Jump to ContentJump to Main Navigation
Social Brain, Distributed Mind$

Robin Dunbar, Clive Gamble, and John Gowlett

Print publication date: 2010

Print ISBN-13: 9780197264522

Published to British Academy Scholarship Online: January 2012

DOI: 10.5871/bacad/9780197264522.001.0001

Show Summary Details

Human Social Evolution: A Comparison of Hunter-gatherer and Chimpanzee Social Organization

Human Social Evolution: A Comparison of Hunter-gatherer and Chimpanzee Social Organization

Chapter:
(p.83) 5 Human Social Evolution: A Comparison of Hunter-gatherer and Chimpanzee Social Organization
Source:
Social Brain, Distributed Mind
Author(s):

Robert Layton

Sean O'Hara

Publisher:
British Academy
DOI:10.5871/bacad/9780197264522.003.0005

Abstract and Keywords

This chapter compares the social behaviour of human hunter-gatherers with that of the better-studied chimpanzee species, Pan troglodytes, in an attempt to pinpoint the unique features of human social evolution. Although hunter-gatherers and chimpanzees living in central Africa have similar body weights, humans live at much lower population densities due to their greater dependence on predation. Human foraging parties have longer duration than those of chimpanzees, lasting hours rather than minutes, and a higher level of mutual dependence, through the division of labour between men (hunting) and women (gathering); which is in turn related to pair-bonding, and meat sharing to reduce the risk of individual hunters' failure on any particular day. The band appears to be a uniquely human social unit that resolves the tension between greater dispersion and greater interdependence.

Keywords:   population density, chimpanzee, Pan troglodytes, human foraging, mutual dependence, social unit

WE COMPARE THE SOCIAL BEHAVIOUR of human hunter-gatherers with that of the better-studied of the two chimpanzee species, Pan troglodytes, in an attempt to pinpoint the unique features of human social evolution. Chimpanzees have presumably followed their own evolutionary trajectory, and we do not assume they represent the social behaviour of any common ancestral species. Humans and chimpanzees nonetheless share a fission-fusion type of social dynamic, although such systems are unusual in primates (Aureli et al. 2008).1 In both species small temporary, taskspecific parties of variable membership form within the residential community that later rejoin other parties, again of variable membership. This shared pattern in spatial distribution within the community can help define the role of the community for both humans and chimpanzees, and that of human bands, which appear absent in chimpanzees. Our primary data for chimpanzees come from six study sites: Budongo, Ngogo and Kanyawara in Uganda; Gombe and Mahale in Tanzania; and Taï', Côte d'Ivoire. Our primary data on hunter-gatherers are taken from ethnographic studies of 26 hunter-gatherer peoples, ranging from those who occupy the same environment as chimpanzees (Mbuti, Aka) and those occupying similar environments on other continents, to those occupying very different environments such as tropical and arctic deserts (because data on some are incomplete, our samples vary in size). In order to compare like for like, we confine our analysis to the observable behaviour of both species. It is, however, difficult completely to disregard intentions. (p.84) Male hunter-gatherers set off with the deliberate intention of hunting, whereas chimpanzee hunts can be more opportunistic; males may be foraging, resting or undertaking a border patrol when a troop of monkeys approach and stimulate the start of a hunt, and chimpanzee hunting party size may thus be determined by patrol size.

The behaviour of both hunter-gatherers and chimpanzees today is affected by the presence of more powerful human neighbours. Neither represents the ‘original condition’ of the species. Because hunter-gatherers survived longest in environments unsuitable for farming, our sample is also biased by its lack of data on hunter-gatherers living in resource-rich environments. Our case studies are in the ‘ethnographic present’ of the time they were recorded. The Mbuti have since been caught up in the civil wars on the Rwandan-Congo border, while the government of Botswana is currently (2008) denying G/wi access to bore water in the Kalahari Game Reserve. Disease had reduced many populations by the time they were first studied by anthropologists, although some have since recovered (Cree—Rogers 1972, 94; Northwest Coast—Rohner & Rohner 1970, 84; G/wi—Silberbauer 1972, 303; Gugadja—Cane 1990, 150). Furthermore, most hunter-gatherers today rely to some extent on cultivated or purchased foods and this may affect not only the size, but even the shape, of foraging ranges. Mbuti territories, for example, consist of long, narrow strips extending from a farming village to the deep forest (Ichikawa 1978, cited in Bahuchet 1992). However, through a comparison of those human communities whose subsistence strategies are closest to those of chimpanzees, we identify consistencies in behaviour that throw light on the species’ characteristic social strategies, and variability that indicates how such strategies have been adapted to specific ecological constraints.

Although the total size of a typical hunter-gatherer community is greater than that of a chimpanzee community, the size of the action sets that forage together is similar. Male chimpanzees move in parties of up to 4–5, while human hunters frequently hunt singly or in pairs. Cooperative drives similar to those claimed for male chimpanzees take place in larger parties of 10–30. Female chimpanzees move in parties of up to seven (O'Hara unpublished data), foraging women in parties of up to ten (see Table 5.1). The size of the group within which day-to-day interaction occurs—the community among chimpanzees and the band among humans—is comparable, ranging from 40 to 150 among chimpanzees and 35 to 80 among hunter-gatherers. However, while the hunter-gatherer band has frequently been treated as the human equivalent of the chimpanzee (p.85) community, we argue that this is incorrect. The chimpanzee community should be compared with the human community, and the band treated as a more enduring fusion phase within a wider pattern of fission and fusion sustained by inter-band movement. We consider the functional benefits of band membership, and of membership in a wider community.

Band or Community?

Social anthropologists have long equated hunter-gatherer society with the band (Radcliffe-Brown 1930; Steward 1936). Men were supposed to remain in the same band all their lives, while women transferred from one band to another at marriage. When it was discovered that the male chimpanzee has to spend his entire life living in the community into which he was born while female chimpanzees typically change community at adolescence, the chimpanzee community was equated with the huntergatherer band (e.g. Foley & Lee 1989 and commentators on Hawkes et al. 1997). The band, however, is a much more fluid and permeable grouping than Radcliffe-Brown and Steward appreciated. Human adults of both sexes often freely change band membership, whereas the chimpanzee community appears to be a closed reproductive unit (Wrangham 2000).

Among chimpanzees, the males of the community defend the boundary of their territory and attack individuals from neighbouring communities that they discover in the border zone (Wilson et al. 2004). Men may range more widely than women within the human band territory, but this is probably due to prey distribution rather than boundary defence (Aka—Bahuchet 1992, 222; Cree—Rogers 1972, 106, 109, 127; G/wi—Silberbauer 1972, 287, 290; Nukak—Politis 2007, Table 6.2; Numamiut—Binford 1976, 169).

Neighbouring hunter-gatherer bands often acknowledge mutual rights of access over each others’ territories. Richard Lee writes that camp composition among the Ju/'hoansi changes monthly and even daily, due mainly to intercamp visiting (1979, 54). Colin Turnbull wrote of the Mbuti (sympatric with chimpanzees), that ‘individuals or whole families sometimes wander…until they are several territories distant from their home territory’ (1965, 96). Visiting between Hadza camps was continuous and it was ‘not uncommon’ for individuals or single families to change camp (Hawkes et al. 1997, 553). Winter band composition among the Cree varies from year to year (Rogers 1972, 121).

(p.86) (p.87)

Table 5.1. Foraging party size, band size, community size and population density in a range of groups and habitat types

Foraging party size

Band size

Community size

Pop. density/km2

 

(a)male

(b) female

range

mean

 

 

Tropical forest

Ache (Paraguay)

15–70

50

<500

0.03

Aka (Central

African Republic)

c. 24–28

500

0.031 (Bagandu [1])

0.28 (Bagandu [2])

0.017 (N'Dele)

Batek De (Malaysia)

1–2

2–3

‘20 families’

c. 80?

200

0.11

Cholanaickan (India)

5–28

20

250

0.6

Mbuti (Zaire)

1–5

2–3

37–62

0.17–0.2

Nukak (Colombia)

1–3

2–11 (mixed parties)

12–63

20–30

125

0.034

Tropical coast

Gidjingali (Australia)

1–3

2–7

14–80

34

300

0.46

Gunwinggu (Australia)

18–44

175

0.05

Tiwi (Australia)

1–3; 15

2–3

40–50

1000

0.4

Yolngu (Australia)

32.7

>500

0.34

Savanna

Hadza (Tanzania)

1

2–10

35–60

750

0.34

Yolngu (Australia)

0.06

Semi-desert

G/wi (Kalahari)

1–2;12

21–85

57

2000

0.07

Ju/'hoansi (Kalahari)

1–2

2–4

9–75 (<141)*

42

460

0.017

Warlpiri (Australia)

25–30

>200

0.01

Western Desert (Aus)

6–30

13.6

250–600

0.01–0.02

Temperate coast

Kwakiutl (Canada)

50–60

280

0.24 [1] 0.57 [2]

Nootka (Canada)

1 to ‘several’

0.4 [1] 0.66–0.77 [2]

Tlingit (Canada)

1–2; 20–60

<50

Boreal forest

Cree (Canada)

15–50

450 [1] 250–1000 [2]

0.004

Khanti (Siberia)

1–6 households c. 14?

800

0.005

Arctic coast

Central Canadian Inuit

Winter 50–150 500

Summer 15–30

0.012–0.005

Hudson Bay

1–8

500–700 (Itivimiut)

150 (Qiqiqtamiut)

0.016

Netsilik (Canada)

2–3

20–30 (mixed)

Winter 50–100

Summer 20–30

260

0.005

Taremiut (Alaska)

30–100 [1] 12–75

336 [1] av 450 [2]

0.05–0.15 [1] 0.07 [2]

Arctic interior

Nunamiut (Alaska)

2; 35

Winter 50–150 (Gubser)

Summer 18–36 (Binford)

0.02

(*) Population at /Xai/xai is exceptionally large (141); the next largest waterhole group size is 75.

Where two sets of figures are given for male parties, the larger figure refers to collective hunting (e.g. fire drives, seals, caribou).

Sources

Ache—Hill & Hurtado 1996, 41–19, 61, 81; Aka—Bahuchet 1992, 219 [1]; Hewlett et al. 1982, 425 [2]; Bahuchet 1988 cited in Kelly 1995; Batek Dé—Endicott 1979, 11; 1988, 110; Cholanaickan—Bhanu 1992, 48; Kelly 1995, Table 6.4 cites similar densities for Hill Pandaran and Paliyan; Mbuti—Bahuchet 1992, 212; Turnbull 1965, 167; Harako 1981, 506–507; Nukak—Politis 2007, 25, 36, 77–78, 164; Gidjingali—Jones 1980, 137; Meehan 1982, 13, 15, 90–105; Peterson & Long 1986, 40–42; White et al. 1990, 177; Gunwinggu—Altman 1987, 15, 22, 23; Tiwi—Hart & Pilling 1960, 33–37; Goodale 1986, 199; Yolngu—Peterson & Long 1986, 135; Williams 1986, 131; Hadza—Woodburn 1972, 193–194, Hawkes et al. 1997, 552–555; G/wi—Silberbauer 1972, 273, 295; Ju/'hoansi—Lee 1979, 53 and pers. comm., 351; Warlpiri—Meggitt 1962, 47; Peterson & Long 1986, 38, 69; Western Desert—Gould 1969, 64; Layton 1986, 43; Peterson & Long 1986, 135; Tonkinson 1991, 71; Kwakiutl—Codere 1950, 50–52 [data for 1853], [1]; Hunn 1994, [2]; Kelly 1995, Table 6.4 citing Mitchell & Donald, and Shalk; Nootka—Drucker 1965, 17, 19, 150, [1] estimated from map; Drucker 1965, 3, [2]; Kelly 1995, Table 6.4 citing Mitchell and Donald; Tlingit—Emmons 1991, 105–132; Cree—Scott 1988, 38, [1]; Rogers 1972, 91, [2]; Feit 1983, 417; Khanti—Jordan 2003, 40–42, 73–74, 85, 251–260; Kelly 1995, Table 6.4 cites same density for Yukaghir; Central Canadian Inuit—Damas 1969, 45, 51, 113; 1984 cited Kelly 1995, Table 6.4; Hudson Bay Itivimiut—Smith 1991, 112; Hunn 1994; Qiqiqtamiut—Guemple 1988, 131; Netsilik—Balikci 1970, xxiii, 56; Kelly 1995, Table 6.4; Taremiut [1]—Andrews 1989, 21, 93–94, [2; community figure for 1880]; Burch 1988, 96–97; Nunamiut—Binford 1976, 204–205, 231, 236; Gubser 1965, 167; Kelly 1995, Table 6.4 citing Hall.

To increase sample size in the box plots we have added secondary data from Kelly on band size and population density for the Bihor, Hill Pandaram, Paliyan, Semang and Andaman Islanders, and population density for the Great Basin (USA), Bella Coola, Haida, Tsimshian, Chipewan, Tutchone, Gilyak, Round Lake Ojibwa, Ojibwa, Yaghan and Yukaghir. We have not used Kelly's data on band sizes for these high latitude societies as these appear to be based on winter aggregations (see discussion in text).

(p.88) It is harder for hunter-gatherers to move between regional communities (Kalahari—Wiessner 1982; Nukak—Politis 2007, 164) and the hunter-gatherer community sometimes acts collectively to defend subsistence resources (Alaskan Inuit—Andrews 1989, 30; Burch 1988, 98; Gubser 1965, 166–167: Gidjingali—White et al. 1990, 177). On the Northwest Coast of North America, the winter village community contained several lineages that came together for mutual defence (Boas 1966, 35–36; Drucker 1965, 47; Hunn 1986, 33–34).

The human community is most frequently characterized as an endogamous breeding unit. The Alaskan Inuit community was 80 per cent endogamous (Burch 1988, 96) and there was a high rate of endogamy among the Nukak munu (Politis 2007, 164). Wobst's pioneering paper predicted, on the basis of computer simulations, that a demographically stable hunter-gatherer breeding population would require a minimum of 175–475 individuals (Wobst 1974, 169). Data in Table 5.1 support this prediction.

Thus the human community seems the best parallel with the chimpanzee community. Central to our argument is the hypothesis that bands are better seen as intermediate groups which have crystallized during human social evolution, emerging as social bonds of cooperation and reciprocal exchange between individuals became stronger during the evolution of modern hunter-gatherer strategies. Equating the regional community, rather than the band, with the chimpanzee community is more consistent with Dunbar's (1993) prediction of group size from brain evolution, although Barton (2006) has recently argued that specific aspects of social complexity, including the tracking of relationships with absent individuals, provides a better explanation for human brain evolution. The capacity to move between bands, according to this hypothesis, would persist from the earlier pattern of fission-fusion within the wider community, but enhanced by specifically human traits such as language, and gift exchange with members of other bands.

In the great majority of societies included in this survey, band recruitment is flexible. The greatest degree of fluidity appears to exist among the Batek De, where, according to Endicott (1979, 10–11), there are no lineages, bands or other corporate groups larger than the conjugal family. In most cases, the core of the band consists of a few bilaterally related families, to which other families attach themselves for a greater or lesser period (Ache—Hill & Hurtado 1996, 66; Aka—Bahuchet 1992, 219; Cholanaickan—Bhanu 1992, 48; Hadza—Hawkes et al. 1997, 553; Ju/'hoansi—Lee 1976, 77, 1979, 58, 65, 338; Yellen 1976, 60; G/wi— (p.89) Silberbauer 1972, 308; Cree—Rogers 1972, 106, 119, 121; Hudson Bay Inuit—Smith 1991, 110). In Australia's Western Desert, individuals began life with potential rights in the father's and mother's bands, and also to the place where they were born, but only become ‘kin to the country’ by looking after it (Layton 1995). Myers (1986: 183) argues that while individuals must affiliate themselves to a residential group, rights to land in the Western Desert are best seen as the outcome of individual claims and assertions rather than descent.

Social Functions of Fission and Fusion Among Hunter-Gatherers

To investigate the adaptive value of the distinctive patterns of fission-fusion behaviour among hunter-gatherers, we look at the activities that take place during both fission and fusion phases.

Meat consumption

One of the most striking differences between hunter-gatherers and chimpanzees is that humans consume much more meat. Stiner (2002) concludes that systematic hunting of large ungulates by humans began about a quarter of a million years ago. Humans living in tropical forest consume up to ten times more meat than chimpanzees (see Table 5.2) and this has consequences both for population density and the social organization of hunting. Hayden (1981) and Marlowe (2001) found that male hunting contributes from between 20–25 per cent to 100 per cent of the human hunter-gatherer diet. Hunting contributes most to the diet in high latitudes, both because there is less plant food for women to gather and because more of the diet is derived from fishing. While information on weight of meat consumed per person is relatively hard to locate, the difference between chimpanzees and central African hunter-gatherers is striking.

As a consequence of our species’ greater reliance on predation, hunter-gatherers occupying the same environment as chimpanzees live at considerably lower densities (Table 5.3; Figure 5.1; see also Grove ch. 19 this volume). In less rich environments humans are even more dispersed (Figure 5.1: p < 0.001). Hunter-gatherer population densities are thus very responsive to habitat type.

Because the human community is considerably more dispersed than the chimpanzee community, the coherence of the community cannot be guaranteed by daily interaction and humans have evolved cultural mechanisms such as gift exchange and classificatory kinship to sustain relationships with individuals beyond the band whom they may meet only infrequently. Why, then, has the human community persisted, despite the greater difficulty of maintaining it? We also consider various functional benefits of retaining the ability to move between bands in the community.

(p.90)

Table 5.2. Meat consumption

 

Kg/individual/day

% meat in diet

Source

Chimpanzee

Gombe (male)

0.055

equivalent to 2–6%

Boesch & Boesch-Achermann

(female)

0.007

of total Kcalories/day

2000, 165

Taï (male)

0.186

(female)

0.025

Human hunter-gatherer

Tropical forest

Mbuti

0.5

30%

Ichikawa 2005, 159

0.45

Hart 1978 in Kelly 1995, 103

1.06

Tanno 1976 in Kelly 1995

Ache

1.78

87%

Hill et al. 1985, cited Kelly

1995, 103; Hill & Hurtado 1996, 65

Nukak (dry season)

0.237

10%

Politis 2007, 75

Nukak (wet season)

0.306

Huaoroni

0.200+

Rival 2002, 75

Batek

0.2

67%

Endicott 1981 in Kelly 1995;

Endicott 1988

Tropical coast

Gunwinggu

0.621

50%+

Altman 1987, 36

Savannah

Hadza

<1.00

50%

O'Connell et al. 2002, 858

Semi-desert

Ju/'hoansi

0.256

30%

Lee 1979, 265 based on 28 days obs.

G/wi (October)

0.03

Silberbauer 1972, 285

G/wi (January and June)

0.22

30%

Western Desert

0.56

15–30%

Gould 1980 in Kelly 1995, 103

Notes: Figures are for meat rather than protein and are taken from data in Hayden 1981, Table 10.3 (where hunting is distinguished from fishing), except Gunwinggu (Altman 1987, 42, 44), Nukak, based on raw weights of fruit and meat (quoted in Politis 2007), and Hadza, for whom O'Connell et al. (2002, 836) give figure of 50% ‘for full-time foragers’ (Hayden quotes a figure of 20% meat in Hadza diet). 87% of Gunwinggu bush protein derived from men's hunting and fishing (Altman 1987, 41), yielding 0.621 kg meat/person/day. Figure for Ache from Hill and Kaplan 1988 in Betzig, Borgerhoff-Mulder and Turke, via Hewlett 1991. Ichikawa's (2005) figure for the Mbuti is based on his report that a medium-sized duiker contains enough meat for 9 adult-days. Silberbauer (1972, 286) calculates that a duiker yields 10 lbs of meat (Lee 1979, 214 gives the meat yield from a duiker as 9.3 kg). Hayden (1981, 353) questions the validity of some data. He points out, for example, that while Gould estimates that Western Desert Aborigines consumed 0.5 kg meat per day, the returns on the hunting trips he observed were much lower. Silberbauer (1981, 486) notes that his data on yields are sometimes estimates.

(p.91)

Table 5.3. Body weight, population densities and foraging ranges, central Africa

 

Male

Female

Body weight in kg (Smith & Jungers 1997)

Chimpanzee

42.7–59.7

33.7–45.8

Pygmies

47.9

42.2

Population densities per km2

Chimpanzee

2.5

Wrangham et al. 1993

Aka

0.31

Hewlett et al. 1982 (Man)

0.28 in Bagandou

Bahuchet 1988, cited in Kelly 1995

0.17 in N'Delé

Hewlett et al. 1982

Mbuti

0.17 to 0.2

Turnbull 1962; Ichikawa 2005

Home range size in km2

Chimpanzee (Gombe)

9.0–12.0 (male)

5.8–11.0 (female; smallest recorded range)

Chimpanzee (Ngogo)

35 (largest recorded range)

Mbuti

260

Turnbull 1968, 134, cited in Abruzzi 1980, 22

Aka

490

Hewlett et al. 1982, 422, 424

The Fission Phase (Foraging Parties)

Although the size of the ‘action sets’ that forage together is of the same order among human hunter-gatherers and chimpanzees, human parties last for several hours while chimpanzee parties are of much shorter maximum duration, ranging from 69 minutes at Gombe (Halperin 1979) to 19 minutes in Budongo (Reynolds 2005). Unlike chimpanzee females, hunter-gatherer women very rarely forage alone. Hawkes et al. (1997, 558) write that Hadza women usually forage as a group. Warner (1958, 129) writes that the women in a Yolngu band frequently forage together as men claim to have practised sorcery on women encountered alone, while Hart and Pilling write that the young wives of Tiwi men are rarely allowed to (p.92) forage without a chaperone (1960, 36–37). According to Tonkinson (1991, 45), Mardu women usually forage in groups because it allows them to be sociable and share child-minding.

Human Social Evolution: A Comparison of Hunter-gatherer and Chimpanzee Social Organization

Figure 5.1. Hunter-gatherer population density as individuals per km2 by habitat type

Hunting techniques

Male chimpanzees hunt animals smaller than themselves, typically monkeys. The most commonly hunted species across sites is the red colobus (Newton-Fisher et al. 2002) although, in the absence of red colobus at Budongo, chimpanzees there hunt black and white colobus and blue duiker. Chimpanzees hunt singly or in the company of others. In group hunts, a single prey is separated from a group and cornered. West African Taï chimpanzees are reported to collaborate, taking on the roles of ‘drivers’ and ‘blockers’ in hunts (Boesch & Boesch 1989). Once killed, any chimpanzee (male or female) present at the place will attempt to gain some of the meat. Active or passive sharing occurs (Newton-Fisher 2007).

(p.93)

Human Social Evolution: A Comparison of Hunter-gatherer and Chimpanzee Social Organization

Figure 5.2. Band size by continent and habitat type

(p.94) Some human hunting strategies appear similar, although using sophisticated technology and communication. In Mbuti net-hunting, for example, several families cooperate, joining their nets together. Men drive small game toward the net, while women and children kill any animal that is trapped (Harako 1981, 519; Turnbull 1965, 154). Aka bands on the same trail periodically form a single camp to enable more effective nethunting (Bahuchet 1992, 226). Meat is always taken back to camp to be divided. Even those who were not present (such as elderly people and children) can expect a share. Fire drives among the Tiwi involve 10–15 men, with women and children acting as beaters (Hart & Pilling 1960, 33–37, 42). Finlayson (1935, 44) describes a similar fire drive in the Australian Western Desert. Inuit men cooperate to catch seals: each seal keeps several breathing holes open in the ice, and one man must stand by each hole to corner the seal.

Humans also have a different hunting strategy, hunting by stealth, where men go out from camp singly or in twos, to hunt larger prey. They creep up on the animal until it can be speared or shot with an arrow. Again, the meat is taken back to camp to share. Human hunting techniques are thus determined by both hunting technology and the behaviour of the prey species. Nunamiut congregate at the spring and autumn caribou hunting sites, located to intercept the migrating herd, but scatter in summer to hunt dispersed game (Binford 1976, 204, 235, 277, etc., cf. Damas 1969). Men in the Cree winter camp combine to hunt caribou, but moose/elk are solitary and are usually hunted by pairs of men (Rogers 1972, 111) or single hunters among the Khanti (Jordan 2003, 250).

A second key variable in human hunting strategies is the predictability of herd locations. Herd movements are more predictable in the arctic and boreal forest than in savannah and semi-deserts. Where the location of game is unpredictable, hunters of the band can reduce the risk of failure by dispersing. Hadza men routinely hunt independently, reducing but not eliminating the risk of failure (O'Connell et al. 2002, 836). Even though they target herd animals, G/wi hunters generally operate in pairs (Silberbauer 1981, 474). Each pair leaves camp in a different direction. Smith attributes the large winter aggregations of the Inuit to the wish to reduce risk in seal hunting through the winter ice by including more than one hunting team in the same camp (Smith 1991, 330). The success of this strategy depends crucially on hunters’ willingness to bring meat back to camp and distribute it beyond the hunting party (Smith 1991, 330; cf. Damas 1969, 51).

(p.95) Fusion Phase 1: The Band

The overnight camp is a defining feature of human hunter-gatherer adaptation. A band is a group whose members meet up at the end of the day, assembling at a camp site which has been pre-arranged and may be used for a number of days at a time. Sharing of resources, such as the meat from game animals, takes place at this base camp. Chimpanzees do not assemble in this way; there are no set base camps and individuals build a new nest site wherever they find themselves at nightfall. The mean number of bands per community in our sample is approximately 17–18, ranging from 5 among the Kwakiutl to 30 in the Western Desert (Australia) and the G/wi, and up to 67 among the Khanti (see Table 5.1). The frequency with which the band moves camp location is variable. The Nukak band moves 70–80 times in a year, the mean distance between camps 3.85 km in the wet season and 7.65 km in the dry season (Politis 2007, 100, 166–167). Cholanaickan bands move weekly (Bhanu 1992, 48). Aka and Mbuti bands move camp on average six times a year; successive sites are usually one hour's walk apart, about 6 km (Bahuchet 1992, 212, 222).

Binford (1980) distinguished between ‘logistic organization’ and ‘foraging’, arguing that the Nunamiut were logistically organized while the San were foragers. Logistic organization is adapted to a highly seasonal environment, where the location of key food resources differs in a predictable way from season to season. Under such conditions, hunter-gatherers can cache seasonally used equipment at the appropriate site and move camp to another resource as it becomes available. Living in fine-grained environments with low seasonal variation, foragers exploit diverse resources and seek to locate their camps at places that will minimize travel distance to all resources. However Yellen (1976) argued that the Ju/'hoansi also strive for logistic organization, but are mapping onto a different resource distribution. Foragers in semi-desert environments must camp near scarce water supplies. The Ju/'hoansi occupy small, widely scattered and short-lived camps in the rainy season, and more permanent ones near secure water sources in the dry season. In the Western Desert of Australia, each band territory contains a permanent or semi-permanent water source that forms the base camp during drought (Layton 1986, 39–41). Hawkes et al. (1997) record that one Hadza band predicated its camp locations on the local abundance of plant foods, moving to fresh plant foods as they ripened.

(p.96) Functional benefits of the band

Among non-human primates, party size is explained as a trade-off between reducing the increased predation risk experienced by small parties, and the increased competition for food experienced by large parties (Sterck et al. 1997; van Schaik 1989). Not only does human band size show no significant variation between Africa, where there is a high predation risk, and other continents (Figure 5.2a: p = 0.201), nor does it vary significantly between habitat type (Figure 5.2b: p = 0.505). Given the sensitivity of forager population densities to ecology, this is remarkable.

We therefore hypothesize that band size is the outcome of trade-offs between social costs and benefits. The flexibility of band composition is consistent with individuals’ need to balance or ‘trade off’ competing benefits of joining different bands, in particular the husband's and wife's parents’ bands. We identify four advantages of living in a band: childminding, prior rights to resources, food sharing and information exchange. The primary cost of band life is competition for food resources within a day's range from camp.

Child-minding

Hawkes et al. (1997) found that Hadza mothers reduce the time they spend foraging at the birth of a new child. Older, weaned children too young to support themselves will therefore suffer. This is where the grandmother steps in: grandmothers spend most time foraging when their infant grandchild is youngest and their weaned grandchildren are receiving least from the mother. Thus a young mother will benefit from living with her parents’ band. While Gurven and Hill (1997, 566) point out that only two of the eight older women tracked in the study were mother's mothers, five out of the eight were matrilateral relatives, and there is therefore evidence for the adaptive value of adult daughters living with maternal relatives. The specific value of the maternal grandmother's role has since been supported by Sear et al.'s (2000) study of infant survival among subsistence cultivators in the Gambia, and Gibson and Mace (2005) in Ethiopia. Leaving children in camp was a common practice (Ache—Hill & Hurtado 1996, 66; Cholanaickan—Bhanu 1992, 40; Ju/'hoansi—Lee 1979, 310, but in the context of Lee's seminal work on the relationship between carrying children when gathering and birthspacing).

The ability of women to gather into old age is an important aspect of hunter-gatherer behaviour, but does not necessarily lead to young women (p.97) remaining permanently in their mother's band. Newly married couples typically live with the bride's parents for a few years before returning to the husband's band (Aka—Bahuchet 1992, 219; Nukak—Politis 2007, 82; Cree—Rogers 1972, 106; G/wi—Silberbauer 1972, 303). Our data show that only one-third to a quarter of women live matrilocally. Despite the flexibility of recruitment to bands, which may be due in part to the marginal environments in which most recent hunter-gatherers live, there is widespread evidence for a patrilineal bias. Terashima (1985, quoted in Bahuchet 1992) reported that about three-quarters of men in a Mbuti band live with their paternal kin, one-fifth with maternal kin and onefifth with their wife's kin. Bahuchet (1992, 219) also found two-thirds of Aka families living with the husband's band, one-third with the wife's. In the Uluru region of the Western Desert, 70 per cent of people joined their father's band, 20 per cent their mother's and 10 per cent the band of a more distant relative (Layton 1983, 25). The numayms (totemic clans) of the Kwakiutl, on the Northwest Coast of North America were recruited bilaterally, but with a slight patrilineal bias (Drucker & Heizer 1967, 10), although elsewhere on the Northwest Coast recruitment was ambilineal or matrilineal (Richardson 1986). There are evidently conflicting advantages of patrilocal and matrilocal residence that individuals must resolve when deciding which band to join.

Prior rights to resources

In two cases among the societies sampled (Batek De—Endicott 1979, 10; Hadza—Woodburn 1972, 193) there are no band prerogatives over foraging in any area. In a few instances the community, rather than the constituent band, is the land-holding unit (see above). Generally, however, the band holds prerogatives, but not exclusive rights, over resources in its territory (Cholanaickan—Bhanu 1992, 39; Ju/'hoansi—Lee 1979, 70; Mbuti—Turnbull 1965, 93, 174, 222–223; Nukak—Politis 2007, 163; Nunamiut—Gubser 1965, 165–167). Peterson and Long argue that even in the rich tropical woodland of Arnhem Land, an Aboriginal band of 40 occupying a territory of 400 km2 would have had to defend a boundary of 70 km, equivalent to 2 km/man. Boundary defence is therefore not practised anywhere in Australia (Peterson & Long 1986: 29). Greeting rituals, where visitors wait for permission to enter a camp, are widespread in Australia (Peterson & Long 1986: 27–28). Peterson (1975) coined the term ‘social boundary defence’ to describe the strategy of controlling access to the group that asserts prior rights to the band territory or, in (p.98) Stanner's (1965) term, ‘estate’.2 However, some cases in which the band defends its territory are known (Ache—Hill & Hurtado 1996, 70, 72; Northwest Coast lineages in summer—Boas 1966, 35–36; Drucker 1965; 47; Hunn 1986: 33–34).

Information exchange

Among the G/wi, all information about game is relayed to hunters when the band camps together (Silberbauer 1981, 474). Information about caribou herd movements is exchanged in the Nunamiut village (Binford 1976, 178). Marshall (1976, 351) wrote that talking maintains good, open communication among band members, releases emotion, sanctions behaviour, and exchanges information about events, plans and movement of people. Men retell hunting exploits, women recall good gatherers.

Food distribution

Meat sharing is probably the best known, and most debated, aspect of band behaviour. Ichikawa (2005) points out that hunting success is uneven, but a large antelope in the Kalahari contains enough meat for 90–135 adult consumption days. Non-human carnivores cope with daily fluctuations through gluttony, whereas primates have a herbivore feeding rhythm. Ichikawa argues humans have inherited the latter pattern, despite their greater dependence on meat, and fluctuating supplies are dealt with by sharing. Sharing is most simply explained as an example of reciprocal altruism, a practice identified by Trivers (1971). Trivers argues that where there is a risk of death, e.g. from starvation, and where no-one knows who will be successful on any one occasion, yet those who are successful in obtaining food get more than their immediate need, it will pay to share, because when the once-successful individual is unsuccessful on another occasion, the debt can be repaid. Both partners will therefore survive whereas, on their own, both would have died. It has been shown that such reciprocity is only reliable where two other conditions hold good (Winterhalder 1990, 1996): individuals must have long-term social relationships, to allow for the balancing of give and take between partners, and it must be possible to detect, and punish, (p.99) ‘cheaters’, who try to avoid sharing when they do well, but take when they do badly. Band organization allows these conditions to be satisfied. The practice of sharing meat and honey within the Ache band increased nutritional status by 60 per cent (Kaplan & Hill 1985b, 233). Among the Nukak, even water and firewood are shared (Politis 2007, 79, 125). Sharing meat helps to keep stress and hostility over food at a low intensity, and fear of hunger is mitigated by knowledge that the receiver will share when he later has food (Ju/'hoansi—Marshall 1976, 357).

Convincing evidence in favour of the risk-reduction hypothesis lies in the nature of the foods that are most widely distributed, primarily large game (Gunwinggu—Altman & Peterson 1988, 78–79; Ju/'hoansi—Marshall 1976, 357). Among the Ache all meat is subject to the same sharing rules but the distribution of gathered foods is less extensive than of meat and honey (Hawkes 1990, 155ff; Kaplan et al. 1990, 114, 124). An Ache hunter, though he will fail on most days to feed a family, would bring in a larger amount than a family could possibly consume often more than once a week (Hawkes 1990, 151; Kaplan et al. 1990, 114).

The care taken to diffuse possible disputes during meat sharing is well known (Aka—Bahuchet 1992, 229–231; Ache—Kaplan et al. 1990, 129; Cholanaickan—Bhanu 1992, 42; Gunwinggu—Altman & Peterson 1988, 80; Mbuti—Ichikawa 2005; Nunamiut—Binford 1976, 235). Among the Netsilik, where successful winter seal hunting depends on the cooperation of a number of hunters, hunting partners were not close kin, but were chosen by the hunter's mother at birth or during the hunter's childhood, creating complex interlocking sets of relationships in the winter camp (Balikci 1970, 133–135).

Against the risk-reduction hypothesis

Two principal arguments claim the risk-reduction hypothesis is inadequate: first, some hunters are better than others, and therefore always give more meat than they receive and, second, hunters often receive little or nothing from their own kills (Hawkes & Bliege Bird 2002; O'Connell et al. 2002; cf. Bahuchet 1992, 229–231). The striking differences in hunting success documented by Lee are well known: of 127 men in Lee's Ju/'hoansi sample, 37 had never killed a kudu, whereas 43 had killed 10 or more (Lee 1979, 243). This may, however, be exceptional. Differential success in hunting affects the Mbuti to a lesser degree: of 10 net-hunters in an Mbuti camp, the most successful obtained 140 kg meat in four weeks, whereas the least successful got 24 kg (Ichikawa 2005). We argue (p.100) that, just as the grandmothering hypothesis placed women's foraging activities in a lifetime perspective, the costly signalling hypothesis for hunting effort (see below) must be placed in a man's lifetime perspective, to resolve the apparent contradiction identified by O'Connell et al. (2002).

Married men

Lee reports that, among the Ju/'hoansi:

Many good hunters did no hunting at all for weeks or months at a time, while their wives and children waited patiently and ate the meat distributed by other hunters…A period of hunting inactivity allows the hunter to enjoy the benefits of some of the reciprocal obligations he has built up. (Lee 1979, 248–249)

Above-average foragers may be willing to give more than their share to avoid the risk of long stretches without food if they suffer injury (Kaplan & Hill 1985b, 237). Kaplan and Hill argue that high producers can expect to be well treated to prevent them joining another band. Wiessner, moreover, has shown that good hunters among the Ju/'hoansi fathered twice as many children as poor hunters, and also had more surviving children (2002, 419; cf. Ache—Kaplan & Hill 1985a).

Single men

Single men among the Ache are the biggest losers because they hunt frequently but most of their game is distributed to other members of the band (Kaplan & Hill 1985b, 233; Kaplan et al. 1990, 127; cf. Gunwinggu—Altman 1987, 134, 139; Altman & Peterson 1988, 80). This requires further explanation. Hawkes and Bliege Bird (2002, 60–61) conclude that hunting is a form of costly signalling, by which meat distribution becomes a medium of communication through which the hunter transmits information to potential mates, allies and competitors. We propose that male foraging strategies must be interpreted within a lifetime perspective. Women can gather effectively into old age, but successful hunting demands good eyesight. Among the Ju/'hoansi age is an important factor in hunting success, although not the only one (Lee 1979, 243–244). Ju/'hoansi men continue to hunt into their fifties, but ‘the camp core of older people want to encourage the sons-in-law to stay with the group permanently, as more hunters mean more meat’ (Lee 1979, 242). Hawkes et al. (1997, 555) write that among the Hadza, older women (p.101) spend significantly more time foraging than females in any other category, whereas the male pattern is quite different, ‘the peak for males comes before marriage’ (italics added). Among Aka, hunting nets are frequently owned by adult men, but it is young men who do the hunting (Ichikawa 2005). As Nicholas Peterson long ago pointed out (pers. comm. to RL), a young man's best asset is his hunting ability, whereas one of the best assets an older man has is a marriageable daughter (cf. Lee 1979, 240). It is when a hunter's eyesight is failing that his daughter is ready for marriage, and he must entice young men to hunt for him. Among the Tiwi of northern Australia, where women are betrothed at birth, a man might bestow his infant daughter on someone he wanted as an ally, or he might bestow her on a younger man as ‘old-age insurance’, in which case he would look for a younger man in his late twenties or thirties who showed signs of being a good hunter and fighting man (Hart & Pilling 1960, 34; Hart et al. 1988, 19). We conclude that while hunting success may be a form of costly signalling for the young, unmarried hunter, surrendering the kill is a long-term investment that will eventually guarantee him meat in old age.

Fusion Phase 2: The Community

Defining the community

Many hunter-gatherer communities number between 250 and 500 individuals (Table 5.1). Where hunter-gatherers live in the same or similar environments to chimpanzees, the community is up to five times larger than the chimpanzee community. A few hunter-gatherer communities are apparently even larger (1,000–3,000), or even 10,000 (Nootka—Drucker 1965, 144), but the literature does not state whether individuals could move between bands throughout this larger population. Drucker describes the Northwest Coast winter village as an alliance of local groups whose territories were contiguous, and calls village communities ‘tribes’ (1965, 70). Boas wrote specifically of the Kwakiutl: ‘The people speaking the Kwa'g-ul dialect inhabit many villages, each of which is considered as a separate unit, a tribe’ (1966, 37). On the Northwest Coast of North America, it seems safer to treat the aggregate of bands/lineages that formed a single winter village as the community. The Yolngu of northern Australia, another apparently very large community of 2,500–3,000, are actually an aggregate of speakers of up to six related dialects; clans speaking the same dialect belong to the same mala. (p.102) Although clans on the border between two mala claimed dual affiliation, the mala seems better to correspond to our ‘community’ (Keen 1982, 632; Warner 1958, 35–36; Williams 1986, 64–65). Among the Inuit there was regular movement between winter villages and here we have taken the named community (-miut = people of a named area) as the community. In the literature on the Cree, the term ‘band’ is applied to the community rather than the band as we have defined it above.

In some of our case studies, the total population comprises a single community (e.g. Batek De, Hadza and Cholanaickan). Like the smaller Nukak communities, these may be isolated remnants whose size may have been reduced by disease or loss of territory (see Bhanu 1992, 31).

Population density

Although there is little difference in weight between central African hunter-gatherers and chimpanzees, human communities occupy much larger territories, both because humans live at lower population densities (Table 5.3), and because their communities are bigger (Table 5.1). We concluded above that this is due to the greater contribution meat makes to the human diet (Table 5.2). The lower population densities of hunter-gatherers, and the larger size of the community, pose serious problems for the coherence of the community. In many cases, unlike the situation in most chimpanzee communities, no individual will know all other members. Such dispersal renders the expensive tissue hypothesis (Aiello & Wheeler 1995) paradoxical. If a higher quality, meat-rich diet was an adaptation associated with increase in brain size, then increased carnivory posed its own cognitive problems for tracking social relationships.

Why do hunter-gatherers sustain inter-band links in the community?

(a) Inter-access between territories

The band territory rarely enables self-sufficiency in subsistence resources. ‘Aborigines, and most other hunter-gatherers, live in environments subject to great fluctuations in the weather and in the abundance of game and plant resources’ (Peterson & Long 1986, 143). When water fails at one waterhole, during drought, people can join relatives or exchange-partners at other waterholes. Rain falls unevenly in the deserts of Australia and southern Africa, and after rain everyone converges on the fortunate area, to exploit its plant foods (see Layton 1986, 26, 34–35; Western Desert—Myers 1986, 183). In the Kalahari, drought occurs two out of five years (p.103) and is severe in one year out of four, but rainfall can vary by a factor of ten over a few miles (Lee 1979, 352). Mutual insurance against local drought was one of the main reasons for maintaining inter-band links among the G/wi (Silberbauer 1981, 459). A well-rounded set of partners is better than a few promising ones; the variety of resources in Ju/'hoansi hxaro partners’ areas is more important than distance (Wiessner 1982, 74, 76). Even in the monsoon zone of northern Australia, Tiwi bands invite others to share local concentrations of resources (Goodale 1986, 201). Among the Inuit and Northwest Coast Native Americans, there was a seasonal fluctuation between dispersed bands in summer and community aggregations in winter. The Cree practised the opposite pattern, with summer aggregation, traditionally at good fishing locations, and winter dispersal into separate bands for hunting.

In more fine-grained environments such as equatorial forest (Bahuchet 1992, 207–210) the function of the community is less clear. Through kinship, an Aka has access to the territories of ten different lineages, and Aka travel furthest to visit relatives (Hewlett et al. 1982, 427), but it is not clear what benefits these kin networks provide. The Mbuti describe a lack of relatives in other bands as ‘“walking emptily” thus emphasizing the necessity to have friends elsewhere for travelling to’ (Bahuchet 1992, 217, quoting Turnbull 1965).

(b) Dispute avoidance

Woodburn (1982) argued that the desire to avoid disputes and overbearing would-be leaders was the main reason for movement between bands among the Hadza. Turnbull and Abruzzi reach the same conclusion with regard to movement between Mbuti bands (Abruzzi 1980; Turnbull 1965, 106, 223). Equalizing band size may be an underlying consideration. A newly married Mbuti couple's residence is usually based on the relative size of the spouses’ home bands (Turnbull 1965, 219).

How often will any two members of the community meet?

The entire community can generally assemble, if at all, only during times and at places where resources are exceptionally dense. If the Northwest Coast winter village can be equated with the community (see above), this is a rare example among recent hunter-gatherers of the whole community assembling in one place, relying in this case on stored food. The Cree form summer aggregations of several hundred, but these are probably facilitated by White Canadian stores (Feit 1983, 417; Rogers 1972, 91). Before the fur trade, summer aggregations met at favourable fishing spots (p.104) for weddings and feasts (Rogers 1972, 107, 123). The winter aggregations of coastal Inuit are undoubtedly traditional, and dictated by seal hunting on sea ice (see above), but aggregations of 50 to 200 were drawn from a community of c. 500 (Damas 1969, 52).

Among the Ache, up to half the community (i.e. up to 250 people) might assemble in one place once every two to three years for ritualized club fights. After six weeks the assembly dispersed, with many changing band membership and new sexual relations established (Hill & Hurtado 1996, 73). A similar ritual is performed by all bands within the Nukak munu (Politis 2007, 82), but this is a smaller group of c. 125 people. The Aka are divided into clusters of bands which live on the same hunting trail. Camps on the same trail meet for several days at a time to perform a ceremony and enable more efficient net-hunting (Bahuchet 1992, 215, 226, 230).

In harsher environments, assemblies are infrequent and short-lived. In good summers, two G/wi bands can combine for a couple of weeks to socialize, exchange news and commodities, and play games. A maximum of 150 people can camp together for short periods (Silberbauer 1981, 459). In the Western Desert, groups of 200 assembled for ceremonies, but could only do so in years of good rainfall and restricted to a period of about two weeks (Layton 1986, 35–37).

How are inter-band links sustained?

(a) Reciprocal gift-giving

Mauss (1954 [1925]) argued that gift exchange was fundamental to human social organization and it may indeed be that exchange enabled the human communities to persist even as their members dispersed over far larger areas than are occupied by chimpanzee communities. In many hunter-gatherer societies, people constantly give each other gifts. Given the impossibility, among most hunter-gatherer communities, of all assembling in one place, the hunter-gatherer community is generally sustained by overlapping networks of kinship and friendship among members (e.g. Aka—Bahuchet 1992, 232; G/wi—Silberbauer 1972, 305; 1981, 463; Lehmann et al. ch. 4 this volume). Among the Netsilik, namesakes exchanged gifts that had to be identical, such as a knife for a knife; the exchanges were not carried out for material benefit but ‘to give expression to…enduring friendliness’ (Balikci 1970, 139). Among the Pintupi (central Australia), gifts are tokens of friendship providing a moral basis for continued and on-going co-residence and cooperation (Myers 1988; cf. (p.105) Tonkinson 1991, 53). The Ju/'hoansi (Kalahari) exchange system called hxaro, which maintains friendships between people in different bands, is well known. Wiessner (1982, 66) described hxaro partnership as ‘a bond of friendship accompanied by mutual reciprocity and access to resources’ (cf. Marshall 1976, 367). Women play the main part in maintaining these partnerships, going on long journeys to visit hxaro partners and giving them ostrich-shell necklaces, water carriers, etc. (Marshall 1976, 363).

(b) Marriage

The combination of band exogamy and community endogamy also underpins inter-band movement. Among the Ndele Aka, men range almost twice as widely as women; the most wide-ranging are young Aka bachelors who leave their camp as individuals (Hewlett et al. 1982, 425; cf. Bahuchet 1992, 222, 233). ‘The main purpose of interband visits’ among the Mbuti ‘is the extension of the social horizon, particularly necessary in view of the preferred [band] territorial exogamy’ (Turnbull 1965, 96–97). Lee (1976, 72) writes that the single Ju/'hoansi male travels more widely and more frequently than the single female, in search of casual liaisons and marriages. In the Western Desert, young men were taken on long expeditions to visit distant bands by their future fathers-in-law (Layton 1995).

(c) Classificatory kinship

Among the Yankunytjatjara, members of neighbouring bands were addressed as brother and sister, and they frequently foraged together. Marriage was expected to take place between members of more distant bands separated by at least 70 km, who were addressed as brother—and sister-in-law and who provided refuge during local drought (Layton 1995). Relationships modelled on kinship are extended among the Ju/'hoansi and Netsilik through the use of namesakes, where two individuals with the same personal name treat each other as siblings (Barnard 1992, 265–281; Balikci 1970). In Australia, ‘section systems’ assign everyone at birth to a category that determines their ‘kin’ relationship to all other members of the community (see e.g. Layton 1999; Read ch. 10 this volume).

(d) Language

The human community sometimes corresponds to the group speaking a distinctive dialect or language, as is the case among the G/wi, and !Ko of the Kalahari (Cashdan 1983, 54; Silberbauer 1972, 273). Community and language or dialect commonly coincide in Australia (Dixon 1997). This (p.106) probably arises from the greater intensity of social interaction within the community, although in some cases the correspondence may arise because the community is isolated from other hunter-gatherers by surrounding farmers or herders, e.g. the Hadza. Elsewhere, the language group is larger than the community. This may be because the region has been colonized by its current hunter-gatherer population relatively recently (e.g. Inuit, Australian Western Desert: McConvell & Evans 1997) while in extreme cases separate dialects may emerge at the band level (Sutton 2001). Mellars (1998) argued that language embodying tense and the subjunctive would be necessary to talk about social relationships with people displaced in time and space characteristic of hunter-gatherers (‘Should this drought continue I may visit Y, whom we allowed to camp with us three years ago’). Terms universal to modern human language include words equivalent to ‘now’, ‘before’, ‘after’, ‘here’ and ‘far’, ‘the same’, ‘other/else’ and ‘because’ (Wierzbicka 1996). Language was probably essential to enable the distinctively human form of fission-fusion society we have outlined.

Conclusions

Our analysis began with the observation that humans and chimpanzees share a form of social organization that is unusual among primates, but we continued by identifying a number of differences between the distinctive patterns of fission and fusion in chimpanzees and human hunter-gatherers. The hunter-gatherer community is larger than the chimpanzee community, yet human population densities are lower. Parties formed during fission phases last longer among humans, and the bands characteristic of hunter-gatherer society are far more durable than the fusion phases of chimpanzee social behaviour. We found that the greater contribution of hunting to the human diet had a number of important consequences. The predatory nature of human subsistence compels a lower population density than chimpanzees can sustain, while the larger size, and frequently unpredictable occurrence of large game, is efficiently exploited by sharing at band camps. Band residence confers a number of other advantages (child-care, territorial prerogatives, information exchange), but these may conflict. Male and female strategies must be assessed within a lifetime perspective.

The persistence of the larger community can be explained by the need to keep membership options in a number of bands open. The persistence (p.107) of the community is facilitated by a number of distinctively human traits: gift-giving, inter-band marriage, classificatory kinship and language. The idea that hunting was crucial to human evolution gained a bad press as a result of the exaggerated claims made by early writers (Ardrey 1967; Dart 1959). We argue that increased carnivory during the course of human evolution may indeed be responsible for some of the most striking differences between human and chimpanzee social behaviour: not aggression and territoriality, but rather sharing, cooperation and exchange.

Note. The stimulus for this chapter arose from two events: the Wenner-Gren funded workshop on fission-fusion societies organized by Filippo Aureli following the 2004 meeting of the International Primatological Society (see Aureli et al. 2008) and a 2006 workshop organized by Robin Dunbar under the auspices of the British Academy Centenary ‘Lucy to Language’ Research Project. We thank Russell Hill for substantial help with creating the box plots, also Filippo Aureli, Catherine O'Hara, Jamie Tehrani and Jeremy Kendall for comments on draft versions, and Peter Sutton for references to language boundaries.

References

Bibliography references:

Abruzzi, W. S. (1980). Flux among the Mbuti Pygmies of the Ituri Forest. In: E. B. Ross (ed.) Beyond the myths of culture, pp. 3–31. New York: Academic Press.

Aiello, L. & Wheeler, P (1995). The expensive-tissue hypothesis: the brain and digestive system in human and primate evolution. Current Anthropology 36: 199–221.

Altman, J. (1987). Hunter-gatherers today. Canberra: Aboriginal Studies Press.

Altman, J. & Peterson, N. (1988). Rights to game and rights to cash among contemporary Australian hunter-gatherers. In: T. Ingold, D. Riches & J. Woodburn (eds) Hunters and gatherers: property, power and ideology, pp. 75–94. Oxford: Berg.

Andrews, E. (1989). The Akulmiut: territorial dimensions of a Yup'ik Eskimo society. PhD thesis, University of Alaska Fairbanks.

Ardrey, R. (1967). The territorial imperative: a personal enquiry into the animal origins of property and nations. London: Collins.

Aureli, F., Schaffner, C., Boesch, C., Bearder, S., Call, J., Chapman, C. et al. (2008). Fission-fusion dynamics: new research frameworks. Current Anthropology 49: 627–654.

Bahuchet, S. (1992). Spatial mobility and access to resources among African pygmies. In: M. Casimir & A. Rao (eds) Mobility and territoriality: social and spatial boundaries among foragers, fishers, pastoralists and peripatetics, pp. 205–257. New York: Berg.

Balikci, A. (1970). The Netsilik Eskimo. Garden City, NY: Natural History Press.

Barnard, A. (1992). Hunters and herders of Southern Africa: a comparative survey. Cambridge: Cambridge University Press.

(p.108) Barton, R. (2006). Primate brain evolution: integrating comparative, neurophysiological and ethological data. Evolutionary Anthropology 15: 224–236.

Bhanu, A. (1992). Boundaries, obligations and reciprocity: levels of territoriality among the Cholanaickan of South India. In: M. Casimir & A. Rao (eds) Mobility and territoriality: social and spatial boundaries among foragers, fishers, pastoralists and peripatetics, pp. 29–54. New York: Berg.

Binford, L. (1976). Nunamiut ethnoarchaeology. New York: Academic Press.

Binford, L. (1980). Willow smoke and dogs’ tails: hunter-gatherer settlement systems and archaeological site formation. American Antiquity 45: 4–20.

Boas, F. (ed. H. Codere) (1966). Kwakiutl ethnography. Chicago: University of Chicago Press.

Boesch, C. & Boesch, H. (1989). The hunting behaviour of wild chimpanzees in the Taï National Park. American Journal of Physical Anthropology 78: 547–573.

Boesch, C. & Boesch-Achermann, H. (2000). The chimpanzees of the Tai forest. Oxford: Oxford University Press.

Burch, E. (1988). Modes of exchange in north-west Alaska. In: T. Ingold, D. Riches & J. Woodburn (eds) Hunters and gatherers: property, power and ideology, pp. 95–109. Oxford: Berg.

Cane, S. (1990). Gugadja of Great Sandy Desert, NW Western Australia. In: B. Meehan & N. White (eds) Hunter-gatherer demography past and present (Oceania Monograph 39), pp. 149–159. Sydney: University of Sydney Press.

Cashdan, E. (1983). Territoriality among human foragers: ecological models and an application to four bushman groups. Current Anthropology 24: 47–66.

Codere, H. (1950). Fighting with property. New York: Augustin.

Damas, D. (1969). Environment, history, and central Eskimo society. Contributions to anthropology: ecological essays. Proceedings of conference on cultural ecology. Ottawa: National Museum of Canada Bulletin 230.

Dart, R. (1959). Adventures with the missing link. London: Hamilton.

Dixon, R. (1997). The languages of Australia. Ann Arbor: University of Michigan Press.

Drucker, P. (1965). Cultures of the North Pacific coast. San Francisco: Chandler.

Drucker, P. & Heizer, R. (1967). To make my name good. Berkeley: University of California Press.

Dunbar, R. (1993). Co-evolution of neocortical size, group size and language in humans. Behavioural and Brain Sciences Evolution 16: 681–735.

Emmons, G. (1991). The Tlingit Indians, ed. F. de Laguna. Seattle: University of Washington Press.

Endicott, K. M. (1979). Batek Negrito religion: the world-view and rituals of a hunting and gathering people. Oxford: Clarendon Press.

Endicott, K. M. (1988). Property, power and conflict among the Batek of Malaysia. In: T. Ingold, D. Riches & J. Woodburn (eds) Hunters and gatherers: property, power and ideology, pp. 110–127. Oxford: Berg.

Feit, H. (1983). Negotiating recognition of Aboriginal rights. In: N. Peterson & M. Langton (eds) Aborigines, land and land rights, pp. 416–438. Canberra: Aboriginal Studies Press.

Finlayson, H. (1935). The red centre. Sydney: Angus & Robertson.

(p.109) Foley, R. & Lee, P. (1989). Finite social space, evolutionary pathways, and reconstructing hominid behaviour. Science 243: 901–906.

Gibson, M. & Mace, R. (2005). Helpful grandmothers in rural Ethiopia. Evolution and Human Behaviour 26: 469–482.

Goodale, J. (1986). Production and reproduction of key resources among the Tiwi. In: N. Williams & E. Hunn (eds) Resource managers: North American and Australian hunter-gatherers, pp. 197–210. Canberra: Aboriginal Studies Press.

Gould, R. (1969). Yiwarra: foragers of the Australian desert. London: Collins.

Gubser, N. (1965). The Nunamiut Eskimos, hunters of caribou. New Haven, CT: Yale University Press.

Guemple, L. (1988). Teaching social relations to Inuit children. In: T. Ingold, D. Riches and J. Woodburn (eds) Hunters and gatherers: property, power and ideology, pp. 131–149. Oxford: Berg.

Gurven, M. & Hill, K. (1997). Comment on Hawkes et al. 1997. Current Anthropology 38: 566–567.

Halperin, S. (1979). Temporary association patterns in free ranging chimpanzees: an assessment of individual grouping preferences. In: D. A. Hamburg & E. M. McCown (eds) The great apes, pp. 491–499. Menlo Park, CA: Benjamin/Cummings.

Harako, R. (1981). The cultural ecology of hunting behaviour among Mbuti pygmies. In: R. V. O. Harding & G. Teleki (eds) Omnivorous primates, pp. 499–555. New York: Columbia University Press.

Hart, C. & Pilling, A. (1960). The Tiwi of north Australia. New York: Holt, Rinehart & Winston.

Hart, C., Pilling, A. & Goodale, J. (1988). The Tiwi of North Australia, 3rd edn. New York: Holt, Rinehart & Winston.

Hawkes, K. (1990). Why do men hunt? Benefits from risky choices. In: E. Cashdan (ed.) Risk and uncertainty in tribal and peasant economies, pp. 145–166. Boulder, CO: Westview Press.

Hawkes, K. & Bliege Bird, R. (2002). Showing off, handicap signaling and the evolution of men's work. Evolutionary Anthropology 11: 58–67.

Hawkes, K., O'Connell, J. F. & Blurton Jones, N. G. (1997). Hadza women's time allocation. Current Anthropology 38: 551–577.

Hayden, B. (1981). Subsistence and ecological adaptations of modern hunter/gatherers. In: R. Harding & G. Teleki (eds) Omnivorous primates: gathering and hunting in human evolution, pp. 344–421. New York: Columbia University Press.

Hewlett, B. (1991). Demography and childcare in pre-industrial societies. Journal of Anthropological Research 47: 1–37.

Hewlett, B., van der Koppel, J. & Cavalli-Sforza, L. (1982). Exploration ranges of Aka pygmies of the Central African Republic. Man (n.s.) 17: 418–430.

Hill, K. & Hurtado, M. (1996). Ache life history: the ecology and demography of a foraging people. Hawthorne, NY: Aldine de Gruyter.

Hunn, E. (1986). Mobility as a factor limiting resource use in the Columbia Plateau of North America. In: N. Williams & E. Hunn (eds) Resource managers: North American and Australian hunter-gatherers, pp. 17–43. Canberra: Aboriginal Studies Press.

Hunn, E. (1994). Place-names, population density and the magic number 500. Current Anthropology 35: 81–85.

(p.110) Ichikawa, M. (2005). Food sharing and ownership among central African huntergatherers: an evolutionary perspective. In: T. Widlok & W. G. Tadesse (eds) Property and equality, vol. 1: Ritualisation, sharing, egalitarianism, pp. 151–164. Oxford: Berg.

Jones, R. (1980). Hunters in the Australian coastal savanna. In: D. Harris (ed.) Human ecology in savanna environments, pp. 107–146. London: Academic Press.

Jordan, P. (2003). Material culture and sacred landscape: the anthropology of the Siberian Khanty. Walnut Creek, CA: AltaMira.

Kaplan, H. & Hill, K. (1985a). Hunting ability and reproductive success among male Ache foragers. Current Anthropology 26: 131–133.

Kaplan, H. & Hill, K. (1985b). Food sharing among Ache foragers: tests of explanatory hypotheses. Current Anthropology 26: 223–246.

Kaplan, H., Hill, K. & Hurtado, M. (1990). Risk, foraging and food sharing among the Ache. In: E. Cashdan (ed.) Risk and uncertainty in tribal and peasant economies, pp. 107–143. Boulder, CO: Westview Press.

Keen, I. (1982). How some Murngin men marry ten wives. Man (n.s.) 17: 620–642.

Kelly, R. (1995). The foraging spectrum. Washington, DC: Smithsonian Institution Press.

Layton, R. (1983). Ambilineal descent and Pitjantjatjara rights to land. In: N. Peterson & M. Langton (eds) Aborigines, land and land rights, pp. 15–32. Canberra: Aboriginal Studies Press.

Layton, R. (1986). Uluru, an Aboriginal history of Ayers Rock. Canberra: Aboriginal Studies Press.

Layton, R. (1995). Relating to the country in the Western Desert. In: E. Hirsch & M. O'Hanlon (eds) The anthropology of landscape: perspectives on place and space, pp. 210–231. Oxford: Clarendon Press.

Layton, R. (1999). The Alawa totemic landscape: economy, religion and politics. In: P. Ucko & R. Layton (eds) The archaeology and anthropology of landscape, pp. 219–239. London: Routledge.

Lee, R. B. (1976). !Kung spatial organisation: an ecological and historical perspective. In: R. B. Lee & I. DeVore (eds) Kalahari hunter-gatherers: studies of the !Kung San and their neighbours, pp. 74–97. Cambridge, MA: Harvard University Press.

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

McConvell, P. & N. Evans (eds) (1997). Archaeology and linguistics. Melbourne: Oxford University Press.

Marlowe, F. (2001). Male contribution to diet and female reproductive success among foragers. Current Anthropology 42: 755–760.

Marshall, L. (1976). Sharing, talking and giving. In: R. Lee & I. DeVore (eds) Kalahari hunter-gatherers, pp. 350–371. Cambridge, MA: Harvard University Press.

Mauss, M. (1954 [1925]). The gift. London: Cohen & West.

Meehan, B. (1982). Shell bed to shell midden. Canberra: Aboriginal Studies Press.

Meggitt, M. (1962). Desert people. Sydney: Angus & Robertson.

Mellars, P. (1998). Neanderthals, modern humans and the archaeological evidence for language. In: N. G. Jablonski & L. Aiello (eds) The origin and diversification of language, pp. 89–115. San Francisco: Memoirs of the California Academy of Science.

(p.111) Myers, F. (1986). Always ask: resource use and land ownership among Pintupi Aborigines. In: N. Williams & E. Hunn (eds) Resource managers: North American and Australian hunter-gatherers, pp. 173–195. Canberra: Aboriginal Studies Press.

Myers, F. (1988). Burning the truck and holding the country. In: T. Ingold, D. Riches & J. Woodburn (eds) Hunters and gatherers: property, power and ideology, pp. 52–74. Oxford: Berg.

Newton-Fisher, N. (2007). Chimpanzee hunting behaviour. In: W. Henke & I. Tattersall (eds) Handbook of paleoanthropology, pp. 1295–1320. New York: Springer.

Newton-Fisher, N. E., Notman, H. & Reynolds, V. (2002). Hunting of mammalian prey by Budongo Forest chimpanzees. Folia Primatologica 73(5): 281–283.

O'Connell, J. F., Hawkes, K., Lupo, K. D. & Blurton Jones, N. G. (2002). Male strategies and Plio-Pleistocene archaeology. Journal of Human Evolution 43: 831–872.

Peterson, N. (1975). Hunter-gatherer territoriality: the perspective from Australia. American Anthropologist 77: 53–68.

Peterson, N. & Long, J. (1986). Australian territorial organisation (Oceania monograph 30). Sydney: University of Sydney Press.

Politis, G. (2007). Nukak: ethnoarchaeology of an Amazonian people. Walnut Creek, CA: Left Coast Press.

Radcliffe-Brown, A. R. (1930). The social organisation of Australian tribes. Oceania 1: 34–63.

Reynolds, V. (2005). The chimpanzees of the Budongo Forest. Oxford: Oxford University Press.

Richardson, A. (1986). The control of productive resources on the Northwest Coast of North America. In: N. Williams & E. Hunn (eds) Resource managers: North American and Australian hunter-gatherers, pp. 93–112. Canberra: Aboriginal Studies Press.

Rival, L. (2002). Trekking through history. New York: Columbia University Press.

Rogers, E. S. (1972). The Mistassini Cree. In: M. G. Bicchieri (ed.) Hunters and gatherers today, pp. 90–137. New York: Holt, Rinehart & Winston.

Rohner, R. & Rohner, E. (1970). The Kwakiutl: Indians of British Columbia. New York: Holt, Rinehart & Winston.

Sear, R., Mace, R. & McGregor, I. A. (2000). Maternal grandmothers improve nutritional status and survival among children in rural Gambia. Proceedings of the Royal Society B 267: 1641–1647.

Scott, C. (1988). Property, practice and Aboriginal rights among Quebec Cree hunters. In: T. Ingold, D. Riches & J. Woodburn (eds) Hunters and gatherers: property, power and ideology, pp. 35–51. Oxford: Berg.

Silberbauer, G. (1972). The G/wi bushmen. In: M. G. Bicchieri (ed.) Hunters and gatherers today, pp. 271–326. New York: Holt, Rinehart & Winston.

Silberbauer, G. (1981). Hunter/gatherers of the central Kalahari. In: R. S. O. Harding & G. Teleki (eds) Omnivorous primates, pp. 455–498. New York: Columbia University Press.

Smith, E. A. (1991). Inujjuamiut foraging strategies. New York: Aldine de Gruyter.

Smith, R. & Jungers, W. (1997). Body mass in comparative primatology. Journal of Human Evolution 32: 523–559.

Stanner, W. (1965). Aboriginal territorial organisation. Oceania 36: 1–26.

(p.112) Sterck, E. H. M., Watts, D. P. & van Schaik, C. P. (1997). The evolution of female social relationships in nonhuman primates. Behavioural Ecology and Sociobiology 41: 291–309.

Steward, J. (1936). The economic and social basis of primitive bands. In: R. H. Lowie (ed.) Essays on anthropology in honour of Alfred Louis Kroeber, pp. 311–350. Berkeley: University of California Press.

Stiner, M. (2002). Carnivory, coevolution and the geographic spread of the genus Homo. Journal of Archaeological Research 10: 1–63.

Sutton, P. (2001). Talking language. In: J. Simpson, D. Nash, M. Laughren, P. Austin & B. Alpher (eds) Forty years on: Ken Hale and Australian languages, pp. 453–464. Canberra: Australian National University Pacific Linguistics.

Tonkinson, R. (1991). The Mardu Aborigines. New York: Holt, Rinehart & Winston.

Trivers, R. (1971). The evolution of reciprocal altruism. Quarterly Review of Biology 46: 35–57.

Turnbull, C. (1962). The forest people. New York: Simon & Schuster.

Turnbull, C. (1965). Wayward servants: the two worlds of the African pygmies. Westport, CT: Greenwood Press.

van Schaik, C. P. (1989). The ecology of social relationships among female primates. In: V. Standen & R. A. Foley (eds) The behavioural ecology of humans and other mammals, pp. 195–218. Blackwell: Oxford.

Warner, W. (1958). A black civilisation. New York: Harper.

White, W., Meehan, B., Hiatt, L. & Jones, R. (1990). Demography of contemporary hunter gatherers: lessons from Central Arnhem Land. In: B. Meehan & N. White (eds) Hunter-gatherer demography past and present (Oceania Monograph 39), pp. 171–185. Sydney: University of Sydney Press.

Wierzbicka, A. (1996). Semantics: primes and universals. Oxford: Oxford University Press.

Wiessner, P. (1982). Risk, reciprocity and social influences on !Kung San economics. In: E. Leacock & R. Lee (eds) Politics and history in band societies, pp. 61–84. Cambridge: Cambridge University Press.

Wiessner, P. (2002). Hunting, healing and hxaro exchange: a long-term perspective in !Kung (Ju/'hoansi) large-game hunting. Evolution and Human Behaviour 23: 407–436.

Williams, N. (1986). The Yolngu and their land. Canberra: Aboriginal Studies Press.

Wilson, M., Wallauer, W. & Pusey, A. (2004). New cases of intergroup violence among chimpanzees in Gombe National Park, Tanzania. Evolutionary Anthropology 5: 46–57.

Winterhalder, B. (1990). Open field, common pot. In: E. Cashdan (ed.) Risk and uncertainty in tribal and peasant economies, pp. 67–87. Boulder, CO: Westview Press.

Winterhalder, B. (1996). Social foraging and the behavioural ecology of intragroup resource transfers. Evolutionary Anthropology 5: 46–57.

Wobst, M. (1974). Boundary conditions for Palaeolithic social systems: a simulation approach. American Antiquity 39: 152–178.

Woodburn, J. (1972). Ecology, nomadic movement and the composition of the local group. In: P. Ucko, R. Tringham & G. W. Dimbleby (eds) Man, settlement and urbanism, pp. 193–206. London: Duckworth.

(p.113) Woodburn, J. (1982). Egalitarian societies. Man (n.s.) 17, 431–451.

Wrangham, R. W. (2000). Why are male chimpanzees more gregarious than mothers? A scramble competition hypothesis. In: P. Kappeler (ed.) Primate males, pp. 248–258. Cambridge: Cambridge University Press.

Wrangham, R. W., Gittleman, J. L. & Chapman, C. A. (1993). Constraints on group size in primates and carnivores: population density and day-range as assays of exploitation competition. Behavioral Ecology and Sociobiology 32(3): 199–209.

Yellen, J. (1976). Settlement patterns of the !Kung: an archaeological perspective. In: R. B. Lee & I. DeVore (eds) Kalahari hunter-gatherers: studies of the !Kung San and their neighbours, pp. 47–72. Cambridge, MA: Harvard University Press.

Notes:

Proceedings ofthe British Academy 158, 83–113. © The British Academy 2010.

(1) Found in only eight species of non-human primate: South American spider monkeys (two species) and muriquis; Asian apes: orangutans, African apes: chimpanzees and bonobos; African cercopithicines: gelada and hamadryas baboons.

(2) Australian anthropologists usually distinguish between band and clan, where the co-residential group is the band, the territorial group the clan (Stanner 1965). The distinction makes most sense where individuals have a presumptive right to a parent's territory from birth, and least sense where rights are only established through residence. Practice generally lies between these extremes, with clans most clearly existing in northern Australia and effectively absent in the Western Desert.