wiessner's observation of the !Kung reaction, in 1974, when the mongongo crop at /Xai/Xai failed, game dispersed and a plague left plant foods almost decimated. In August, the !Kung subsistence work effort decreased, instead people sat around talking and making handicrafts and eagerly gathering information from visitors about conditions in other areas. By mid-septemeber, families took their hxaro gifts and left /Xai/Xai to visit relatives in other areas (Wiessner 198277). The delivery of hxaro gifts is often stated as the pretext for visiting distant relations (Bleek & Lloyd 1911: 281, 283, 377; Tanaka 1980: 121; silberbauer 1981: 240; wiessner 1982, 1983, 1984) but the underlying economic motive must be strong when resources are depleted or when social tensions arise. !Kung women in the Tsumkwe area, who have become heavily reliant on their husbands' cash wages, are anxious about their own diminishing economic roles as plant gatherers and therefore make prlific numbers of headbands as gifts to safeguard their networks of reciprocity (Wiessner 1984: 198).

Socio-economic stress may affect not only the quantities of gifts made, but also their styles. Stress may accentuate style (Hodder 1979:450) and can therefore predict stylistic differences between hxaro items from times of plenty and those from times of stress. The widest variety of San headband styles is observable at Tsumkwe, where social change is rapid. Nonetheless, when people from Tsumkwe take gifts to ‘bush’ neighbours they select conservative designs, thereby demonstrating their respect for traditional values (Wiessner 1984:218). It is possible that, in the past, this same type of conservatism served to keep artefact designs intact.

Environmental stress may result in two different and opposing needs: first, the need to gain access to a neighbouring territory and, secondly, the need to keep neighbours out of one’s territory. These conflicting needs may influence the type of style that is emphasized.

Where it is possible to transfer membership from one band to another, hunter-gatherers will try to project a positive self-image to the hxaro partner by using individual styles. !Kung headbands arc good examples of this assertive style (Wiessner l 984:207-8). Unlike headbands, San metal arrows from within a single language group do not appear to exhibit many stylistic differences: Wiessner suggests that perhaps morphological standardization is important here because compatible styles encourage intergroup communication by playing down differences (Wiessner 1983:271-2). However, different San language groups have distinct arrows and these serve to mark and maintain boundaries (Wiessner 1983). This type of ‘group’ style may be emphasized together with increased territoriality.

Calling on reciprocal ties and moving in with relatives can, it seems, only work where there are local shortages; regional shortage may encourage territoriality and few reciprocal exchanges (Cashdan 1983:54). The abundance of resources decreases from the !Kung area in the well-watered north to the !Ko area in the south. Tightly-knit social units are most pronounced among the !Ko and the units become looser in membership as one moves north into !Kung territory. The !Ko display great territoriality: they rarely travel outside the territory of their own nexus and do not have outside marital and social ties. !Ko respond to scarcity by closing ties with outsiders (Cashdan 1983).

Regional shortages may influence not only the content of aggregation camps but also their distribution. The aggregation venue must be an area where resources are generous enough to support a large population for at least a few weeks. Although low biomass areas may have the potential to feed small bands, such areas may not be occupied unless they are contiguous with territories able to sustain periodic aggregations. Thus, aggregation "centres" may develop. These centres may have shifted many times during the Stone Age to accommodate regional deteriorations or ameliorations in the envirnment. Shifting aggregation centres, and thus demographic shifts, may explain some of the temporal and spatial variability observed in the Stone Age.

Band size may have fluctuated through time. During the Holocene bands seem to have been fairly small, but this may not have been the case in the later Pleistocene: H.J. Deacon suggests that the hunting of large migratory herd animals in the later Pleistocene may have correlated with large local group organization, low population density, large territorial range and the absence of fixed territorial boundaries (Deacon 1976:163). This type of organization might imply social relations quite different from those of small band societies. In large groups, spouses could be found within the group, the group could be ritually self-sufficient and there would be little need to make use of hxaro ties with distant groups. Hxaro might be weakly developed under these circumstances or might be completely absent.

The introduction of hxaro or increased popularity of hxaro may coincide with the breaking down of large groups into smaller bands. It would also seem the appropriate time for the practice of aggregation to be adopted.

I now attempt to use the ethnographic data on aggregation camps and reaction to stress to construct a predictivemodel for hxaro relationships. I should, for example, expect to find few hzxaro gifts where:
1. little visiting takes place, for example in a dispersal camp or in an area where band aggregations are absent;
2. people emphasize territoriality, perhaps because of regional resource shortage;
3. pioneer movements take place into unoccupied territory;
4. band size is large year-round, enabling marriage partners to be sought year-round within the band rather than only during aggregation phases (resources need to be abundant).

On the other hand, hxaro gifts might be numerous in a particular region when:
1. it becomes the aggregation centre of a region;
2. resource shortage is localized and it is possible to move in with relations in a better endowed area;
3. bands immigrate into the territory of others;
4. population is generally dispersed in small groups and bands are dependent on each other for spouses, ritual and social support;
5.aggregation phases are lengthy or frequent.

I now apply this model, together with the aggregation and dispersal model to each Later Stone Age industry in my research area, beginning with the most recent.

The post-Wilton: demise of the Transvaal hunter-gatherers

The pre-pottery, past-Wilton assemblage in Jubilee contains many bone points and shafts, beads and a variety of jewellery, together with the manufacturing debris from these artefacts . There are also many formal stone tools, particularly standardized, micrdithic scrapers made on a wide variety of imported siliceous rocks. I interpret this an an aggregation assemblage. Apart from Jubilee, there is only one other site in my Magaliesbcrg research area that might have been an aggregation camp. This is Rissik Estate, which is a large open site with a high proportion of formal tools and many cores. Further afield, Bushman Rock Sheller (Plug 1961) and Oliebomnpomt (Mason 1962) also look like aggregation sites. The Olieboompoort assemblage, dated 870 ± 50 BP, is remarkably similar to the post-Wilton at Jubilee.

Sites that I interpret as dispersal phase camps are, however, cammon. Cave James once again provides the best example of a casual stone assemblage but other Transvaal sites that look like dispersal phase camps include Fort Troje (500 ± 80 EP), Elizabeth Shelter, Kloofendal and possibly Hope Hill (Wadley 1987), Hennaps River Cave, Pietkloof Cave, Magabeng Shelter, Glenferness (Mason 1951; 1962) and Groenvlei Shelter (Malan & Van Niekerk 1955).

In the late Holacene hunter-gatherers moved into some shelters that were far smaller than any previously occupied, for example, Fort Troje and Kinofendal Shelter. Perhaps bands were much smaller than before because of regional environmental stress. Certainly, at several past-Wilton sites there is evidence for broadened subsistence bases because freshwater mussels, bullfrogs, lizards, hare, leguaan and crabs were harvested for the first time. This is a typical response by people suffering environmental stress (Cohen 1975). Strain on local resources may have been exacerbated after the arrival of stock-owners and at many southern African sites this arrival coincides with a broadening of the hunter-gatherer resource base (Deacon 1984). Micromainmalian evidence from Jubilee points to reduced grass cover perhaps caused by overgrazing in tbe area after 1840 BP (Avery 1987). The foragers’ scheduling cycle may have been disrupted so that they were obliged to make the most of local resources in small territories.

In Jubilee, there is a marked difference between assemblages associated with the early and later contacts with pottery makers. Initially the winter seasonality pattern remained intact and tbe artefact tradition was unchanged. After c. 1300 BP, however, formal tool-making declined and remains of spring and summer fruits were represented. Perhaps Jubilee and many of the tiny sites were the last refugia for hunter-gatherers in the Transvaal. With their mobility restricted the hunter-gatherers may have been prevented from aggregating in the traditional way. If people visited less often, there would be few hxazo partnerships and few hxaro gifts. Without aggregation, people would have been denied access to religious succour and social reproduction with its input of information, reciprocity and material exchange. Even physical reproduction may have declined because of the difficulty of obtaining spouses from outside the immediate household. All of these changes would have had deleterious repercussions on social relations of production. The final demise of the hunter-gatherers in the Transvaal may have come about, not so much through starvation as through social strangulation.

The Witton dilemma: cultural florescence or struggle for survival?

The pattern of mid-Holocene settlement seems restricted and the interior of South Africa may have been sparsely occupied. A contraction in the geographical spread of populations may be inferred from the low visibility of archaeological sites dated between 9000 and 4500 BP (Deacon 1974, 1982, 1984; Deacon & Thackeray 1984). Drier and warner mid-Holocene conditions in the interior of southern Africa may have caused low diversity and concentrations of resources, making the region marginal for habitatian (Deacon & Thackeray 1984). Evidence from the mid-Holocene levels of Wonderwerk (Humpbreys & Thackeray 1983) supports this inference.

None of the environmental data from Jubilee Shelter give convincing support for substantial change during the mid-Holocene but in Hope Hill, micromammalian remains tentatively suggest a more arid environment at 4400 BP (Wadicy & Turner 1987).

As I mentioned previously, the inability of the land to apport an aggregation phase may have forced people to move from their traditional territories. Perhaps this is why mid-Holocene people settled in the better watered coastal and escarpment areas of southern Africa. This may mean that some bands previously living in the interior emigrated to territories that were already occupied. As newcomers, tbey would have needed to set up fresh networks of gift reciprocity in order to gain access to resources. After the arrival of farmers in South Africa the option to move to new territory may have been impossible, but it was probably never easy at any time.

During the process of immigration, gift exchange should intensify in the host territory. There should be increased numbers of formal, standardized tools, arrows and beads, jewellery, art, bonework and leatberwork. This certainly is the case in the Wilton Industry of the Cape (Deacon 1982,1984), and also in the Wilton of Jubilee Shelter. Stone arrowheads (for example, segments) would have been ideal for expressing assertive style to woo a potential hram partner. Individual style could be expressed through the different use of rock type and colour. Such stylistic variations would not jeopardize the aerodynamic function of a stone arrowhead because standardization of shape and size could still be maintained.

The classic Wilton, with its formal, standardized artefacts looks like a Stone Age florescence. Perhaps nothing could be further from the truth and the Wilton may be a responce to difficult conditions when hunter-gatherers were expending a great deal of time creating alliances.

Territorial expansion in the Oakhurst

The technological changes that mark the beginning of the Oakhurst Industrial Complex, at about 12000 BP, coincide only broadly with the environmental amelioration at the end of the last glacial (Deacon 1982, 1984). There is, in fact, a lag of at least 2000 years between the environmental change and the technological change. The Oakhurst is found in a wide range of southern African environments, including Karoo scrub, grassland and bushveld, and accordingly, it would be imprudent to explain the technology as a direct response to specific environmental conditions. In the Cape, Oakhurst industries seem to coincide with a new emphasis on plant foods and the capture of small rather than large bavids, but in the Transvaal, there are some indications that the change in hunting pattern was not as drastic (Klein 1984).

In contrast to the mid-Holocene Wilton assemblages, the earlier Oakhurst assemblages are visually prominent on the landscape. The end of the Pleistocene may have witnessed a gradual expansion of people into different southern African habitats (Deacon & Thackeray 1984: 383). Indeed, Oakhurst tools are prolific, from the Matopos southwards, and the development is, as Goodwin and Van Let Lowe (1929) suggested, an essentially southern African development There are many excavated Oakhurst assemblages in the Transvaal: Magazine Shelter, Zevenfontein, Uitomst Cave, Kruger Cave, Glenferness, Hennops River Cave (Mason 1951, 1962, 1974) and Bushman Rock Shelter (Plug 1981). It is therefore no surprise to find a proliferation of Oakhurst sites in the Magaliesberg.

In Jubilee the Oakhurt is characterized by a low proportion of retouched tools. The retouch is dominated by scrapers and miscellaneous retouch on baked siltstone, diabase or hornfels. There are high frequencies of flakes: each bucket of excavated deposit contained an average of 198,9 flakes. Most flakes are siltstone and they were probably imported from open site quarries such as Silkaatsnek, Serpent Quarry or Xanadu because there are few siltstone cores in Jubilee. These large Oakhurst quarries do not appear to have been used as home bases; they contain vast quantities of quarrying and primazy mapping debris but only rare retouched or utilized pieces.

It is in the Oakhurst that worked bone and ostrich eggshell beads first appear in Jubilee Shelter and in Uitkomst Cave (Mason 1962:308). In Bushman Rock Shelter the Oakhurst contains a wealth of bonework and ostrich eggshell beadwork (Plug 1982). Some Oakhurst sites in the Cape seem to have particularly high frequencies of worked bone, especially bane paints: Nelson Bay is a good example (Deacon 1982, 1984) and at Boomplaas there are more polished bane tools at c. 12000 BP than at any other time (Deacon 1984). As a general southern African trend, it seems that bone points are farmore common in the Oakhurst and early post-Wilton than in the Wilton. Humphreys (1979) suggests that this is because arrows made of bone alternate in popularity with those of stone (segments). This trend can possibly be explained by demographic shifts and the need to send distinct sylistie messages to neighbours.

I have already mentioned that fine, stone arrowheads could be used to impress hxaro partners during the Wilton, which seems to be associated with demographic contraction. In contrast, both the Oakhurst and the early post-Wilton (pre-pottery) seem associated with demographic expansion. When bands move into new territory where land and resources are regionally plentiful, as may have happened during the Oakhurst, it is theoretically possible for them to move in with neighbours. Social rules for band membership might then emphasize inclusion rather than exclusion. People would want to play down the differences between themselves and their neighbours because this would make it easier to create ties of fictive kinship. Gifts exchanged between neighbours should reinforce this intention; because all bone arrowheads are similar they would make ideal gifts. The new technology and subsistence pattern in the Oakhurst seems to have been accompanied by social rearrangements that went further than mere demographic expansion. Bands seem to have formed into small groups with higher densities than was previously the case (Deacon 1976). Such groups would have been particularly vulnerable because small territories carry a great risk of resource failure. The survival of small groups depends on cordial inter-group relations and, therefore, the Oakhurst bands would have had considerable motivation to create gifts that fostered affiliations.

The shift to smaller groups during the Oakhurst may also have influenced the trend towards hunting small, browsing bovids in preference to the gregarious, large bovids that were hunted in greater numbers in the Pleistocene. Hunting stealthily in pairs, using bow and arrow, is socially quite different from large group hunting with spears. Driving game and finishing it off with spears, sticks or stones is a large group activity (Lee 1979:234) that does as depend on the element of surprise and can involve women and children.

When only a few men hunt with bow and arrow, this serves to to ‘privatize’ the hunt and its products. The two techniques imply different conventions for the redistribution of meat. Even today, the San have no speciflc rules for the division of large animals such as elephant that are hunted by large groups with spears; people just takewhat they want (Biesele 1975). In contrast, strict rules apply to the division of large antelope lahunted with bow and arrow(Marshall 1976). A change in social relations is implied when hunters change from being primarily spear hunters to being primarily arrow hunters because there is a change in the way that people gain access to resources. New gender relations may be part of the new social relations, and an issue to be researched further is the possibility that meat and women were simultaneously privatized. The role of wives as plant food gatherers, excluded from the hunt, may well have its origin at the time when arrow hunting became the generally accepted technique for obtaining meat. The invention of the arrow may, as Vinnicombe (1972:202) suggests, be a technological revolution but its general acceptance and widespread use may also have marked a social revolution.

In addition to bone paints, a number of other artefacts seem to become widely accepted between 12 000 HP and 8000 BP: ostrich eggshell beads, ostrich eggshell waterbottles, decorated ostrich eggshell and bore dstones (Deacon 1984:290-1).

The San belief in the shamanistic powers of the medicine man can also be traced back into the Oakhurst. Engraved art mobilier from Wonderwerk Cave dated to the early Holocene, has abstract engraved designs that are interpreted as trance hallucinations (Thackeray et al. 1981). Similar ‘ladder’, ‘chevron’ or grid designs are well known on open-air rock engravings, either on their own or in conjunction bth naturalistic animal (and occasionally human) depictions. The Transvaal open site rock engravings are possibly contemporaneous with the Oakhurst because there is a regular contiguity of Oakhurst tools and rock engravings (Goodwin & Van Riet Lowe 1929:155,172-5; Mason 1962:303) but it is not yet possible to date these engravings.

All these material representations give the Oakhurst and subsequent Stone Age industrial complexes a closer relationship to the modem San organization than any previous industrial complex The Oakhurst may therefore be the starting point far the type of social relations that we now recognize in modern San hunter-gatherer societies.

The Early Later StoneAge (ELSA): mistaken identity or different social relations

Oakhurst and ELSA sites are so different that it is not difficult to conclude that the makers of these distinct industries did not share the same social relations. Few ELSA sites have been found in southern Africa and, of those discovered, all are cave sites containing informal microlithic assemblages.

In Cave James the ELSA assemblage is unstandardized and contains few retouched pieces other than miscellaneous retouch. Flakes are abundant and average 760 per bucket of deposit; this may, in part, be eplained by the high percentage of quartz used. The Jubilee ELSA is almost indistinguishable from the Cave James ELSA (Table 2) and the informal nature of both assemblages may imply that both sites represent dispersal phase camps. The problem with this interpretation is that all of the presently recovered ELSA sites in southern Africa have informal assemblages and look like dispersal phase sites. In the Cape, the ELSA undergoes a technological change and becomes slightly si more formal after c. 15 000 BP when bladelets become characteristic of these assemblages (Deacon 1982, 1984). nonetheless, even these assemblages do not fit the aggregation model satisfactorily.

Three possibilities come to mind to eplain the general informality of the ELSA and its distribution in southern Africa. First, if the ELSA people had similar relations of production to those of the modem San, aggregation and dispersal must have taken place but the appropriate sites are mat being recognized. This may be the problem because no open sites that can be convincingly attributed to the ELSA have been found, and it would be curious if they did not exist.

The second possibility is that ELSA aggregation sites did not look like their later counterparts and that the model is presently not flexible enough in its definition. Artefacts later used for hxaro do occur in small numbers in some ELSA sites, but their use was clearly not widespread. Hxaro itself might be a Holocene innovation to facilitate aggregation and, thus, the third possibility is that the ELSApeople did not aggregate or disperse. Their relations of production were therefore different from those of Holocene people. The pattern of large, local group organization, law population density, large territorial range and weak territoriality suggested for the late Pleistocene in the Cape (Deacon 1976) may explain the lack of aggregation and dispersal. Where bands are large and widely separated, it might be possible and, indeed, more desirable to get a mate within the ‘home’ group than from a distant group. ‘This would explain the apparent lack of hxaro items in ELSA camps.

Fluctuating social relations

I have attempted to show that aggregation and hxaro exchange are critical to understanding modern San hunter-gatherer behaviour and probably also hunter-gatherer beaviour throughout most of the Later Stone Age. I have also suggested that kinship and gender relations dominated Later Stone Age relations of production from the time that the first Oakhurst industries appeared, and that a different set of social relations prevailed in the earlier ELSA industries. The complexity of material culture in the Holocene implies far greater sophistication of social organization than previously existed, but it is unlikely that the Oakhurst was the first or the only time in the Stone Age when a change in social relations took place. The Stone Age may have experienced several fluctuations in social relations in response to changing demographic, social and economic conditions. It is even possible that there were times during the Pleistocene when Holocene type conditions prevailed, thereby fostering social relations similar to those of the Holocene. Although there is not as much scope in hunter-gatherer societies as in agricultural and industrial societies for varying the dominant relations of production, it is none the less possible to obtain some variety by juxtaposing several potentially dominant elements. Even if only four of these are selected: physical strength, kinship, gender and religion, they can be ranked or combined in a number ofways to create different sets of social relations.

Clearly the aggregation model presents several problems that I have not been able to address here, and I outline some of these elsewhere (Wadley 1987:91). Notwithstanding these theoretical and technical shortcomings, I believe that the recognition of aggregation and hxaro as crucial elements in Stone Age life is an important threshold. Knowing the nature and the difficulty of the challenge is a stimulus for future research.

Acknowledgements: I thank B. Ward for figure 1 and I also wish to thank, again, all the people who were involved with my southern Transvaal project (Wadley 1987). I am also grateful to the HSRC and the University of the Witwatersrand for their financial support.


Avery, D. M. 1987. Micromammalian evidence for natural vegetation and the introduction of farming during the Holocene in the Magaliesberg, Transvaal. South African Journal of Science 83:221-5.

Biesele, M. A. 1975. Folklore and ritual of !Kung hunter-gatherers. Unpublished PhD thesis: Harvard University.

Binfard, L. R. 1983. Working at archaeology. New York: Academic Press.

Binneman, J. 1984. Mapping and interpreting wear traces on stone implements: a case study from Boomplaas Cave. In: Hall, M., Avery, G., Avery, D. M., Wilson, M. L. & Humphreys, A. J. B. (eds) Frontiers: southern African archaeology today: 143-51. Oxford: British Archaeological Reports International Series 207.

Bleek, W. H. I. & Lloyd, L. 1911. Specimens of Bushman folklore. London Allen.

Brooks, A. 1984. San land use patterns, past and present: implications for southern African prehistory. In Hall, M., Avery, G., Avery, D. M., Wilson, M. U & Humphrey; A. I. B. (eds) Frontiers: southern African archaeology today: 40-52. Oxford: British Archaeological Reports International Series 207.

Cable, C. 1984. A model for terminal Later Stone Age subsistence strategies in southern Natal. In: Hall, M., Avery G., Avery D. M,, Wilson, M. L & Humplireys, A. I. B. (eds) Frontiers: southern African archaeology today: 167-81. Oxford: British Archaeological Reports Tensional Series 207.

Carter, P. L. 1970. Late Stone Age exploitation patterns in southern Natal. South African Archaeological Bulletin 25: 55-58.

Cashdan, E 1983. Territoriality among human foragers: ecological models and an application to four Bushman groups. Current Anthropology 24:47-66.

Clark, J. D. 1959. The prehistory of southern Africa. Harmondsworth: Penguin.

Clark, J. D. 1970. The prehistory of Africa. London: Thames & Hudson.

Clark, J.D. 1977. Interpretations of prehistoric technology from ancient Egyptian and other sources. Part II: prehistoric arrow forms in Africa as shown by surviving examples of the traditional arrows of San Bushmen. Paleorient 3:127-50.

Cohen, M. N. 1975. Archaeological evidence for population pressure in preagricultural societies. American Antiquity 40:471-5.

Conkey, M. W. 1980. The identification of prehistoric hunter-gatherer aggregation sites: the case of Altamira. Current Anthropology 21:609-30.

Deacon H. J. 1976. Where hunters gathered: a study of Holocene Stone Age people in the eastern Cape. Claremont: South African Archaeological Society Monograph Series 1.

Deacon. H. J. & Thackeray, J. F. 1984. Late Pleistocene environmental changes and implications for the archaeological record in southern Africa. In: Vogel, J. C. (ed.) Late Cainozoic palaeoclimates of the Southern Hemisphere: 375-90. Rotterdam: Balkema.

Deacon, J. 1972. Wilton: an assessment after 50 years. South African Archaeological Bulletin 27: 10-45

Deacon, J. 1974. Patterning in the radiocarbon dates for the Wilton/Smithfield complex in southern Africa. South African Archaeological Bulletin 29: 3-18

Deacon, J. 1982. The Later Stone Age in the southern Cape, South Africa. PhD thesis: University of Cape Town.

Deacon, J. 1984. Later Stone Age people and their descendants in southern Africa. In Klein, R.G. (ed.) Southern African Prehistory and Palacoenvironments: 221-328. Rotterdam: Balkema.

Dirks, R. 1910. Social responses during severe food shortage and famine. Current Anthropology 21:21-32.

Ehrenreich, J. 1983. Reply to Cashdan, B. Territoriality in human foragers. Current Anthropology 24:58-9.

Godelier, M. 1975. Modes of production, kinship and demographic structures. In: Bloch, M. (ed.) Marxist analyses and social anthropology: 3-27. London: Malaby Press.

Godelier, M. 1977. Perspectives in Marxist anthropology. Cambridge: Cambridge University Press.

Godelier, M. 1978. Infrastructures, societies and history. Current Anthropology 19:763-71.

Goodwin. A. J. H. 1945. Some historical Bushmen arrows. South African Journal of Science 41:429-43.

Goodwin, A. J. H. & Van Riet Lowe, C. 1929. The Stone Age cultures of South Africa. Annals of the South African Museum 27:1-289.

Guenther, M. G. 1986. The Nharo Bushmen of Botswana. Hamburg: Helmut Buske Verlag.

Hodder, I. 1979. Economic and social stress and material culture. American Antiquity 44:446-54.

Humphreys, A. J. B, 1979. The Holocene sequence in the northern Cape and its position in the prehistory of South Africa. Unpublished PhD thesis: University of Cape Town.

Humphreys, A.J.B. & Thackeray, A. I. 1983. Ghaap and Gariep: Later Stone Age studies in the northern Cape. South African Archaeological Society Monograph Series 2.

Klein, R. G. 1980. Environmental and ecological implications of large mammals from Upper Pleistocene and Holocene sites in southern Africa. Annals of the South African Museum 81:223-83.

Klein, R. G. 1981. Stone Age predation on small African bovids. South African Archaeological Bulletin 36:55- 65.

Klein, R. G. 1984. The large mammals of southern Africa: Late Pliocene to recent. In Klein, R. G. (ed.) Southern African prehistory and palaeoclimates: 107-46. Rotterdam: Balkema.

Laughlin, C. D. & d' Aquili, E.G. 1979. Ritual and stress. In: d’ Aquili, E. 0., Laughlin, C. D. & MeManus, J. (eds) The spectrum of ritual: a biogenetic structural analysis: 280-317. New York: Columbia University Press.

Lee, R. B. 1979. The !Kung San: men, women and work in a foraging society. Cambridge: Cambridge University Press.

Lewis-Williams, J. D. 1984. Ideological continuities in prehistoric southern Africa: the evidence of rock art. In: Schrire, C. (ed.) Past and present in hunter-gatherer studies: 225-51. Orlando, florida: Academic Press.

Malan, B. D. & Van Niekerk, S. C. 1955. Die Later-Steentyd in Transvaal. South African Journal of Science 51:231-5.

Manhire, T., Parkington, J. & Van Rijssen, B. 1983. A distributional approach to the interpretation of rock art in the south western Cape. The South African Archaeological Society Goodwin Series 4:29-31

Marshall, L. 1969. The medicine dance of the !Kung Bushmen. Africa 39:347-81.

Marshall, L 1976. The !Kung of Nyae Nyae. Cambridge: Harvard University Press.

Mason, R.J. 1951. Excavation of four caves near Johannesburg. South African Archaeological Buletin 6:71-9.

Mason, R. J. 1957. The Transvaal Middle Stone Age and statistical analysis. South African Archaeological Bulletin 12:119-37.

Mason, R. J. 1962. The prehistory of the Tranavaul. Johannesburg: Witwaterstand Press.

Mason, R.J. 1974. Later Stone Age plant remains from Kruger Cave, Olifantsnek, Rustenburg District. South African Journal of Science 70:376.

Mazel, A. D. 1983. Eland, rhebuck and cranes: identifying seasonality in the paintings of the Drakensberg, The South African Archaeological Society Goodwin Series 4: 34-7.

Mazel, A. 1988. People making history: the last ten thousand years of hunter-gatherer communities in the Thukela Basin. Unpublished PhD thesis: University of Cape Town.

Moore, J.A. 1983. The trouble with know-it-alls: information as a social and ecological resource. In: Moore, I.A. & Keene, A. S. (eds) Archaeological hammers and theories: 173-92. New York: Academic Press.

Parkington, J. 1977. Follow the San. Unpublished PhD thesis: University of Cambridge.

Plug, I. 1981. Some research results on the late Pleistocene and early Holocene deposits of Bushman Rock Shelter, eastern Transvaal. South African Archaeological Bulletin 36:14-21.

Plug, I. 1982. Bone tools and shell, bone and ostrich eggshell beads from Bushman Rock Shelter (ERS), east- em Transvaal. South African Archaeolagical Bulletin 37:57-62.

Sampson, C. G. & Bowman, B. 1985. Variations in the size of arcbaealogicai surface sites attributed to the Seacaw River Hushmen. South African Journal of Science 81:321-21

Sliberbaner, 0. B. 1981. Hunter and habitat in the central Kalahari Desert. Cambridge: Cambridge University Press.

Tanaka,J. 1980. The San hunter-gatherers of the Kalabari. Tokyo: University of Tokyo Press.

Thackeray, A. I., Thackeray, 3. F., Beaumont, P. B. & Vogel, J. C. 1981. Dated rock engravings from Wonderwerk Cave, South Africa. Science 214:64-7.

Turner, G. 1986. Faimal remains from Jubilee Shelter, Transvaal. South African Archaeological Bulletin 41: 63-8.

Vinuicombe, P. 1972. Myth, motive and selection in southern African rack art. Africa 42:192-204.

Wadley, L 1984. Subsistence strategies in the Transvaal Stone Age: a preliminary model. In: Hall, M., Avery, G., Aver~ D. M., Wilson, M. & Humpheys, A. J. H. (eds) Frontiers: southern African archaeology toda)c 207-14. Oxford: British Archaeologicai Reports International Series 207.

Wadley, L. 1986a. Segments chime: a mid-Holocene Wilton site in the Transvaal. South African Archaeological Bulletin 41:54-62.

Wadley, L. 1986b. A social and ecological interpretation of the Later Stone Age in the southern TransvaaL Unpublished PhD thesis: University of the Witwalersrand.

Wadley, L. 1987. Later Stone Age hunters and gatherers of the southern Transvaal. Oxford: British Archaeological Reports International Series 380.

Wadley,L & Verhagen, B. Th. 1986. Filling the gap:a new chronology for the Holocene Stone Age of the Transvaal. South African Journal of Science 82:271.

Wadley, L. & Turner, G. 1987. Hope Hill Shelter: a Later Stone Age site in the southern Transvaal. South African Journal of Science 83:98-105.

Wiessner, P. 1977. Hxaro: a regional system at reciprocity for reducing risk among the !Kung San. Unpublished PhD thesis: University of Michigan.

Wiessner, P. 1982. Risk, reciprocity and social influences on !Kung San economies. In: Leacock, H. R. & Lee, R.B. (eds) Politics and history in band societies: 61-84. London: Cambridge University Press.

Wiessner, P. 1983. Style and social information in Kalahari San projectile points. American Antiquity 48:253-76.

Wiessner, P. 1984. Considering the behavioural basis for style: a case study among the Kalabari San. Journal of Anthropological Archaeology 3:190-234.

Winterbaider, B. & Smith, E. A. (eds) 1981. Hunter-gatherer foraging strategies: ethnographic and archaeological analysis. Chicago: University of Chicago Press.

Yellen, J. E. 1977. Archaeological approaches to the present. New York: Academic Press.

back to top of page