yanomamo reader for HRAF (YANREADG)

Yanomamö, Varying Adaptations of Foraging Horticulturalists 

Originally prepared for Just in Time Anthropology series, Prentice Hall and Simon & Schuster supplemental readings for Ember and Ember's Anthropology, 8th edition.

by Raymond Hames (c) 1995

Introduction

The documentation of behavioral variation in cultural anthropology is key to scientific description and explanation. Early ethnographers were content to describe modal patterns of behavior to give readers an idea of what was typical, expected, or conventional in a given culture. In order to understand variation in cultural practices anthropologists engaged in cross-cultural comparison with individual societies as data points or exemplars of particular traits. While comparative or cross-cultural approaches have been enormously productive, they are not the only useful approach to a scientific understanding of cultural variation. Within each society individuals or even whole regions may vary enormously in terms of how they conduct their social, economic, and political lives. Accurately documenting this intracultural variation and attempting to associate it with hypothetical factors is an important alternative approach. This is not to say that intracultural comparisons are superior or in competition with cross-cultural approaches. In fact, I would expect them to compliment each other. For example, one might demonstrate cross-culturally that warfare is strongly associated with a particular environmental variable. This then might lead us to test that proposition within a particular cultural group if that environmental variable had enough variation.

The goal of this chapter is to describe variation in Yanomamö economic activities at cross-cultural, regional, and individual comparative levels. I will first compare Yanomamö horticultural adaptation to other horticultural groups. The striking finding here is that compared to other horticulturalists the Yanomamö spend an enormous amount of time in the foraging activities of hunting, gathering, and fishing. In many ways they behave like hunters and gathers, peoples without agriculture. I will then turn to a regional comparison of Yanomamö economic adaptations by comparing how highland and lowland Yanomamö adapt to the rain forest. Here we will find that highland Yanomamö are much more dedicated to a sedentary horticultural life than lowland Yanomamö. Finally, I will turn to an analysis of individual Yanomamö to describe how sex and age determine the division of labor and the amount of time that individuals work.

Geography, Demography, and Environment

The Yanomamö are a tribal population occupying the Amazonian border between Venezuela and Brazil. In Venezuela, the northern extension of the Yanomamö is delimited to the north by headwaters of Erebato and Caura rivers, east along the Parima mountains, and west along the Padamo and Mavaca in a direct line to the Brazilian border. In Brazil, they concentrate themselves in the headwaters of the Demini, Catrimani, Araca, Padauari, Uraricoera, Parima, and Mucajai rivers. In both countries the total area inhabited by the Yanomamö is approximately 192,000 km2. Dense tropical forest covers most of the area. Savannas are interspersed in forests at high elevations. In general, the topography is flat to gently rolling with elevations ranging from 250 to 1,200 meters.

Area Exploited

The area village members exploit in the course of their economic activities is probably best characterized as a home range. Home ranges differ from territories because they are not defended; but like territories home ranges tend to be used exclusively by a single group. This exclusiveness is not determined by force but by the following simple economic considerations. Important food resources tend to be evenly distributed in the tropical forest. When Yanomamö establish a new village they intensively exploit and deplete resources near the village. Through time, they must travel greater distances where higher return rates compensate for greater travel costs needed to reach areas of higher resource density. At a certain point they will begin to reach areas that are exploited by neighboring villages and if they were to travel still further they would begin to enter areas close to neighboring villagers that have been depleted. At this point, is not economic to travel further since the costs of gaining resources increases (more travel time) while resource density decreases. Thus the borders of home ranges are established with some overlap with the home ranges of adjacent villages.

Where warfare is intensive home ranges may become more like territories if enemy villages are neighbors. It is difficult to determine whether resource areas being defended so that foragers Yanomamö are not threatened by the presence of enemies or whether a powerful village decides to press its advantage over a weaker neighbor by expanding its home range and transforming it into a territory. The way in which Yanomamö verbally rationalize their reasons for warring complicates this matter further. Yanomamö may claim that they go to war in order to avenge an insult, a previous killing, an abduction of a woman, or the illness causing spells cast by a neighboring shaman. Therefore, Yanomamö explain war in terms of vengeance for harm caused by an enemy. The problem here is that neighboring villages invariably have members who have done one of more of the above to a neighbor or a neighbor's ancestor. Why some past depredations are ignored or acted upon may be determined by economic (territorial) and political factors (opponent's strength or perceived threat).

Table 1 displays home range estimates for a number of Yanomamö villages. Differences in home range may be the result of ecological differences in resource density or the distribution of neighboring villages. In addition, Colchester's large estimates (from a highland village) differ markedly from the rest and may be the result of how home range was defined. With the exception of Colchester, all other ethnographers defined home range as the area regularly exploited in the pursuit of wild resources. Colchester's definition includes the maximum distance ever traveled to gain resources.

 Demography and Settlement Pattern of Yanomamö Villages
 

Although ethnographers have done extensive and excellent demographic research on Venezuelan Yanomamö, a complete census for Venezuelan and Brazilian Yanomamö is lacking. Current estimates are 12,500 and 8,500 Yanomamö in Venezuela and Brazil, respectively, for a total of 21,000. However, the figures for Brazil may be significantly less because of epidemics and white-Yanomamö fighting caused by incursions of Brazilian gold miners starting about 1987. I discuss this problem later. In Venezuela and Brazil there are approximately 363 villages ranging in size from 30 to 90 residents each. But some Venezuelan villages in the Mavaca drainage may reach 200 or more. Chagnon (1974) provides evidence that warfare intensity is associated with village size: where warfare is intense, villages are large. People are forced to associate in large villages to both deter attackers and enable themselves to mount effective counterattacks. Population density ranges from about 6.7 km2/person to 33.5 km2/per person.

Anthropologists consider stable settled life one of the important consequences of the agricultural revolution. Although the Yanomamö are agriculturists, villages are unstable in duration, location, and membership. A typical Yanomamö village (shabono) is in the shape of a giant circular "lean-to". Each house or apartment section of a village has a roof and back wall but no front or side walls. Individual family lean-to's are joined in a circle. When a Yanomamö sits in his hammock and looks left or right he sees his next door neighbors and if she looks straight ahead she sees a broad plaza and the dwelling places of neighbors on the other side of the village. A village structure rarely last more than a few years before the roof thatch beings to rot and the entire village is filled with vermin. On such occasions, a new village may be constructed adjacent to the old one.

Yanomamö villages are relocated about every five years for economic and political reasons. The practice of shifting cultivation forces the Yanomamö to use extensive tracks of land. This is because garden land is used for only two to three years and then abandoned to the encroaching forest. Through time gardens become increasingly distant from the village. When gardens or easily accessible garden land become too distant the village may move several kilometers to be in the midst of good garden land. Raiding provides a political cause for village relocation. When a village is repeatedly raided by a more powerful enemy the entire village may be forced to relocate. Such moves are designed to put as much distance as possible between themselves and an enemy and may cause great privation due to loss of easy access to productive gardens (Chagnon, 1974).

Highlands and Lowlands

There is good reason to suspect that there are fundamental differences in environmental quality for Yanomamö who occupy highland and lowland elevations. Defining a precise metric boundary between the highlands and lowlands is impossible at this point. However, one can provisionally define highland populations as villages found in areas higher than 500 to 750 meters of elevation and occupying areas of highly dissected and hilly terrain with small fast flowing streams and occasional savannas. The lowland environment is flatter with slowly moving, larger streams and rivers. This highland-lowland distinction appears to have important implications for the fundamental economic activities of gathering, hunting, fishing, and agriculture which I summarize immediately below.

General ecological research provides considerable evidence that plant biomass and diversity are negatively correlated with altitude. Holdridge's detailed research indicates that plant diversity and forest biomass decreases with increasing altitude (Holdridge et al., 1971:674-678; Beard 1955). Ethnobotanical research by Lizot (1984:54, Table 2) shows that there are a greater variety of edible plants available to lowland groups, more plants are restricted to lowland environments, and, on average, more edible plants are available on a monthly basis for lowland groups. Evidence that highland areas are less productive for gathering is suggested in Colchester's (1984) data on a highland group: they spend less time to gathering than seven of the eight highland Yanomamö groups on which I have time allocation data (Table 2). In addition, the cultural geographer Smole (1976:41) notes a decrease in edible plants with increasing elevation. Obviously, one cannot positively conclude that gathering is more productive in lowland areas since plants differ enormously in their food value, processing costs, density, and seasonal availability. Nevertheless, we have limited data that shows that the rate of return in gathering wild forest resources is much greater in lowland than in highland areas (Colchester, 1984).

It is well-established that fish are more abundant and larger in the wider, slower moving rivers found in lower elevations (Goulding, 1980; Lowe-McConnell 1977). A comparison of Saffirio's upland and lowland sites (Saffirio and Hames, 1983) and Lizot's riverine and upstream sites (Karohi-teri and Kakashiwä, respectively, 1984:197, Table 7) reveals that groups living along large streams or rivers consume twice as much fish and other aquatic prey (frogs, caiman and crabs). In addition, these lowland groups gain fish at efficiencies (kg/hr) two to three times higher than highland groups (Hames, 1989:64, Table 6).

It is my impression (based on informant statements) and that of other Yanomamö researchers (e.g., Smole 1976:181, 227) that game animals are much less abundant in higher elevations. Colchester (1984:300, Table 2) notes that the highland Sanema Yanomamö hunt at an efficiency of 1.8 kcal/hr versus 2.8 kcal/hr for Lizot's lowland groups (see also Hames 1989:64, Table 6). A review of the ecological and biogeographical literature on altitude and animal biomass gradients suggests that huntable biomass declines with altitude. Unfortunately all of these studies make comparison across rather than within geographic regions (e.g., Eisenberg, 1980; Eisenberg et al., 1979; Terborgh and Winter, 1980).

We have no convincing comparative data to indicate significant differences between highland and lowland areas as being more useful for agricultural pursuits. Agricultural productivity is a complex interplay of many factors such as soil quality, quantity and distribution of rain, and temperature extremes. It is clear, as I show below, that there are significant differences in garden size corresponding to a highland and lowland divide. What is not clear, however, is whether these differences are the result of environmental, economic, or socio-political factors to be discussed below.

Yanomamö Economics

Economically, the Yanomamö, along with most other tribal peoples living in the tropics, are classified as shifting cultivators because most of their dietary calories come from horticultural pursuits. However, a significant amount of time is allocated to the foraging activities of hunting, gathering, and fishing. In fact, the Yanomamö allocate more time to foraging activities than they do to agriculture (see below). Their dedication to foraging is greater than any other Amazonia group and any other horticultural group that we know of (Hames, 1989). Given this huge investment in foraging activities it might be more accurately to refer to the Yanomamö as "foraging horticulturalists". In this section I describe the basic productive components of the Yanomamö economy.

Technology

Until the middle 1950s the Yanomamö relied on a locally produced "stone age" technology. This technology was dependent on locally produced non-metal resources. Since that time much of their traditional technology has been replaced by steel cutting tools (machetes and axes), aluminum pots, and other industrial items given or traded to the Yanomamö primarily by missionaries. The main impact of such introductions has been to reduce labor time and increase Yanomamö dependence on non-Yanomamö to satisfy these new needs (Hames, 1979). In many instances the Yanomamö no longer possess the skills to make or use traditional technology such as clay pots or fire drills.

Gardening

In shifting cultivation, forest is cleared with machetes, axes, and fire. The newly opened forest is then planted with plantains, root crops, and a large variety of plants that serve as relishes, medicines, and technology sources. After about two to three years of cultivation, the garden is abandoned to the encroaching forest. In most years Yanomamö add to the size of a current garden by clearing adjacent forest and cease to work old areas of the garden and let it naturally revert to scrub and later to forest. Men do nearly all the heavy work involved in clearing such as slashing the undergrowth and felling large forest trees. Men and women work together to plant the garden and women are responsible for the nearly daily trips to the garden to harvest and weed.

There is considerable variation, as Figure 2 clearly indicates, in the amount of land under cultivation per capita in Yanomamö villages. Per-capita land cultivated in highland villages is 2419 m 2 (±1247) compared to 558 m 2 (±81) for lowland villages, nearly a five-fold difference that is statistically significant (two-tailed t-test, <0.05, see inset to Figure 2). There are potential ecological and economic explanations for these differences. Garden land may not be as fertile in higher elevations (Smole 1976:36-37), or basic crops may not be as productive because of the cooler temperatures in highland areas (Simmonds, 1962). As a result, highlanders are forced to increase garden size to produce the same quantity of plantains as lowlanders. Or, as I have already argued, since foraging success (efficiency) is greater in lowland areas, lowlanders may gain a larger fraction of their diet from foraging which lessens their dependence on garden food.

There is a basic contrast in subsistence crops relied upon by highland and lowland groups which may have far reaching consequences in helping us understand their economic differences. Lowland groups rely on plantains and bananas as the basic subsistence crop while some highland groups rely more heavily on manioc, a very productive root crop. Where either crop is staple it contributes up to 40% or more of all dietary calories. This difference leads us to ask two questions: (1) why do some highland groups depend on manioc; and (2) what impact does dependence on one or the other have on the overall Yanomamö economy. Some anthropologists suggest that manioc is a recent introduction from neighboring groups such as the Ye'kwana. Where manioc has been introduced the Yanomamö have taken it up because it appears to be a 40% more efficient source of calories than plantains. However, this alleged advantage may disappear since the comparative data on efficiency do not consider processing costs. Some varieties of manioc become poisonous ("bitter") soon after harvesting and must be detoxified. In addition, many varieties require a laborious process of peeling, grating, and baking before consumption. It is perhaps significant that the only highland group in Table 2 (Sanema) allocates more time to food preparation ("Cook") than any other group. Plantains, in contrast, are easily peeled and quickly cooked by roasting or boiling. Other anthropologists make an opposite argument by suggesting that manioc was aboriginal with the Yanomamö and it was replaced by plantains (an Old World crop introduced by the Spanish more than 400 years ago) because plantains were a more efficient producer of calories. Unfortunately, these researchers have no quantitative data to back up their claim. Clearly relatively simple research could help settle this issue.

Whether or not manioc or plantains are ancient or recent introductions, dependence on one or the other has a strong impact on the Yanomamö economy. The key issue here is not one of efficiency but of reliable and predictable yield. Tropical forest peoples tend to rely on crops that can be harvested over a long time. In contrast to temperate horticulturalists, many tropical peoples who grow root crops or plantains do not harvest their entire field during a single harvest period and store the crop to tide them through seasons when crop growth is impossible or risky. Instead, tropical cultivators stagger-plant throughout the year so that what is needed can be harvested from the field every few days or week. For example, one can harvest manioc 6, 12, or 18 or more months after it is planted. This allows the manioc cultivator considerable flexibility in insuring a steady and reliable yield. Plantains, in contrast, have much less flexibility. Although Yanomamö attempt to stagger-plant plantains to gain a reliable weekly or half-weekly yield, a variety of environmental factors thwart this strategy. Dry spells can hasten maturation while prolonged wet spells slow maturation. Also, heavy winds that accompany violent thunderstorms may blow down plantains with heavy maturing racemes. The Yanomamö can salvage blown down bunches of plantains by hanging them in the village. When plantains ripen there will be a momentary glut of food but a lack of plantains in the near future when they were scheduled to mature.

Heavy dependence plantains by lowland groups may help us to understand their greater reliance on gathering compared to manioc producing highland groups. Since plantains are far less reliable than manioc, lowland Yanomamö may be forced to gather because of periodic underproduction of plantains. I provide evidence for this proposition in Table 2. The only highland group on which we have time allocation (Sanema, in the table) shows that the average adult spends only 26 min/day gathering which is the lowest figure in the table and fully one-half of the mean time allocated to this task for lowland villages.

If manioc is more reliable than plantains we must ask why don't plantain growing lowlanders grow more manioc than they currently do? Since lowland gathering is about twice as efficient as highland gathering (Colchester, 1984; measured by of kilocalories of food per caloric unit of work), it may mean that unreliable plantain production is buffered by highly productive gathering. Furthermore, Table 2 shows that overall work effort for the highland Sanema is essentially identical to the mean level of work for lowland villages. I suggest that this is clear evidence that plantain gardening coupled with a high reliance on gathering causes no discernible hardship in overall work effort.

Foraging

Foraging is the simple extraction of resources from the environment without any attempt to modify the environment (as in agriculture) to increase the yield of that which is harvested. Hunting, gathering, and fishing are the basic foraging activities. Foraging is most ancient technique humans use to exploit the environment and an adaptation that humans share with all other animals. What is interesting about Yanomamö foraging is the large amount of time they allocate to it. As time allocation statistics in Table 2 indicate, Yanomamö allocate more than twice as much time foraging as they do gardening.

As I mentioned above, given the amount of time the Yanomamö spend foraging on a daily basis, the term foraging horticulturalists might be an apt designation for their economic adaptation to the tropical forest. The logic of this designation is further reinforced by the Yanomamö practice of waiyumö. Waiyumö or trekking is camping in the forest and subsisting mostly by foraging. Trekking usually occurs in the dry season when there is an abundance of forest fruit and the dryness makes walking and camping out pleasant. The probable motivation behind most trekking is to save travel time by taking advantage of abundant vegetable resources distant from the village. However, trekking may also be stimulated by a shortage of garden food or the presence of powerful enemies. If the latter is the cause, then trekking is a kind of refuging adaptation designed for concealment against enemies. Treks may last a week to more than a month and normally include all village members. Dependence on wild resources is not total during treks since young men are sent from forest camps to gardens to harvest plantains if the wild resources are scarce.

Hunting

Hunting is the main source of dietary protein for the Yanomamö. As we shall later see in the time allocation statistics, hunting is in essence is a male activity with important social and ritual functions. Bows and arrows, which measure approximately 2 meters in length, are the main weapons of the hunt. The long arrow is not accurate beyond about 30 meters. This is of little significance in dense tropical forest where it is rare to have a clear shot at a greater distance (Hames, 1979). Skills in locating game and stalking it to a short distance are abilities that differentiate good from poor hunters. Game sought ranges from 1 kilogram birds, to 25 kilogram peccaries, and to the occasional 175 kilogram tapir (the largest terrestrial animal in South America). Yanomamö quivers contain large lancelotate tips for big game, poisoned pencil-shaped tips for monkeys, and harpoon points for birds and small terrestrial game. Because of their great a hunter can carry no more than three arrows on a hunt; however, the quiver also contains a repair kit of thread, resin, and a hand tool to repair arrows that have been damaged.

Although most hunting is done by individuals or pairs, organized group hunts occur under two important circumstances. If a hunter discovers a herd of white-lipped peccaries (sometimes numbering more than 50 individuals), he carefully notes the location and immediately returns to the village to alert other hunters who return to cooperatively hunt the herd. To prepare for a feast (reahu) organized hunting parties travel great distances and may continuously hunt for a week in order to amass a large quantity of game to provide high quality meals for visiting allies on a variety of social occasions (reahu and braiai rituals). During these excursion hunters especially seek highly esteemed game such as peccaries, turkey-like birds (Crax sp.), and monkeys.

Gathering

The harvesting of wild plant resources is an activity that includes all ages and sexes and is commonly organized by families and groups of families. Important resources include honey, palm fruits, brazil nuts, palm heart, and cashew fruit. Men specialize in the risky task of climbing trees to shake loose fruit or to sever fruit laden branches. The peach palm is especially important. It is planted in gardens but it only begins to yield several years after a garden has been abandoned and it continues to bear for a decade or more. The Yanomamö assert that peach palms are owned by those who planted them and it is not uncommon for disputes to arise over ownership. With the exception of the peach palm, the Yanomamö make little or no effort to harvest fruit trees or palms so they may be harvested on a sustained basis. The Yanomamö fell small, fruit ladened trees to make harvesting easier while they never fell forest giants such as Brazil nut or cashew trees because of the enormous labor required.

Fishing

Most Yanomamö villages occupy areas between major rivers that are crossed by small streams. Fish found here fish seldom are larger than a few kilograms. Nevertheless, fishing is widely and avidly pursued by all ages and sexes, especially in the dry season through hand catching, stream poisoning, and archery. In the dry season small streams begin to shrink leaving fish in large ponds or cutoffs. The Yanomamö will use a vegetable poison to stun fish and cause them to rise to the surface where they can be grabbed or shot with a miniature bow and arrow. Women sometimes will jointly push a long and broad palm frond through the water to herd fish towards a bank to trap them.

Marriage and Family

Yanomamö marriage rules are prescriptive such that marital partners ought to be cross-cousins. Ideally, mates are double cross cousins, a result of the practice of sister exchange. Women are typically marry soon after their first menses to men in their early twenties. Although marriage is patrilocal, a husband must live with his parents-in-law for several years performing bride service. This rule may be relaxed for high status males. Polygyny is permitted and 10-20% of all males at any time are polygynists. Ideally, polygyny is sororal and the levirate and sororate are practiced. Men and women average 2.8 marital partners during their lifetime with about 75% of those marriages ending as a result of divorce with the balance as a result of death of one of the partners (Melancon, 1982).

Monogamous or polygynous nuclear families are the rule among the Yanomamö. Deviations from this pattern occur when aged parents live closely associated with married children or when newlyweds dwell with one or the other's parents. Each family has a garden or gardens and is responsible for basic subsistence activities.

Social Organization

Each Yanomamö village is an autonomous political entity, free to make war or peace with other villages. Coalitions between villages are important: nevertheless, such coalitions tend to be fragile and ephemeral. Although the Yanomamö are an egalitarian people, age, sex, and personal accomplishments are important in status differentiation. Yanomamö men acquire high status through valor in combat, accomplished oratory, and expertise in shamanism. However, high status cannot be inherited, it must be earned. Mature men dominate positions of political authority and religious practice. Local descent groups play important roles in regulating marriages and settling disputes within the village.

Political Organization

The village headman is the dominant political leader and comes from the largest local patrilineage. When a village is large or when two local descent groups are approximately equal in size, a village may have several headmen. To lead the headman must rely on demonstrated skills in settling disputes, representing the interests of his lineage, and successfully dealing with allies and enemies. Styles of leadership vary: some headmen lead through practiced verbal skills while other resort to bullying tactics. Concerted action requires the consensus of adult males. However, an individual is free to desert from collective action if it suits him.

Social Control

Conflicts typically arise from accusations of adultery, failure to deliver a betrothed woman, personal affronts, stinginess, or thefts of coveted garden crops such as tobacco and peach palm. For men, if such conflicts move past a boisterous shouting match, a variety of graded, formal duels may occur. If a fight becomes serious, respected men may intervene to cool tempers and prevent others from participating. Frequently, duels end in a draw which allows each contestant to preserve his dignity. For women, dueling is rare. Instead, a direct attack is made by the aggrieved using hands and feet or makeshift weapons.

Conflict

Warfare or feuding is endemic among the Yanomamö. While the initial cause of a conflict may be frequently traced to a sexual or marital issue, feuds are self-perpetuating since the Yanomamö lack any formal mechanisms to prevent aggrieved parties from exacting the amount of vengeance or counter-vengeance they deem sufficient once a conflict has started. The primary vengeance unit is the lineage, but coresident non-kin have some obligation to assist since coresidence with a feuding faction is seen as implicit support of the faction by the faction's enemies. Most combat is in the form of stealthy raids. The goal is to quickly dispatch as many of the enemy as possible (who are frequently found on the outskirts of the village engaging in mundane activities), abduct nubile women if possible, and return quickly home. While the primary goal is to kill mature men believed to be responsible for a previous depredation or their patrilineal kin, unrelated co-villages may be killed if there is no safe opportunity to kill primary targets. Endemic warfare has a profound effect on politics and settlement size and location. Each village needs at least one allied village it can call upon for assistance if it is overmatched by a more powerful enemy; and village size and distance between villages tend to increase with the intensity of conflict. Peace between villages may develop if conflict has remained dormant for a long period and there is a mutual need for an alliance in the face of a common enemy. It begins with a series of ceremonially festive visits. If old antagonisms do not flare, visits may lead to joint raids and intermarriage between villages that strongly solidify an alliance. Proximity of missions and government agencies has had little impact on warfare

Major Changes

Over the last twenty years most Yanomamö have become totally dependent on outside sources of axes, machetes, aluminum cooking pots, and fish hooks and line. These metal goods have replaced much of their stone-age technology. Most of these items have come from missionaries as gifts and wages. Through mission organized cooperatives, the Yanomamö recently have begun to market baskets and arrows and some agricultural products.

Another impact missionary presence has had on the Yanomamö is a distortion of their traditional settlement pattern. Yanomamö attempt to gain easy access to mission outposts by moving their villages near to a mission. As a result, the normal spacing of about a day's walk between villages has diminished dramatically. For example, around the Salesian mission at Mavaca there are five villages and numerous small settlements within one day's walk with a total population of about 900 people. This population density is unprecedented for the Yanomamö and has led to severe depletion of wild resources. In addition, a significant fraction of that population no longer lives in traditional communal village but rather in small settlements of two to three houses occupied by a few families. Finally, missionaries have failed to gain significant numbers of Yanomamö converts to Christianity. The Yanomamö have enormous pride in their culture and have strong doubts about the authenticity or superiority of Christian beliefs.

However, the greatest change and threat to the Yanomamö are the thousands of Brazilian gold miners who have infiltrated Yanomamö territory in Brazil and who have again (August 1993) illegally entered into Venezuela and this time killed a dozen men, women, and children. The situation in Brazil is much similar to the United States in the 1800s when Whites expanded against the Native Americans. Miners bring epidemics of measles and influenza that lead to high mortality rates among the Yanomamö. Gold processing pollutes streams with mercury killing fish and ruining a village's water supply. And open warfare between miners and Yanomamö has killed numerous Yanomamö and disrupted village life.

TIME ALLOCATION

In the West, we tend to think of the work as something done away from the home for forty hours a week. In subsistence based tribal populations this sort of definition is as inadequate for them as it is for us. While it is true that the Yanomamö, for example, leave the village to travel to garden, forest, and stream to acquire resources, much work takes place in the village. But the same thing is true in the West. Driving to work, mowing the lawn, shopping for food, washing clothes, and all those other household chores that must be done are not what we would call leisure time activities. We do these tasks to maintain our material well-being. I believe that most would define leisure time activities as including dining out, going to the movies, visiting friends, and sports activities. Therefore, one can define work as all those activities we must do in order to maintain or enhance our material existence. Clearly, adults in the West work more than 40 hours per week if we use this expansive definition of work. Researchers who have investigated time allocated to work in the West show that urban European and North Americans work on average 55 to 65 hours per week, or 7.8 to 9.3 hours per day, seven days a week (Hames, 1989).

Table 2 below presents time allocation data for adults on a basic set of work activities in eight different Yanomamö villages. The table reveals that the Yanomamö work about six hours per day or 42 hours per week. This is significantly less than the 55 to 66 hours of work accomplished in modern societies. Furthermore, if we compare Yanomamö and related simple tropical horticulturalists to other economic formations (e.g., hunters and gatherers, pastoralists, agriculturists, etc.) we find that they are among the most leisured people in the world (Hames, 1989).

Although Table 2 shows little variation in overall labor time (mean 361, SD 7.88) there is considerable variation among villages in time allocated to various subsistence tasks. Much of the variation is attributable to local conditions such as season in which the researcher collected the data, the degree to which a village is associated with missionaries, special environmental condition, etc. Nevertheless, the only highland site, Sanema, shows some interesting patterns. This village allocates the most time to gardening, the least time to fishing, the second least amount of time to gathering, and the second most amount of time to hunting. Gathering and gardening times are probably related. Since the density of wild sources of plant food is lower in highlands, foraging is not as productive which leads highlanders to spend more time gardening. Related to this is the higher reliability of manioc gardening which makes gathering less of an important alternative source of vegetable foods.

The extremely low amount of time highlanders allocate to fishing and the relatively high amount of time they allocate to hunting are also related. Highlanders do little fishing because of difficulties exploiting steep and narrow highland streams. Since fishing and hunting are the only ways of gaining sufficient high quality protein to the diet and fishing is unprofitable, highlanders are forced to hunt more intensively. Another way of expressing the contrasting dependence on foraging and gardening in highland and lowland locales is to note that the highland population spends the least amount of time foraging and has the lowest ratio of foraging time to gardening time (2.13:1.0 compared to a mean of 4.43:1.0 for the lowlanders).

Division of Labor

As figure three indicates, women spend significantly more time in food preparation, fishing, gathering, and in child care than do men while men spend more time hunting than women. From what we know about the division of labor cross-culturally these differences in time allocation are not surprising. In all cultures males either dominate hunting or it is wholly a male activity. Although the data indicate that Yanomamö women do almost no hunting, some qualifications are necessary. Yanomamö women occasionally accompany men on hunting forays to act as spotters and assist in the retrieval of game. Rarely do they ever make kills while with men. However, they occasionally make fortuitous kills of their own while gathering or fishing. Such kills are made without the use of bows and arrows.

Although the data show no significant difference between men and women in gardening there are important differences in garden tasks performed. Men almost exclusively do the heavy work of felling large trees, slashing the undergrowth, and burning the resulting debris prior to planting. Both sexes share in planting while the daily tasks of weeding and harvesting fall almost exclusively to women. The pattern of men doing tasks which are dangerous and/or take them far from home is consistent with Judith Brown's model of the division of labor (Brown, 1970). Brown suggests that women tend to dominate tasks that are compatible with simultaneous child care. Such tasks are not dangerous, can be accomplished near to home, and can be interrupted and resumed with no loss of efficiency.

While Brown's model usefully captures much of the variation in the division of labor among the Yanomamö and other groups it does not explain why women who are post-menopausal or otherwise unencumbered with intensive child care do not, for example, engage in hunting or tree felling. For dangerous and arduous activities as tree felling it is probable that other models such as those that focus on physical strength differences (Murdock and Provost, 1973) or task linkages (Burton and White, 1984) are required to compliment Brown's model. However, lack of female participation in hunting may require yet another explanation. Hunting is a highly skilled activity that is not easily learned and requires frequent practice to maintain proficiency. On average, little Yanomamö boys spend 60 to 80 minutes per day playing with bows and arrows and spotting, tracking, and stalking small birds and other tiny game near the village. It may be the case that women don't hunt because they never acquired the skills necessary to become proficient hunters.

The question of whether men or women work more can only be answered if we have a reasonable definition of work. Generally, economic anthropologists define work as all of those activities required to directly maintain and enhance survival and reproduction. Thus, it includes rather obvious activities as the provisioning and preparing of food, construction and repair of tools and shelter, and the acquisition and management of fuel (or firewood, in the case of the Yanomamö. If we use this definition of work, the means in Tables 3 and 4 indicate that Yanomamö women work about 12 minutes per day more than men but the difference is not statistically significant. This finding is rather interesting since in the vast majority of horticultural tribal populations on which we have time allocation data women work significantly more than men (Hames, 1989). The only societies in which men work significantly more than women are hunters and gathers. That Yanomamö men and work approximately equally is therefore consistent with the point made earlier that they can be best characterized as foraging horticulturalists since time allocation patterns are intermediate between horticulturalists and hunters and gatherers.

Some may consider this definition unnecessarily restrictive since it ignores a task that is critical for the long-term survival of the Yanomamö, the task of child care. In order to analyze the impact of child care on overall labor time differences I must restrict my analysis to data I collected on four Yanomamö villages (Mishimishimaböwei, Rakoiwä, Krihisiwä, and Bisaasi) since none of the other studies collected such data. When I include direct child care activities (feeding, nursing, holding, etc.; see Hames, 1992), female time allocation increases by 43 minutes per day while male time allocation increases by only 8 minutes per day. If child care activities are added to conventionally defined labor then Yanomamö women work more than men.

Table 3. Yanomamö Female Time Allocation

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
   
  Table 4. Yanomamö male time allocation.

Sources for tables 3 and 4:  1=Hames, 1989; 2=Lizot, 1978; 3=Hames, in press; 4=Lhermillier, 1974; 5=Colchester, 1984;6=Good, 1989.

  Status and the Allocation of Labor

The Yanomamö like all people exhibit strong individual differences in the amount of labor they perform that are independent of sex. Factors such as age, number of dependents, and marital status should logically help us to understand much of the variation. For example, one would expect that a married couple with numerous dependent children to labor more than newly weds with no dependents. Such a prediction is based on a number of assumptions such as each family is wholly responsible for supplying their economic needs and economic resources are freely available. While this latter assumption is correct for the Yanomamö, the former is suspect, as I will later explain. In this section I will examine the degree to which age determines individual labor time allocation.

Child Labor Trends

On the basis of our own experiences we expect that the amount of work one does will increase with age and that it eventually begins to diminish when one retires or becomes physically incapacitated. We also tend to believe that childhood should be a carefree time with little in the way of work responsibilities - a time for play, exploration, and learning. An examination of Yanomamö time allocation data will allow us to evaluate all of these ideas, and since the Yanomamö are relatively typical representatives of the tribal world we can get a sense of whether our Western experience is in any way typical over the history of humankind.

Figure 4 shows the amount of time children from ages 5 through 18 allocate to labor time activities. As can be easily seen, labor time does increase with age of child. The rate of increase in uneven only because small sample sizes in some of the age groups. Over the bar chart I have superimposed adult male and female labor time. You will note in this graph that adult females work significantly more (421 min/day) than males (372 min/day). These figures are different from the ones given in the 10 village since they derive from the four villages I studied. I used this restricted data set because it is the only one broken down by age. As the figure indicates, boys and girls begin to achieve adult labor time levels by the time they become teenagers.

The data presented seems to indicate that childhood is brief and children are quickly recruited into the family work force. To some extent these figures are an artifact of the method I used to collect time allocation data. If I could not observe someone when I was sampling behavior I had to rely on reports of what they were doing. For example, someone would tell me that all the members of a particular family were in the forest gathering wild palm nuts or weeding the garden. When I was able to accompany families on their economic activities I found that children did work but not as hard or as constantly as adults: they worked about 40% to 80% as much as adults when, for example, a garden was being weeded. Nevertheless, the data tell us something important about work and family life that provides a strong contrast to what occurs in the urban West. Yanomamö children work along side their parents and are important to the household economy. The family unit does not chronically separate in the morning as children and adults go their different ways to school and work only to rejoin in the evening. School for a Yanomamö child is in the context of the family economy where they learn how to hunt, gather, garden, fish, and all the other activities that are necessary for them to become competent adults.

Adult Labor Trends

If we extend the analysis of time allocated to work across the entire life span we expect to see labor time increase to a point then decrease. This pattern, an inverted U-shaped curve, is evident for both men and women in Figure 5. However, the shapes of the curves are quite different. Male allocation of labor time begins at a lower level but increases rapidly until it peaks around age 35 and then rapidly decreases thereafter. Females, in contrast, begin at initially higher levels, ascend more slowly to a peak at age 50 and then decrease their efforts much more slowly. The last point is rather interesting since the curve shows that women at ages 30 years and 60 years work about the same amount of time. The factors that account for these patterns are quite complex. Women engaged in active child care (i.e., nursing) work less than women who are not (Hames, 1992). This fact probably accounts for female labor time peaking after menopause. Male labor time decreases quite rapidly after age 35, but I am not sure why the rate of decrease is so much greater than for women. There are two interrelated possibilities. Since males have higher rates of mortality than women at all ages they may also have higher rates of decrepitude than women. That is, their ability to do physically demanding labor may decrease more rapidly. Related to this trend is that men increase labor time activities to relatively sedentary tasks such as manufacturing and the gardening tasks of weeding and harvesting with age while significantly decreasing their efforts in hunting and clearing new gardens.