Gendered interpretations are not new to Stone Age archaeology. For many years, archaeologists uncritically labeled Stone Age man as “man the hunter,” “man the butcher,” “man the meat provider,” and “man the tool maker.” Since cultural development is often considered to be linked to technological development, the unspoken assumption was that man was the agent of change. This left woman as the passive bearer of children and collector of plant food, so it is not surprising that Stone Age woman went almost unnoticed in the archaeological record. She was often absent from archaeological reports except where she was mentioned in her capacity as “woman the plant food gatherer.” Such gender stereotypes illustrate an androcentric and ethnocentric view of gender roles; they trivialize the division of labor and, indeed, social relations. I shall show that these gender stereotypes are inappropriate when applied to modern hunter-gatherer people and that they are almost certainly inappropriate for people in the past.
The old gender assumptions have had a stifling effect on interpretation of social behavior. It is only since 1984, with the publishing of Conkey and Spector’s seminal paper, that self-conscious gender studies have been undertaken. Gender centers on social values invested in sexual differences; gender relations are an integral part of social theory (Gilchrist 1991:498) and cannot be ignored by anyone wishing to make social interpretations of the past. The new generation of gender studies has concentrated on looking afresh at old issues (Moore 1994:6–7), and new critiques are made from gendered perspectives without seeking gender attribution (Dobres 1995a).
The “man the hunter” interpretation of the Stone Age came about through the use of androcentric ethnographic accounts. In the 1970s Murdock and Provost (1973:207) claimed that hunting was one of the most obviously gender segregated activities in hunter-gatherer ethnography, and their argument was so persuasive that contradictory reports of gender equality in some hunter-gatherer societies were generally disregarded as aberrant or idiosyncratic. However, since the 1970s, reports of hunter-gatherer gender equity have increased and the stereotype of “man the hunter” has begun to be deconstructed. I begin to do so here by examining some of the ethnographically recorded variability in the way that meat is obtained. In so doing I am aware of the problems associated with the use of ethnographic analogy (see Gifford-Gonzalez, Chapter 7) but, due to space limitations, I concentrate on examples from southern Africa and draw on selected ethnographies from elsewhere to illustrate specific points about, for example, the organization of labor and exploitation of resources. Additional hunter-gatherer ethnographies from around the world could be drawn on to expand my arguments.
The best-known exception to the male-only hunting rule is that of the Agta women of the Philippines who are known to hunt occasionally with bow and arrow (Estioko-Griffin and Griffin 1981). Chipewyan and Tiwi women also hunt big game regularly and, like the Agta women, they use the same weapons as men (see Kent, Chapter 1). The Mbuti net hunters of Zaire include women in the hunt, yet men and women do not perform the same role. Men hold the nets and spear the prey while women drive the game and act as beaters (Ichikawa 1983:58). This cooperative form of hunting with a large number of personnel is highly productive (Ichikawa 1983:63). Women and children also take part as beaters in Begbe, an unpoisoned arrow hunt, and they drive the duiker in the same way as in the net hunt (Harako 1976). Begbe takes place only in the dry season when interband cooperation is observed.
Among the Inupiat of Alaska, women are seen as pivotal to the hunt and ritually attract animals, thus being classed as hunters (Bodenhorn 1990). One successful Inupiat hunter said, “I’m not the great hunter, my wife is!” (Bodenhorn 1990:61). He was referring to his wife’s ability to attract animals through her generous behavior rather than to her ability to shoot with a rifle (Bodenhorn 1990:62). Nonetheless, some Inupiat women are skilled with rifles, are good at paddling, and may take their place in whaling crews or on hunts with their spouses (Bodenhorn 1990:60). What is really important is that it is not necessary for a woman to accompany her husband on the hunt in order to be regarded as a hunter: women’s activities such as sewing, butchering, and meat sharing are all classed by Inupiat as hunting skills (Bodenhorn 1990:65).
Basarwa women at Kutse, in Botswana, may also accompany their husbands on the hunt and, conversely, men may spend much time gathering plant food with their wives or collecting firewood or water (Kent 1995:521; 1996b:130). Kutse women and men manufacture snares, and meat caught in these belongs to the person who made or owns the snare (Kent 1995:518). Bow and arrow hunting and spear hunting, on the other hand, are restricted to men (Kent 1995:518), though, unlike some other Basarwa groups, there is no restriction on a woman touching a bow and arrow (providing that she is not menstruating) or using a man’s knife or spear for cutting (Kent 1995:520). Furthermore, women occasionally use digging sticks to kill small animals and will check their husband’s traps when men are away from the settlement. Women regularly butcher small animals and may also butcher animals such as springbok when men are busy (Kent 1995:520). Thus the concept of “man the hunter” and “woman the gatherer” has a fuzzy boundary at Kutse (Kent 1995:520). The majority of Kutse activities are not gender linked; therefore, most tools are not associated with only one sex: digging sticks and spears, for example, are not exclusively women’s and men’s tools, respectively. Although men use spears most often for hunting, women may use spears for stirring food and as knives (Kent 1995:525).
The Ju/’hoansi (previously known as !Kung) of the Kalahari use a wider range of meat-getting techniques than the Basarwa at Kutse: bow and poisoned arrow hunting, large-group spear hunting on foot, spear hunting with the use of pit-traps, spear hunting on horseback, snaring, running down small game, and scavenging game killed by lions. Most Ju/’hoansi bow and arrow hunting, spear hunting, and hunting by snares takes place in the dry/hot and dry/cold seasons, and hunted meat is scarce at other times of year (Wilmsen and Durham 1988:63). Rain and dew fall from November to May, the wet season, and the dampness causes the sinew parts of bows, arrows, and spears, as well as the cordage of snare lines, to malfunction (Lee 1979:208; Wilmsen and Durham 1988:70, 72). In addition, prey are dispersed and spoor obscured during this wet season (Wilmsen and Durham 1988:72).
Ju/’hoansi archers hunt successfully either in small groups or as individuals because hunting with a poisoned arrow relies on stealth, surprise, and a good aim. Ju/’hoansi women are specifically barred from bow and arrow hunts and they do not participate in butchering animals (Draper 1975b; Marshall 1976:97, 287). Though they are not present at the hunt, Ju/’hoansi women contribute considerably to hunting success because they observe game movements and the condition of the veld when they collect plant foods, and they supply hunters with this valuable information (Draper 1975b).
Spear hunting on foot by large groups of hunters is no longer practiced by Ju/’hoansi, though Lee (1979:234) described an elephant hunt as it was recounted to him:
They set grass fires on one side and the people came in on the other side. The dogs worry it, then when it raises its ears the people throw in their spears one after another. They didn’t put poison on their spears. They just gathered many men together to throw their spears.
Lee comments that the spear hunt on foot requires the coordination of many people and suggests that this is no longer possible among 20th-century Ju/’hoansi.
The use of pit-traps was observed ethnographically at !Gi in Botswana (Brooks 1984). Animals caught in the traps were dispatched with spears. The pit-trap and spear technique may have been widespread earlier this century because it was also used by the Batwa, of Lake Chrissie in Mpumalanga, who no longer exist as a distinct cultural group (Potgieter 1955:22). The hunters excavated pit-traps for game, and then formed rows in the veld and frightened approaching game driven there by other band members.
Modern conditions have brought changes to Ju/’hoansi hunting patterns. Traditional bow and arrow hunting, for example, has been largely replaced by spear hunting where hunters have obtained horses. Ju/’hoansi men of /Xai/Xai (Cae Cae) now ride horses and throw metal-tipped spears for eland hunting and use dogs and spears for warthog and carnivore hunting (Wilmsen 1989:230–231), as do the ≠Kade of the central Kalahari (Osaki 1984). These modern spear hunts, which are viable even for solitary hunters, are efficient because the equestrian hunter hunts half as frequently as the archer on foot yet obtains almost two and a half times as much total return (Wilmsen and Durham 1988:78; Wilmsen 1989:252). The ≠Kade use the new hunting techniques to facilitate changing social relations. Meat ownership has moved from the owner of the weapon to the owner of the horse, thereby enhancing the wealth and status of those men with access to a cash economy. ≠Kade men sell some of their hunted meat, while the horse owners have the potential to become a wealthy elite relative to the non-horse-owning ≠Kade. The ≠Kade changes are linked to circumstances brought about by the demographic shift to a massive settlement of some 500 people. The ≠Kade demographic change was, in turn, brought about through contact with pastoralists and villagers, and also through the availability of store-bought foods and a regular supply of borehole water.
Snaring of animals is also a highly productive means of obtaining meat. Rope snares, made of plant fiber, are most often used by Ju/’hoansi for catching antelope and game birds. A noose is made from the rope and this is attached to a sapling that acts as a spring. The device is activated by a wooden trigger (Lee 1979:142). Most of the steenbok, duiker, and birds captured by Ju/’hoansi are taken in snares (Wilmsen and Durham 1988:71) that are most often set by old men and young boys whose mobility is limited (Lee 1979:207). Nisa, the unusual Ju/’hoansi female interviewed by Shostak, reported that Ju/’hoansi women and girls set snares (Shostak 1981:91) and that she (Nisa) had also run down and killed animals such as steenbok and small kudu (Shostak 1981:94, 102). Nisa was, however, an atypical Ju/’hoansi woman and should perhaps not be considered representative of her group (Kent, Chapter 1). Although Ju/’hoansi women and girls often collect tortoise, snakes, lizards, and birds (Lee 1979:235), they are not reported to have set snares by anyone other than Nisa.
Scavenging of animals killed by lion or other predators is also an important meat source (Wilmsen and Durham 1988:71). Women spend much time in the veld, and they sometimes come across dead animals and alert the other camp members to their find (Shostak 1981:93, 101).
This brief review of selected ethnography confirms that the stereotype of “man the hunter” and “man the meat provider” is flawed. It is true that men generally reserve the right to handle weapons (though Kent [1995] observed a Kutse woman firing her husband’s bow at a springbok that wandered into camp), but it is not universally true that women are excluded from playing an active role in the hunt or the meat quest. In the Mbuti example, women are indispensable to the hunt even though they do not deal the death blows. In the Kalahari example we see that the regular meat supply is provided by methods other than hunting with weapons, that is, by snaring and collecting, and we find that it is often old men, children, and women who bring this meat back to camp. Women at Kutse set traps for small game and collect small creatures. Thus Basarwa men are not the only meat providers and, by the same token, Basarwa women are not the only plant food providers. Out of 116 person-days of gathering, 21.5 were performed by Ju/’hoansi men (Lee 1979:262).
Among the Inupiat, women are regarded as more important to the hunting process than men, even where their contribution is restricted to sewing items that are needed for the hunt. Given this wealth of evidence it is clear that the interpretation of women as nonhunters or as exclusively plant providers is largely a western construct and has little to do with hunter-gatherer perceptions. There is a tendency for westerners to define hunting as exclusively the killing of large game with weapons such as bow and arrow or spear, but many hunter-gatherer societies regard the meat quest in a broader light (Casey, Chapter 5; Kent, Chapters 1, 3).
The hunter-gatherer ethnography described here displays considerable variability in the details of gender relationships, and it is clear that change has occurred this century, not only in hunting techniques employed, but also in the organization of labor. While the modern hunter-gatherer change has been fueled by, for example, contact with pastoralists, farmers, white settlers, and merchants, there is no reason to believe that change and variability were not also features of Stone Age life. Indeed, there may even have been a greater range of gender relationships in the past than there is at present (Paynter 1989:385); and it is possible that social complexity was greater in the Stone Age before hunter-gatherer societies became marginalized and fragmented (Schrire 1984b).
Southern African Stone Age people seem to have hunted more large than small game during the Pleistocene because sites dated to this time period have high frequency percentages of bone from large, gregarious, mobile grazers. The hunters would thus have been tied to the seasonal migrations of plains game whose movements would have been controlled by the quality of the grazing (Deacon 1976:116). In contrast, many sites post-dating 10,000 BP, particularly in the Eastern Cape, are dominated by small, shy, nocturnal, territorial browsers that can be snared more effectively than hunted (Deacon 1984:256). In addition, the more recent sites have a wider range of prey than before, including birds and dangerous animals. Melkhout-boom and Boomplaas caves (Figure 4.1) are good examples of the change from grazing to browsing antelope: grazers such as hartebeest, wildebeest, and buffalo, and mixed feeders such as eland, were hunted extensively before 10,000 BP, but thereafter proportions of large grazers declined in favor of small antelope such as grysbok, steenbok, and duiker (Deacon 1976; 1984).
Figure 4.1. Southern African sites mentioned in the text. BC: Border Cave; BP: Boomplaas; E: Edgehill; JS: Jubilee Shelter; KC: Kruger Cave; KRM: Klasies River Mouth; MHB: Melkhoutboom; RCC: Rose Cottage Cave; W: Wilton; WE: Welgeluk.
It is generally assumed that big-game hunters who used a Middle Stone Age (MSA) technology (predating about 22,000 BP) used spears that were either stone-tipped or made of wood or bone. Triangular flakes and unifacially or bifacially retouched points are thought to have been hafted for use as spearheads. By about 22,000 BP all traces of MSA industries disappeared throughout southern Africa (Wadley 1993), and they were succeeded by Later Stone Age (LSA) assemblages, some of which were microlithic and bladelet-rich. Although no LSA tools look like spearheads, it is possible that the people of the bladelet-rich Robberg Industry hafted bladelets to form composite barbed spearheads (Mitchell 1988) and that spear hunting continued unabated. Between about 12,000 BP and 11,000 BP the microlithic Robberg Industry gave way to the macrolithic Oakhurst Industrial Complex, which appears to contain no stone tools suitable for use as either spearheads or arrowheads. However, sites from this period contain bone points similar to those used by both Ju/’hoansi and other southern African hunters as arrowheads and linkshafts. Regular arrow hunting may, therefore, date to this time period.
Although bone points are not common anywhere until after 10,000 BP, they were invented considerably earlier. A few bone points were found in the 38,000 BP levels of Border Cave (Beaumont 1978) (Figure 4.1) and in the 21,000 BP levels of Boomplaas Cave (Deacon 1984b:292), and a single enigmatic bone point was found in a Middle Stone Age context predating 60,000 BP in Klasies River Mouth (Singer and Wymer 1982) (Figure 4.1). Thus arrow hunting may have been known for tens of thousands of years, but it was not popular initially and probably did not alter the traditional pattern of spear hunting for large game. The change to hunting small browsers at approximately 10,000 BP cannot, therefore, be linked to the moment of the invention of the arrow. If technological invention influenced the change, then the invention may have been in the form of the modest snare-line trap because it seems possible that the small, territorial browsers were more often snared than hunted (Deacon 1976; Klein 1981). It is easier to trap shy, nocturnal, solitary antelope than to stalk them in wooded areas. It seems likely that people were using snares for hunting small browsers such as steenbok, grysbok, and gray duiker because snaring is most likely to have resulted in the catastrophic mortality rates (where prime-age adults were killed in about the same proportions as they occur in live herds) seen in faunal assemblages (Klein 1981:62). Klein thinks that Middle Stone Age people may also have used traps because the Middle Stone Age browser samples, though small, also have a catastrophic mortality profile. Thus both arrow hunting and snaring may have had their origin in the Middle Stone Age.
At Melkhoutboom the relative quantity of unworked bone fragments to stone artifacts is not constant throughout the sequence; more bone relative to stone is present in the later Pleistocene (Deacon 1976:110), and this trend is also apparent at other southern African sites. Therefore, the post-10,000 BP popularity of bow and arrow, and the targeting of small antelope, does not seem to have increased the amount of meat brought back to camp sites. The shift to hunting small browsers may, however, have been beneficial in other ways. Since the browsers live in fixed territories that do not change seasonally, they can be hunted or trapped anywhere at any time of year, in contrast to the migratory, grassland animals (Deacon 1984b:256). This may mean that smaller meat packages were obtained more reliably and regularly than when mobile herds were followed. Furthermore, when the hunters set traps close to home, they brought the game to them. The reverse is necessary for arrow or spear hunting. When hunters were freed from following game migrations, people could schedule their movements to take advantage of plant food seasonality. Deacon’s (1976) work in eastern Cape sites suggests that, indeed, plant food staples rather than game movements dictated seasonal mobility after 10,000 BP. At Melkhoutboom the underground corm of the Watsonia appears to have been the plant staple, and Deacon (1976:105) suggests that the seasonal mobility of the Cape Folded Belt populations was closely tied to Watsonia ecology.
If gender relationships in the Late Stone Age bore any relationship to those of the Ju/’hoansi in the Kalahari, then it is possible that after 10,000 BP the new importance allocated to plant seasonality gave women a large measure of control over band mobility. Conversely, band mobility may have been tied to the migrations of large game during the Pleistocene, and this may have meant that women had less control than men over band mobility before 10,000 BP. Of course this assumption is based on the premise that only men were involved in the spear hunting of the large game, an assumption that may be entirely without foundation. Today, Basarwa women do not take part in spear hunts, though these modern spear hunts probably bear little resemblance to those of the Pleistocene. Today spears are made of metal and the hunts are often conducted with the aid of dogs or from horseback. Such accoutrement enable men to hunt alone or in small groups. No dogs are present in any Stone Age sites, nor did dogs appear in southern Africa before about 400 AD (Plug 1996, personal communication), thus people in the Pleistocene did not have access to metal, dogs, and horses. Spear hunting is thus likely to have relied on large group size for game drives. Pleistocene people may have conducted themselves rather like the Ju/’hoansi in the elephant hunt described by Lee, or perhaps like the Mbuti net hunters. Mbuti band size averages between 50 and 60 people, which allows efficient net hunting by 10 or 11 men carrying spears and nets and a similar number of women who drive the game (Ichikawa 1983). Mbuti cannot have dispersal camps with fewer than 20 adults because this is the minimum number needed to hunt effectively. If a male-only hunting party was employed for net hunting, a band size of well over 100 would have been a requirement. Presumably spear hunters not using nets would have needed even more personnel than the Mbuti, and it seems unlikely that nets would have been employed in the Pleistocene spear hunts of southern Africa. Although plant fiber cordage is present in several Late Stone Age sites (Deacon 1984b:292), no netting has been found in Pleistocene sites, even where organic preservation is good. Thus women are likely to have been employed in the spear hunt for at least part of the year because, judging by the size of southern African caves and shelters occupied during the Pleistocene, band size is unlikely to have been larger than modern Mbuti band size. The size of the cave and shelter sites with later Pleistocene deposit provides some indication of demography, because several of these sites could house up to 50 or 60 people; some of the sites, however, would not have accommodated more than about 30 people. It is therefore difficult to see how the people occupying these shelters could have organized drives and spear hunts for large game without using women’s labor.
Participation in the hunt seems to ensure that women can claim rights over a portion of the hunted meat. If women were present at the hunt in the Pleistocene, they may have had a better share of meat than if they remained at home and waited for meat to be brought back to them. The principle that women present at the hunt get a better share of meat than those who wait at home can be illustrated by the Mbuti and Ju/’hoansi examples that I have already mentioned. Mbuti women who have participated in the hunt, and who have carried their husband’s game back to camp, receive a forequarter of the animal in the first round of meat distribution (Ichikawa 1983:69). In Ju/’hoansi society, however, where women do not take part in arrow hunting, it is only male hunters who receive meat in the first round of sharing. It seems, then, that the Mbuti woman’s share symbolizes her partnership in the hunt. Hunter-gatherer women in the Pleistocene may have found themselves in a similar position to Mbuti women and they may therefore have had more of a share in large game meat than modern Ju/’hoansi women.
Food sharing that is equitable or otherwise can sometimes be detected through the stable carbon isotope composition of human bone collagen. Unfortunately, skeletal samples are not abundant enough in the Pleistocene to make a study of gender similarities or differences in diet. Holocene skeletons are, however, far more abundant, and Sealy’s study of 74 Western Cape skeletons, most of which post-dated 3000 BP, shows unequal access to resources, based on gender (Sealy 1992; Sealy et al. 1992). The pre-3000 BP skeletons display few gender differences, but in the post-3000 BP sample the stable carbon isotope composition of the human bone collagen suggests that, while female skeletons were similar to those of their pre-3000 BP predecessors, on the whole male skeletons had more positive 13C values. This implies that after 3000 BP men’s diet was enriched with marine foods, for example, seal meat and fish. Sealy’s study implies that women consumed less marine food than men but more terrestrial food (presumably plants but possibly terrestrial animals). After about 3000 BP, men seem to have eaten a considerable amount of food away from the home base, and sex-segregated work parties may have become the norm. Parkington (Chapter 2) suggests that the change in men’s diet coincided with the establishment of “mega-middens” of marine shells between 3000 and 2000 BP. It may then be unwise to assume (as many archaeologists do—for examples see Barich, Chapter 6) that the gathering of shellfish was always women’s work.
Of great interest is the chronological difference in the skeletal evidence; the gender similarity in the small pre-3000 BP sample suggests that gender relationships in the Western Cape may have changed only after about 3000 BP. This is an important observation, showing that archaeologists must not expect gender relationships to remain static through time, and that they should also anticipate regional differences in social behavior.
Social, including gender, relationships may also have been affected by other changes that took place at the beginning of the Holocene and during it. Demographic change is evident in the early Holocene (Wadley 1993). It is implied by a noticeable increase in the occupation of caves and rockshelters. By about 10,000 BP many rockshelters that can hold no more than 10 to 15 people were occupied, and thus a demographic shift to smaller bands of higher densities is inferred (Deacon 1976:163). By 3000 BP the groups seem even smaller and the density even higher because many more tiny shelters are occupied (Deacon 1984b:232–236, 322). A few large sites were, however, still occupied and these may have represented occasional aggregation camps. Large-scale aggregations may have taken place only a few times in a decade (Wadley 1987), and under such circumstances it may have been impossible to spear hunt in large groups as a regular means of obtaining meat. Arrow hunting, however, relies more on stealth and aim than on large group cooperation and may be effectively conducted by individuals or small groups. Thus the social and demographic changes after 10,000 years ago in southern Africa may have promoted the greater popularity of arrow hunting and snaring. Large frequencies of bone points in sites post-dating 10,000 years ago (Deacon 1984b) do indeed suggest that arrows were in regular and frequent use by this time.
In some ways the shift to individual or small group hunting with bow and arrow or snares involved a shift to individual or family-based “ownership” of meat products. On the one hand, this may have adversely affected women’s access to game meat if, as among modern Ju/’hoansi, women were excluded from the hunt. On the other hand, the Agta, Tiwi, and Chipewyan examples suggest that women may have been hunting in their own right.
Individual “ownership” of meat products by either men or women after 10,000 years ago would have necessitated reorganization of the sharing rules that applied during the Pleistocene. In some of the post-10,000-year-old levels in Rose Cottage Cave, in the eastern Free State of South Africa, the spatial distribution around several hearths of animal bone from, for example, springbok and warthog suggests that the modern Basarwa practice of sharing large game with everyone in a camp may also have existed (Wadley in press b). Interestingly, the same type of widespread distribution may also apply to the bones of Procavia capensis, the small rock hyrax that, today, would not be shared in a Ju/’hoansi camp. At Kutse, however, small creatures may be shared by less successful Basarwa hunters who wish to maintain a sharing bond but have nothing else to contribute (Kent 1996b:151). As happens today, we may imagine that hunter-gatherer women in the Stone Age were also active trappers and collectors of small animals, so many of the faunal remains in the Stone Age sites may have been contributed by women. Consequently, the women probably butchered and shared these meat provisions with their immediate families and perhaps also with other band members.
Other important changes are apparent in the South African archaeological record between 4000 and 3000 BP. The changes include the appearance of broad-spectrum gathering and an increase in the frequencies of ornaments, elaborate burials, and portable and parietal art (Deacon 1984b) that often displays shamanistic imagery (Lewis-Williams 1984). The broad-spectrum gathering included increased plant food gathering in several areas, as well as the harvesting of quantities of small creatures such as crabs, toads, lizards, and freshwater shellfish. Marine crustaceans were collected from southern African coastal sites as far back as the MSA, but freshwater shellfish and other aquatic creatures appear in quantities only after about 4000 BP in Stone Age sites in the middle Orange River area, the Eastern Cape, the Northern Cape, eastern Free State, and Gauteng (Deacon 1984b).
In the Eastern Cape sites of Wilton, Scott’s Cave, and Melkhoutboom, frequencies of freshwater mussels increased in the topmost deposits at about 2000 BP (Deacon 1976). In other Eastern Cape sites, Welgeluk, and Edgehill, freshwater mussels, fish, and tortoise increased much earlier, at about 4000 BP (S. Hall 1990:88–98).
At Jubilee Shelter, in Gauteng, shellfish became prolific after 3000 BP (Wadley 1987:117), and at Kruger Cave, also in Gauteng, extensive shellfish middens postdated 1300 BP (Mason 1988).
Archaeologists’ explanations for the appearance of gathered shellfish and aquatic creatures vary, though most discount the role of environmental change (Deacon 1976:52; S. Hall 1990). Parkington (1980:83) sees higher population density as necessitating a general trend toward broad-spectrum gathering. Deacon elaborates this idea, suggesting that fish and molluscs were a food supplement for hunter-gatherers inland and that their collection may have been stimulated by local adjustments in group territories and annual movements, as a result of the influx of pastoralists and Iron Age people and the consequent demographic changes (Deacon 1984b:265). The Welgeluk, Edgehill, and Jubilee Shelter evidence suggests, however, that the contact situation with pastoralists and farmers should not be emphasized, because the broad-spectrum gathering was under way before their arrival.
It is difficult to gauge whether the broad-spectrum gathering preceded or postdated the population increase. From my own theoretical standpoint, I favor a social stimulus for such gathering. As I pointed out earlier, the broad-spectrum gathering is accompanied by a florescence of activities that are related to shamanistic practices. Among the Ju/’hoansi anyone may become a shaman, but experienced and aged shamans seem to have been particularly respected (Katz 1982). A desire for the longevity of valued shamans may have prompted the gathering of foods that would particularly benefit the aged. Shellfish would fulfill such a need. Shellfish are protein-rich and can be a supplement or substitute for meat; the soft shellfish “meat” is a particularly valuable source of protein for the aged (and the young) who might have difficulty chewing hard meat (Claassen 1991:279). On the one hand, then, the apparent population increase after 4000 BP may be linked to an increased life expectancy of the aged and the very young through a better diet than in previous times. On the other hand, the shellfish may have provided better diets for all age groups of the population: shellfish and fish caught in winter have high fat contents and this fat would have assisted with the metabolism of lean game meat. African game is particularly lean in the winter months, and protein from this meat cannot be metabolized in the absence of carbohydrate or fat (Speth and Spielmann 1983). As Gifford-Gonzalez points out in Chapter 7, a shortage of fat and carbohydrate may have limited hunter-gatherers’ reproductive success. The consumption of freshwater mussels may, then, have improved women’s diets with the result that there was a positive effect on hunter-gatherer birthrates.
Shellfish were probably collected in winter (S. Hall 1990). In the summer rainfall areas this would also be the dry time when most hunting and snaring were undertaken to prevent summer dew and rain from ruining the effectiveness of hunting equipment. Shellfishing would therefore compete with hunting and snaring time; in other words, it was being carried out when protein supplies were at their highest, not their lowest. Consequently it is unlikely that shellfish were regularly gathered by hunters, though they could have been gathered by snare-line operators whose time was less constrained. This means that shellfish could have been gathered by both men and women.
Among many hunter-gatherer groups the collection of shellfish usually forms part of the food package brought home by women (Murdock and Provost 1973:207), but we presently have no means of knowing the gender identity of the shellfish gatherers in the southern African Stone Age. When women who are normally plant-food gatherers begin collecting shellfish, we need to know whether tasks were reallocated or rescheduled, whether some form of specialization developed, or whether the women were simply working harder (Claassen 1991:277). If women were working harder, this could imply that they were supporting intensified male sociopolitical activities (Collier and Rosaldo 1981; Hastorf 1991:148) or that they were establishing some independence from the hunters on whom they depended for meat. If the gathering of shellfish was a form of resistance, then we must assume that some type of social change took place between 4000 and 3000 BP, and that this involved greater gender segregation and less meat sharing outside the group of hunters than was previously the case. This is a possible interpretation because, as I have already shown, skeletal evidence suggests that western Cape people were far more gender-segregated after than before about 3000 BP. Perhaps gender segregation also intensified in other parts of southern Africa in the last few thousand years.
In sites such as Rose Cottage Cave and Jubilee Shelter there are discrete activity areas for bead making, bone point manufacture, stone tool knapping, and plant food processing. Such areas may represent the type of gender-segregated work parties that are evident in Ju/’hoansi aggregation camps, but it is necessary to approach such interpretations with extreme caution. As Dobres (1996, personal communication) points out, it is unwise to make assumptions about gender attribution in the past, and I have already shown that gendered stereotypes of hunters and gatherers are inappropriate.
Dobres’s circumspection is well founded, as the preliminary study of lithics at Rose Cottage Cave shows. A residue analysis of stone tools (Williamson 1996) has revealed that both blood and plant residues are frequently present on a single stone flake. This suggests that a sole person might be involved with both meat butchery and plant food processing. Further archaeological support for this type of behavior comes from the microwear analysis of stone segments from Jubilee Shelter in the Magaliesberg of Gauteng (Wadley and Binneman 1995). Segments have long been assumed to have been arrowheads, particularly because, in Egypt, segments were found mounted as part of projectile heads (Clark et al. 1974). The Jubilee Shelter segments contain only plant polish on their cutting edges, suggesting that while they may have been used as arrowheads on occasions, they were not exclusively used as such. Of course the presence of the plant polish does not imply that the tools were not used by men, but it does suggest that no secure gender attribution can be made.
We should not assume from these examples that women merely borrowed men’s tools without ever making their own. Gero (1991) has convincingly argued that Stone Age women are unlikely to have waited passively for men to produce tools for them and that they are most likely to have produced their own flakes. In Rose Cottage Cave it is not possible to identify the gender of the stone knappers, yet the spatial evidence tends to support Gero’s claim (Wadley in press a, in press b). Stone flakes have diffuse distribution across the excavation grid in many of the occupation levels, suggesting that flakes were used for many activities in many places. With such a distribution it seems most likely that both women and men were using the flakes. Not all archaeologists are equally optimistic about the possibility of recognizing gender in the spatial patterns of Late Stone Age sites (for example, Kent, Chapter 3), but Casey’s conclusions (Chapter 5) are similar to my own. She notes that “basic tools” have wide spatial distribution in FKWMH, Ghana, and are likely to have been used by both men and women (Chapter 5).
In this brief paper I have shown that a diverse array of gender systems may have operated in the past. Of significance for archaeologists is the conclusion that hunter-gatherer gender relations were as open to change in the past as they are today. The well-worn labels “woman the plant food gatherer,” “man the hunter,” and “man the meat provider” are not appropriate for the Stone Age any more than they are appropriate for modern hunter-gatherers. Such gender stereotypes mask the subtleties of gender relations, render women almost invisible in the Stone Age, deny that they could ever have participated in the hunt, and imply that they played a timeless, unchanging role. The evidence presented here suggests that women’s and men’s roles may have changed in the past, and that, where inequalities existed between Stone Age men and women, these are unlikely to have been accepted passively. What is most important is to realize that Stone Age social and gender arrangements were not static and that ethnographically recorded examples of the male-oriented hunt, with their associated ritual and strict meat-sharing rules, should not be seen as an immutable model for the Stone Age.
I thank Bonny Williamson for the residue analysis of the Rose Cottage Cave tools and Kim Sales for drawing Figure 4.1. I am also most grateful to Sue Kent, Simon Hall, and Marcia-Anne Dobres for useful discussions. I am particularly grateful to Sue Kent for all the information she provided on the people of Kutse.
