sign in sign up
<<
Home , Behavioral ecology, Foraging

UC Davis UC Davis Previously Published Works

Title Social and the behavioral of intragroup transfers

Permalink https://escholarship.org/uc/item/5565r5zn

Journal Evolutionary , 5(2)

ISSN 1060-1538

Author Winterhalder, B

Publication Date 1996

DOI 10.1002/(SICI)1520-6505(1996)5:2<46::AID-EVAN4>3.0.CO;2-U

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California 46 Evolutionary Anthropology

Social Foraging and the of Intragrou p Resource Transfers

BRUCE WINTERHALDER

Two stalk, isolate, and kill a red colobus monkey. An attendant question: is sharing as practiced primatologist notes that parts of the prey are relinquished selectively to onlooking among modem foragers an early or scroungers (Fig. 1). A human forager returns to camp mid-afternoonwith a freshly late development in our prehistory? killed, medium-sized ungulate. Later in the day, an ethnographer observes that The models of behavioral ecology shared portions of the animal have found their way into the cooking pots of most or (boldface indicates a term defined in all of those in the small band. Examining a prehistoric scatter of food residues, an the Glossary) described here empha- ethnoarcheologist wonders when early hominids began to scrounge or share food, size the importance of intragroup re- and with what consequences for our . All of these settings represent one source transfer among social problem: the analysis of intragroup resource transfers among social foragers. New foragers, but give them different and studies in the behavioral ecology of transfers show them to be more commonplace more diverse explanations than those in nature, more complicatedand variable, and more subject to comparativeanalysis just cited. Among the key ideas are tol- than has been appreciated. erated theft, marginal evaluation, scrounging, trade, by-product coop- eration, reciprocity and Food sharing has been a routine ob- holder of a fresh kill could not prevent risk-minimization, and showing off. I servation in hunter-gatherer studies. or control to benefit. This is nascent will describe each of these ideas, then Scrounging is a more recent, but not sharing at best, incidental to the rare return to a discussion of resource uncommon observation for some pri- instance of . With chimps and transfers among hominids, nonhu- mates. At one time, the explanations modern foragers anchoring the end- man , and extant foragers. for these practices seemed clear. points of a presumed continuum, dis- The premises and methodology of Hunter-gatherers, living in small, sta- putes about the evolution of hominid human evolutionary or behavioral ble groups, enact their social alle- often have turned on one ecology should be familiar to most giances and secure a more regular diet through the institutionalized sharing of food. Obvious group benefits are served. For chimpanzees, prey trans- fer has been seen as something the

Bruce Winterhalder studies models as a means of understanding the origins of behavioral diversity among foragers and peasant food producers. He has undertaken ethnographic field work with hunter-gatherers of the boreal forest in Canada and with the Quechua, agriculturists in a tropical high-mountain region of Peru. He has published on diet selection, risk avoidance and sharing, predator-prey , and plant domestication and exchange. With Eric Smith, he is co-editor of the volume Evolutionary Ecology and Human Behavior. He is Professor of Anthropology and Chairman, Curriculum in Ecology, University of North Carolina, Chapel Hill. [email protected]

Keywords: Hunter-gatherers, primitive Figure 1, Alpha male shares a colobus monkey carcass with members of the hunting economics, risk, exchange, sharing, scrounging party. Photo courtesy of Craig Stanford, ARTICLES Evolutionary Anthropology 47

GLOSSARY Behavioral ecology - the field that applies makes the effort to locate and procure. theory to the study of behavioral in a socio-ecological Realized value - the total marginal value of a resource to setting. A subfield of evolutionary ecology. the individual procuring it, corrected for losses or gains arising By-product cooperation - individual A acts to its own fit- through tolerated theft, reciprocity, or trade. ness benefit. Individual B enhances the of that action to Reciprocal cooperation - individual A acts to the benefit A, and to itself, by cooperating in the performance of the task. of B at some cost to A, expecting that B will reciprocate in the Synonyms, by-product , pseudo-reciprocity. future. The individual benefit is greater than the cost, providing Detection opportunity - the chance of securing access to a a mutual benefit for the two-way interaction. Preferred to syno- resource either by producing or scrounging it. nym, reciprocal . Encounter-contingentforaging - optimal diet selection Risk - the likelihood of a resource shortfall and, conse- model that assumes a concurrent search for randomly arrayed quently, a negative effect on fitness. resource items. The decision to pursue an encountered item is Scrounger one who lives by appropriating from contingent on its profitability relative to the efficiency of forag- - ing for items of higher rank. Also called the diet selection or diet producers, who pay the costs of locating and procuring them. choice model; preferred to synonym, diet breadth model. Scroungers’ share - the portion of a resource patch or Evolutionarily stable (ESS) - the tactic or mix of packet remaining when scroungers show up and begin to con- tactics that will evolve to an equilibrium that cannot be changed sume it along with the producer. or “invaded by any other options within the set. ESS analysis, Scrounging -appropriating part of a resource that another based on , is used to predict the evolution of pheno- individual has located. Both opportunists and scroungers en- types with frequency-dependent fitness. gage in scrounging. Synonym, tolerated theft. Evolutionary ecology -the study of adaptation and biologi- Sharing - 1) the mutual, uncoerced, two-way transfer of cal design in an ecological context, using the theory and concepts portions of a like resource or service, sometimes with a delay, as of neo-Darwinism. in reciprocity-based food sharing; 2) a one-way transfer among Exchange - the two-way flow of goods and services (e.g.. kin without equivalent return in kind, as when a mother shares trade and some forms of reciprocity-based sharing. food with a child. Marginal value - the value, in fitness, utility, or other cur- Trade -a voluntary, two-way transfer of unlike resources or rency, of the nth unit of a resource to an individual already hold- services, with or without delay. ing n-1 units. See total value. Tolerated theft - a resource transfer occurring when the Models of circumstance - portrayals of an expected func- holder of a packet relinquishes portions of low marginal value tional relationship between environmental factors and a set of that are being contested by an individual to whom those portions phenotypic or behavioral options. havea high marginalvalue. Synonym, scrounging. Because theft Modelsof mechanism- constructs that examine if and how connotes property rights, scrounging in most is a better neo-Darwinian evolutionary processes might generate hypothe- cases label for this phenomenon. sized phenotypes. Total value - the sum of the marginal values of x units of a Opportunity costs - the cost of a behavior, as measured by the benefits associated with alternative activities that it pre- resource. cludes. Transfer - a neutral word for the movement among indi- Opportunist -one who acts flexibly either as a producer or viduals of good and services, not presupposing a particular a scrounger, depending on which opportunity presents. mechanism such as scrounging, sharing, or trade. Producer priority - the proportion of a resource packet Unilinear evolutionism - the approach taken by social an- consumed or otherwise held aside by the producer, which thus thropology and archeology in the late nineteenth and early twen- is not subject to the depredations of scroungers. See scroungers’ tieth centuries, which portrayed societies as developing through share. a single sequence of stages graded by increasing complexity, Producer - one who lives entirely !?om resources that she/he technological prowess, or energy capture.

readers of this jo~mal.l-~Earlier re- surfeit has little fitness incentive in gers); ii) the packet is larger than an views on food transfers among hu- risking defense of low-value portions. individual can consume immediately; mans have been written by Kaplan and The potential taker, less successful in and iii) consumption of the packet is Smith,’ and Hawkes.6 The pri- the food quest and perhaps facing characterized by diminishing mar- mate literature has been reviewed by hunger, has a large fitness incentive in ginal returns. It does not matter de Waal.9 pressing a claim on what it sees as whether the packet is from an animal high-value portions. One gains by or plant source. Assuming that com- MODELS OF CIRCUMSTANCE avoiding unnecessary strife, the other petitors meet as equals, transfers from by threatening it. Transfers generated the replete to the hungry will occur un- Tolerated Theft by this mechanism are known as tol- til an equilibrium is established at Individual-level selection should fa- erated theft.lOJ1Such transfers are which the last portion has an equal vor social foragers who contest the un- likely to occur if i) the food item is marginal value for the giver and equal possession of foodstuffs within discovered individually and sporadi- taker. a group. The holder of a temporary cally (that is, not in phase among fora- Tolerated theft requires only indi- 48 Evolutionary Anthropology ARTICLES

Box 1. Marginal Valuation and Evolutionary Currencies

When King Richard Ill declares “A ! a horse! my kingdom for a horse!” we can be reasonably sure he is engaging in marginal analysis. Mis- fortune on the battlefield can upend the relative value of an equine mount and continued sovereignty. In less than dra- matic circumstances, humans make 2 vp such comparisons all the time. Other -m species routinely behave as if they do > +e so as well. a The evolutionary ecology hypothe- Y. 0 ses that explore these are for- mulated around cost-benefit accounting. The balance of accounts allows us to trace the likely fitness consequences of behavioral alternatives. However, tally- ing up these costs and benefits entails VO refinements in the way they are concep- QO Qe QP Qs1 QS2 tualized and measured. First, in behavioral ecology, all costs and Packet Quantity benefits include opportunity costs and benefits.Any behavior should be assessed relative to the alternate flness-enhancing options available Often, this will not be the case. For example, a meal-sized to the organism. Direct measures of outcome for a particular portion of food makes a high contribution to well-being, but behavior are subordinate to potential trade offs among be continued consumption of like increments adds less and haviors. Thus, the fitness gains of foraging at a certain net less to well-being. Satiation sets an asymptote in the rela- yield must be assayed against the fitness gains associated tionship between quantity and fitness value. with other activities that might occupy the same time. The above figure shows the fitness gain associated with Second, the value of an activity changes as it is performed food packets of increasing size. G (for giver) has a packet or as a function of the time since it was last performed. of quantity (Qp) and (Vp).It is relinquished in small portions Because of this, the decision to switch behaviors is a mar- to a scrounger (T, for faker),who begins the transaction with ginal decision. As hunger is sated, continued foraging re- no resources (Qo,Vo).Under conditions of tolerated theft, the turns less and less value for effort. Correspondingly, while transfers will equilibrate at Q,. With their marginal values attention is on the food quest, the demands of other fitness- equal, the incentive for T to acquire additional portions is enhancing activities grow more urgent. This linkage be- precisely balanced by the incentive for G to resist ceding tween opportunity costs and marginal analysis requires only any more. Note that the packet has been divided equally, but that we postulatescarcity of time, resources, or both; organ- that G has lost considerably less value (V,-V,) than T has isms for which fitness depends on at least two mutually ex- gained (Ve-Vo). The aggregate value for the two of them is clusive activities; and a nonlinear relationship between greater after the tolerated theft (2V,>Vp+V,,), although the ag- either the duration of an activity or the time since it was last gregate resource quantity has not changed. Variations on this performed and the contribution of that activity to fitness. graphical treatment allow for complications due to the holding Ideally, we would focus our models and observations on of resources other than food packets, differing numbers of tak- marginalanalysis of fitness in terms of opportunity costs and ers, and other benefits. In practice, we compromise. We often settle for Several points in this paper turn on the appraisal of mar- direct rather than opportunity cost measurements, measure ginal values. Note that the holder of avery large packet (as,) proxy currencies instead of fitness, and assume a linear might relinquish much of it (Qs,) with virtually no loss of relationshipbetween quantity and value. Each simplification value. Much is ceded; nothing of value is lost. Conversely, has advantages. Direct measurement allows us to isolate the small quantity propitiously gained by an individual with one behavior for study. The desirability of a greater as op- no resources may have extraordinarily high value. For ex- posed to a lesser foraging efficiency is a reasonable prem- ample, scrounging models use constant units (“items”) of ise, whatever the opportunity costs. Because fitness rarely food. However, with even a small degree of producer priority, is amenable to direct measurement, we observe instead the finder of a packet cedes much less value to scroungers more tractable proxy currencies such as net acquisition rate than would be indicated by the amount scroungers might of energy or number of items exchanged. Further, we as- consume. Similarly, a hungry scrounger might obtain value sume that fitness is a constant function of units of proxy. that is disproportionate to the quantity it is able to secure. ARTICLES Evolutionary Anthropology 49

vidual-level selection for its initiation. tionary costs and benefits in terms of tween resource selection and distribu- The conditions favoring it are less resource quantity. Consider two exam- tion.8,I4-’6They should change the way stringent than those that might pro- ples. First, much of a very large packet in which we apply and evaluate long- duce reciprocity-based sharing. Blur- might be subject to appropriation standing models of production, such as ton Jones makes the very plausible without any loss of value to its finder. for encounter-continent foraging.” suggestion that sharing has its origins Graphically, we would envision the Because behavioral ecology hypotheses in tolerated theft, which became suffi- packet holder’s value as lying along the often turn on small differences in a bal- ciently routine in small, relatively sta- horizontal asymptote both before and ance of costs and benefits, marginal ac- ble groups that it could prime the after redistribution (Box 1). Despite counting may be critical in a wide evolution of reciprocity by meeting having a reduced portion, the holder variety of circumstances. these additional conditi~ns.’~-l~Most has suffered no loss in the currency of important are regular role reversals Producers, Scroungers, and among a small set of givers and takers, Opportunists stable group membership, the pros- pect of an indefinite future of trans- Social foragers that harvest one or fers, and a clear advantage to the Producers expend more resources in packets have the interaction. energy to locate option of obtaining food by producing it or by scrounging from the produc- resource packets; tion of others. The rewards of active Marginal Analysis scroungers avoid search foraging are diminished by scroung- Tolerated theft assumes that a re- ing; the advantages of scrounging are source packet has decreasing mar- costs by appropriating heightened by the chance of obtaining ginal value to its holder as it increases portions of a producer’s a meal with minimal effort. In a popu- in size. Formal analysis of this premise lation of producers, the pure with graphical models14 not only packet; and scrounger has a significant advan- shows how resource transfer distrib- opportunists produce or tage. Easy taking diminishes the at- utes costs and benefits among the par- tractiveness and the fitness ticipants, but also how such transfer scrounge as the advantages of procuring one‘s own can be affected by group size and eco- occasion presents. food. Because the costs and benefits of logical conditions, and can interact these tactics are frequency-depend- with models of production (resource ent- as long as there are a few of selection). them, scroungers do well relative to In tolerated theft, the benefits to a evolutionary significance. Second, producers, but do poorly if they be- taker are greater than the costs to a central to resource choice models is a come too common, so that more and giver (see Box 1).This follows from the ranking of resources by their net en- more of them compete for a smaller marginal nature of the transaction. ergy return per unit of time spent in harvest-the goal of analysis is to de- Distribution does not change the pursuit and handling. Energy value scribe the evolutionarily stable amount of the resource, but does en- typically is determined by the edible strategy or combination of strategies. hance its total value to the partici- mass captured. But if that mass is sub- Among the several recent models of pants. This is the advantage required this situation,18J9 I will summarize ject to tolerated theft, the realized for the evolution of reciprocal coop one developed by Vickery et al.*O There value derived by the forager may be eration. It underwrites the shift from are three tactics: Producers expend less than would be inferred from the tolerated theft to sharing. energy to locate resource packets; size of the resource. The loss to toler- This enhancement of a packet’s scroungers avoid search costs by ap- value by its transfer grows with the ated theft will lower the ranking of the propriating portions of a producer’s number of recipients. With increasing resource, perhaps sufficiently that it packet; and opportunists produce or “n,”the low-value portions ceded by a drops from the optimal diet. The half scrounge as the occasion presents. By giver concentrate on the highest incre- of a hare remaining after tolerated definition, producers and scroungers ments of marginal value that can be theft may not be worth the effort of adopt pure tactics. The equilibrium realized by takers. If the situation is capturing the whole animal. By con- mix of these behavioral options is de- one of reciprocity, with a receiver in- trast, if a forager expects a realized termined by group size, producer pri- debted by the value derived from its value that is enhanced through reci- ority, and the opportunist’s detection portion, then a self-interested giver procity, he or she may well find it op- opportunities. will distribute portions as widely as timal to harvest a resource that Producer priority measures the possible, and to the neediest among otherwise would be ignored. The indi- degree to which the finder of a packet the supplicants. In sharing, a motive vidual use value of a wildebeest may is able to monopolize its consump- of maximum self-interest may well not justify its pursuit unless there are tion, thus lowering the scroungers’ give the appearance of selfless gener- delayed returns to be realized through share. Producer priority will be high osity. exchange. if a packet can be quickly consumed Marginal valuation also highlights Models of food transfers among so- or if the scroungers are slow to detect the pitfalls of accounting for evolu- cial foragers reveal the linkages be- and reach it. Detection opportunity 50 Evolutionary Anthropology ARTICLES

Box 2. The Producer-Opportunist-ScroungerGame 1.o compromises that individual must make in order both to produce and to scrounge. Detection opportunity can range 1.9 0.5 from 0 to 2. At a value of 0, the opportunist, in trying to detect both types of feeding situations, succeeds at locating nei- ther. At a value of 2, the opportunist detects all of the oppor- 0.0 tunities garnered by either producers or scroungers, thus 1.o doubling its chances of feeding relative to either of the pure tactics. Finally, if a producer consumes a large portion of a 1.6 0.5 patch before being joined by scroungers (high producer pri- ority), the scroungers’ share is diminished. Figure 2 depicts an outcome of this game. There are 10 0.0 foragers and 20 food items per patch. A/F, the scroungers’ 1.0 share, vanes from 0 ( the producer gets all 20 items; com- E .-0 plete producer priority) to 1 (scroungers arrive before the .c r + 1.3 0.5 0 producer can consume even 1 item; no producer priority). Q 0 2 The variable cis the relative ability of an opportunist to de- n tect feeding situations that would be found by a specialist 0.0 producer. The variable h is the same for the opportunist relative to a scrounger. For simplicity, we assume that op- portunists’ abilities are symetrical (c=h). Then Oe=c+k=2 1.c measures the relative frequency with which the opportunist detects feeding opportunities.The five graphs show a range of values for c+hcl. Several things can be observed here. For c+h there are 1.o no opportunists. If the scroungers’ share (NF) is less than 1/N (0.1 in this example), the population is composed en- tirely of producers. Put differently, producer priority must be 0 to <1 0.5 greater than 0.9 to exclude scroungers from the equilibrium. Otherwise, there is a two-tactic mix of producers and 0.0 scroungers, with the former decreasing in frequency as NF 0.0 0.2 0.4 0.6 0.e 1.0 increases. AIF Above c+h =1, opportunists prevail, either alone (to increas- Across a wide variety of species, social foragers are af- ing degree as c+h approaches 2), in combination with produc- flicted by scrounging (toleratedtheft). Individuals having the ers (low-to-moderate A/F), or in combination with scroungers fortune (and paying the costs) to locate a food patch or (high No.Apopulation composedonly of producerscontinues packet may inadvertently signal the discovery to group to be restricted to A/?=el/N, but it is progressively squeezed out members who hasten to appropriate part of it. This tolerated as detection opportunities approach 2. theft has consequences for social tactics, foraging strate- In virtually all circumstances, the equilibrium is composed gies, and subsistence viability. of only one or two of the options. Scrounging, whether by Considerthree tactics. Producers actively seekto procure scroungers or opportunists, is present over most combina- food. Scroungers avoid search and pursuit costs by feeding tions of scroungers’ share and detection opportunity. It is only from the finds of producers. Opportunists do both, pro- likely to be absent only from small groups (low N) in which ducing and scrounging as the chance arises. Vickery et al. c+kl and producer priority is very high. In general, increasing (Table 1, p. 854)20have derived the evolutionarily stable the group size (N) shifts the boundaries between tactics to the strategy (ESS) proportions of these tactics as a function of left. Thus, larger groups allow increased likelihood and fre- three variables: the relative frequency of the feeding oppor- quency of scroungers. tunities available to the opportunists, the percentage of a Three cautions are germane. First, this is only one set of packet or patch that scroungers seize, and group size. ESS transition thresholds and frequencies. Others are possible analysis of this situation is necessary because the payoffs with changes in Nand the other fixed parameters (5c, and to scrounging and alternative tactics are frequency-depend- h). Second, because opportunists both produce and ent. Relative to producers, scroungers do very well when scrounge, this chart gives only a rough sense of the total they are few and opportunities to parasitize producers are amount of production and scrounging in the economy. Fi- abundant. Scroungers do poorly when they become com- nally, it cannot be inferred that all of these combinations mon. Not only are there fewer chances to cadge, but more represent viable populations. At high scroungers’ share scroungers compete for each item. (A//=),the ESS approach can produce an equilibrium that The chance that an opportunist will feed depends on the leaves everyone with too little food to survive. ARTICLES Evolutionary Anthropology 51 will be low if the opportunist, by divid- Trade more likely if a in estrus is pre- ing his or her attention, fails to find as Analyses of trade appear infre- sent. The hunting group grows larger and is more likely to be successful as many feeding opportunities as does quently in the literature on foraging the number of estrous in- either a producer or scrounger. Detec- perhaps because it is rare in extant tion opportunities will be high if op- hunter-gatherer societies. Exceptions creases. Further, males favor their al- lies and exclude their political rivals portunists freely can avail themselves are those cases in which foragers trade of the feeding situations available to regularly with horticultural neigh- from such distributions, suggesting both producers and scroungers. This bors.22223 Relevant models are avail- trade for political support. effectively doubles their encounters able from micro-economics,24 but with feeding opportunities relative to they rarely have been applied formally. By-product Cooperation the opportunities provided by either Transfers may also result from evo- of the pure tactics. lutionary mutualism. In by-product Box 2 shows a typical outcome of cooperation. an organism undertakes this kind of analysis. Opportunists are an action with a personal benefit that excluded from the evolutionary equi- will be augmented if it is performed librium if they detect fewer feeding The of cooperatively, whether actively or in- events than are available to either pro- hominid scavenging Conversely, noncoop- ducers or scroungers (detection op- may have depended on erators pay a cost if they opt to avoid portunities ~1).The equilibrium most joint action. By-product cooperation likely is a mix of producers and quickly locating and may be common in nature.29Like tol- scroungers. It will be limited to pro- then cooperatively erated theft, it generates conditions ducers if the scroungers’share falls be- that could prime the evolution of reci- low the reciprocal of group size. By defending a feeding procity cooperation. contrast, if opportunists have a detec- site. The hominid Social foraging provides an exam- tion advantage (detection opportunities discovering a freshly ple. Assume the optimal diet of a fora- >1), then the most likely outcome is op- ger includes a medium-sized packet, portunists either alone or in combina- killed carcass would highly ranked by its realized value. In tion with producers and scroungers. enhance the value he seeking the advantages of high forag- The producer-scrounger model or she was able to ing efficiency, group members inad- hints at the complex behavioral dy- vertently cooperate in providing one namics affecting social foragers derive from it by sharing another scrounging opportunities. whose diet includes resources that are the opportunity - and Foragers can fail to cooperate, passing subject to tolerated theft. These dy- by the resource in question, only at namics might appear to require tacti- particularly its their own expense. Hominid scaveng- cal skills that are too complex for appropriation and ing presents a second case. Eth- nonhuman species. However, work defense - with others. noarcheological work suggests that with spice finches has shown that the productivity of hominid scaveng- these alter their foraging behav- ing may have depended on quickly lo- iors in the ways predicted.21 cating and then cooperatively Group living usually is analyzed for defending a feeding site.30The hominid its advantages in, for example, locat- Because trade implies incommensu- discovering a freshly killed carcass ing food or detecting and deterring rate goods or services, it is difficult to would enhance the value he or she was predators, but scrounging draws at- measure the relative costs and benefits able to derive from it by sharing the op- tention to its costs. Under a wide range of such exchange. portunity - and particularly its appro- of conditions we should expect groups Two empirical examples are notable priation and defense - with others. of social foragers to experience toler- for their similar proposals, although ated theft, whether practiced by the data in both cases are quite lim- Risk and Reciprocity scroungers or opportunists (but not ited. Kaplan and Hillz5suggested that Cooperation both) amongst producers, scroungers more productive AchC hunters trade Tolerated theft occurs among fora- amongst opportunists, or opportun- foodstuffs for reproductive advan- gers whose diet depends significantly ists amongst themselves (Box 2). Self- tages such as opportunities on relatively large resource packets or interest pushes the mix of tactics to an and better treatment of their off- on patches that are encountered spo- equilibrium at which no tactic garners spring. Stanford26 believes that chim- radically and asynchronously by indi- more than another. Among other con- panzees capture game - primarily vidual foragers. Because these sequences, this can strongly depress colobus monkeys -both for its nutri- encounters are also likely to be unpre- the average rate of food harvest in tional significance and because it can dictable, subsistence risk is likely to such groups. In simulation models, it be exchanged for sexual access and affect the evolution of resource trans- can jeopardize survival.1gFaced with political support. He observes that fers. scrounging, self-interested foragers most hunting occurs during the food- Ethnographers routinely report come to underproduce. scarce dry season. Hunts also are that food sharing ensures band mem- 52 Evolutionary Anthropology AR JICLES

synchrony in their short-term success Box 3. The Ambivalence of Sharing rates. High day-to-day variation of success in hunting and gathering of How are we to understandgenerosity that must be evoked by a firmly tendered course elevates the indwidual’s risk of request? This question is at the heart of an essay by Nicolas Peterson, titled periodic shortfalls. But with even “Demand Sharing: Reciprocity and the pressure for generosity among fora- modest degrees of asynchrony in the gers.”= yield among foragers, pooling is quite Using Australian and other ethnographic materials, Peterson corrects a wide- effective in mitigating the likelihood spread mis-impression: although ethnographers and secondaty sources often of harmful shortfalls. Further, most of gloss hunter-gatherersharing in an idiom of unsolicited and beneficent giving, the benefit of sharing can be achieved the record more often speaks of demands and claims, and occasional resistance in small groups of six to eight fora- to these claims: ger~.~~ “Public pressure on individual Anbara to share was virtually irresistible, so We now have good quantitative evi- various counterstrategieswere adopted by the diligent to prevent exploitation by dence of this phenomenon. For in- the lazy or manipulative. The most effective of these, in Hiatt’s view, was eating stance, based on their field work with during food collection, so that the greater part of a person’s produce was in an AchC forager-horticulturist s, Kaplan advanced state of digestion by the time he or she returned to camp ...” (p. 860). and Hill6have calculated the improve- Peterson explains the discrepancy between anthropological representation ments in nutritional status of indi- and native practice by the positive value that Western culture and Western viduals achieved through food anthropologists place on altruistic giving. A secondary explanation is the ten- sharing: 20% with the sharing of dency in ethnographic interpretation to rely on idealized, normative statements honey alone, 40% with the sharing of rather than on behavioral practices in specific situations. meat alone, and 80% with the sharing Peterson also considers the advantages of what he calls “inertial sharing,” in of all foods. which the burden of reciprocity is put onto the taker rather than the giver. An Food sharing among hunter-gather- uninvited offer can appear rude if refusalto accept is ruder yet, and if the potential ers typically has been explained by its recipient does not wish the obligation to reciprocate. Further, faced with highly seemingly obvious group benefits. Be- complicated socio-economic bookkeeping of past transactions among kin and havioral ecologists prefer an explana- camp mates, the holder of a resource might be tempted to simplify and just tion in terms of individual self-interest respond to demands as they are made: ”Difficult decisions are avoided; the onus and mechanisms that are consistent is placed on others: discrepancies in the evaluation of relationships are not laid with neo-Darwinism. In by-product bare; and an excellentexcuse is provided for not meeting some obligationswithin cooperation, an individual advances the context of behaving generously” (p. 864). Faced with awkward calculations its own interest while incidentally pro- of giving, the victim of tolerated theft may have reason to appreciate being taken. viding a common good. In reciprocity It is tempting to speculate that the overly romanticized interpretationsof shar- cooperation, an individual provides a ing in contemporary foraging societies have influenced the recent rejection of common good at some cost, expecting exchange as an element in accounts of hominid origins. Such interpretations that in the future similar behavior by invite a reaction in which the paleoanthropologistplays Hobbes to the ethnog- others will more than compensate for raphers’ Rousseau. By contrast, Peterson’s review couches the ethnographic that cost. Because the delay carries a record in terms that are consistent with predictions of the accompanying text. chance of , analysis of reci- The multiple origins and motives for food transfers conflate self-interest and procity cooperation requires tit-for-tat benevolence, giving the practice its ambivalent quality. They also help to explain and related game theory models.33 how foragers committedto sharing might nonetheless go out of their way to avoid If foraging groups, whether of early a situation in which it would be expected. hominids, nonhuman primates, or The Yolngu (of Arnhem Land) chastise supplicants who overhear talk of food modern hunter-gatherers, are com- with the line, “Your ears have fingers” (p. 862). A similar rebuke must have posed of small bands with relatively occurred many times among the early hominids. stable membership, they provide the context needed to make reciprocity- based food sharing an evolutionarily stable tactic.12J3Food sharing may be bers a more regular day-to-day diet. pants more evenly than they initially routine, providing an indefinite pros- Behavioral ecology models have re- were acquired. Most models assume pect for future transactions. Because that each participant gets an equal fined this empirical generality by iden- such sharing is a highly visible activity tifying more precisely the ecological share. This is formally analogous to of most adults, cheating can be moni- contexts in which sharing is effective the risk reduction achieved by the tored and sanctioned. Resource cap- in reducing risk and by drawing our farmer who plants multiple, dispersed ture is fortuitous, ensuring regular attention to the varying forces that agricultural plots.)’ Three variables role reversals. Marginal valuation may motivate it. determine its effectiveness: the num- shows the fitness gains of regular shar- Food sharing pools the yield from ber of foragers who participate, the ing; risk adds to those benefits by en- spatially separated foraging expedi- day-to-day variation in their individ- hancing the marginal value of food tions, allocating shares to the partici- ual foraging success, and the degree of that unpredictably is in short supply. ARTICLES Evolutionary Anthropology 53

Although tolerated theft may have a family or the high-variance, poten- DISCUSSION preceded and catalyzed sharing, it ap- tially high-reward option of showing parently has remained a factor in off before the whole . By hunter-gatherer transactions. Al- this hypothesis, women are advised to Mechanisms and Lacunae though generalized discussions of marry the town grocer but flirt with a At their best, studies of behavioral sharing stress social obligation and generous buccaneer, and men to con- ecology combine models of circum- personal generosity, close scrutiny of sider carefully which role represents stance and models of me~hanisrn.~~ first-hand ethnographic accounts re- their best opportunity. The former models show how specific veals transactions fraught with am- Foragers often pursue large game; environmental conditions constitute bivalence. Peterson34 characterizes when they do so, it is the men who do adaptive costs and benefits; the latter the Australian materials with terms the hunting; and when these men are identify which of these costs and bene- like “demand sharing” and “mutual successful, their harvest is distributed fits are potentially consequential taking,” descriptions that ring true for widely, benefiting group members be- within neo-Darwinian and related nondustralian cases as well (Box 3). yond themselves and their families. mechanisms of evolutionary change. As recorded in ethnographies, sharing The show-off proposal is attractive be- Tolerated theft and scrounging, mar- encompasses diverse practices mixing cause it neatly explains these three ginal valuation, trade, and by-product generosity, tit-for-tat reciprocity, and correlated observations with one idea. cooperation can be matched to sim- barely disguised social extortion. Be- ple, individual-level selection. Risk havioral ecology models couched in minimization and reciprocity-based terms of tolerated theft, marginal valu- sharing require selection for recipro- ation, scrounging, and reciprocity lead cal co~peration.~~,~~The show-off hy- us to expect such variation in the form In the show-off pothesis invokes the mechanism of and motives of this behavior. .42 hypothesis, men seek Kin selecti0n~3.~has not been large game not to eat linked with particular models of cir- Showing Off cumstance, perhaps because it has Ethnographies record that men or to feed their families, seemed self-evident that food would typically are the hunters, and that but in order to cultivate be shared to advantage among related hunting often is a risky, high-variance individuals. Here the most interesting activity, especially if it is for large support from a coterie evidence is negative. Although good game. Hawkes’6 believes that these of allies and paramours. data are rare, there does not appear to men are providing a “public good,” a be a strong kin bias in band- or com- service costly to them and valuable to munity-level ~haring.6.~~~~~In fact, the others. Transfers of large game are un- marginal advantages of widespread compensated by reciprocity or trade, Nonetheless, this hypothesis has had a distribution within a group that in- cludes individuals who are not kin leaving a lower realized benefit to such mostly critical reception and has gen- hunters than they might gain from al- may swamp the direct fitness benefits erated lengthy debate.35-37It has been ternative subsistence pursuits. Be- to be achieved by exchange that is lim- difficult to show that large game are cause this appears to defy the ited to relatives. “Reciprocalaid giving assumption that indviduals act from public goods. For instance, if the shar- ... can generate a substantial selective self-interest, Hawkes seeks to explain ing of large game is a matter of by- force independent of hunting in terms of the less tangible product cooperation, then there is no even when performed among related benefits it might garner. Primarily, need to invoke intangible benefits to animals.”47 these are the attention, potential social explain the pursuit of such game. Unfortunately, we lack models of influence, and, perhaps, the mating Hawkes argues that the alternative circumstance that are specifically opportunities these men might receive risk-minimization (or sharing) hy- linked to two additional kinds of selec- from their conspicuous displays of pothesis is weakened by the difficulty tion, interdemic (or group) selection48 food sharing largess. In the show-off of enforcing reciprocity. The show-off and dual inheritance appro ache^.^^ hypothesis, men seek large game not proposal not only entails the same dif- Because interdemic selection and to eat or to feed their families, but in ficulty, but also recreates the public dual inheritance appear more congen- to the evolution of group-level, co- order to cultivate support from a cote- goods problem by supposing that de- ial operative behaviors, they may be rie of allies and paramours. layed and somewhat diffuse social at- The evolutionary mechanism here especially promising venues for inves- tention is the hunter’s reward.38 If is sexual selection. Females will prefer tigating exchange. everyone collects, why should anyone a mate who is intent on regularly pro- Despite these lacunae, half a dozen visioning her children, but should be bother to repay? Finally, in practice it models of circumstance already have willing to encourage any males who may be difficult to separate this hy- been more or less matched to an equal sporadcally provide a windfall of re- pothesis from that of trade, in which number of evolutionary mechanisms. sources. Males have two options: the sexual selection alters the value males Each model is peculiar to different en- low-variance but relatively con- and females assign to the elements in vironmental or evolutionary condi- strained fitness option of provisioning the transaction. tions. Taken together in combinations 54 Evolutionary Anthropology ARTICLES set by the environment and history of the degree to which it depresses aver- data as better supporting an interpre- a population, they predict diversity age food harvests. tation of reciprocity than one of toler- and considerable flexibility in the ex- Primates ated theft or social enhancement. By pression of intragroup exchange be- strict definition, this is reciprocity in The chimpanzee food transfers de- haviors. Whereas ethnographic trade. The counter-flows entail unlike scribed by StanfordZ6might be ex- summaries have tended to stress commodities and services. plained by some combination of broad generalizations (e.g., band-level In other cases, the practice of pri- scrounging, trade, and showing-off. sharing50), behavioral ecology theory mate reciprocity has been securely es- Stanford’s observations are of a popu- would have us attend to the variants. tablished in more naturalistic settings. lation infrequently and incompletely Some species- rhesus macaques liv- Hunter-Gatherers, Primates, and dependent on medium-sized packets. ing at Cay0 Santiago are an exam- the Hominids Because meat supplements a diet oth- ple55.56- separate to forage, with erwise low in risk, these chimpanzees individuals giving food calls that sig- Hunter-Gatherers may be an instance of social foragers nal to their fellows the discovery of a A common generalization from whose transfers lack the ecological rich patch. Hauser and MarlerS6de- hunter-gatherer ethnographies is that conditions necessary to establish the scribe these calls as “honest” signals of band-level organization involving regular, reciprocity-based exchange of reflecting food possession, noting that radial foraging from a base camp by “Those who fail to signal and are individuals or small groups. Food caught with food are apparently pun- often is consumed as it is procured, ished.”Because of this aggressivepun- with the surplus, and especially the ishment, non calling females that were bulk of medium-to large packets, car- ... studies show detected by conspecifics ate less food ried back to camp to be shared at the reciprocity-based from their discoveries than did those end of the day51 Recast in the vari- sharing to be a that gave calls (the sample of males ables of the producer-scrounger was not large enough to make this model, this scenario reveals links to widespread capacity in comparison). The temptation to re- other features of the foraging mode of the order, when evoked main quiet upon discovery of a patch production. Partial consumption of and the attempt by conspecifics to en- foodstuffs as they are gathered en- by appropriate force signaling suggests that this is sures a high degree of producer prior- evolutionary reciprocity rather than by-product co- ity. Concurrently, end-of-day sharing operation. Females call more than sharply reduces any conflict between circumstances. males, and their call rate is positively the detection of producing and correlated with the number of female scrounging opportunities. By ensur- kin in the vi~inity,5~suggesting an ele- ing that the individuals benefit first ment of kin selection. Although not ex- from their own production and that ethnographic hunter-gatherers. The plicitly analyzed in terms of the producers will miss few opportunities “fission-fusion”26reworking of group transfer models presented here, the to scrounge, this arrangement elimi- membership also may impede the de- general form, and especially the care- nates pure scroungers from the evolu- velopment of reciprocal cooperation. fully documented nuances of rhesus tionarily stable mix of tactics. By contrast, experimental studies macaque food calls, might well reward Producers and opportunists will based on the provisioning of nine such an effort. dominate, an observation that is con- adult, corral-living chimpanzees give Hominids sistent with the ethnographic record. clearer evidence of reciprocity in food Central place sharing may be a group’s sharing.9 Observation of 4,653 food in- In the 1970s, Plio-Pleistocene homi- best defense against by teractions showed that exchanges are nids were described as hunter-scaven- pure scroungers. balanced among dyads, although not gers who transported portions of medium-to-large to central Routine scrounging by opportun- in a tit-for-tat manner. Instead, prior camps to be shared among the mem- ists in a mix of producers and oppor- grooming and aggression are linked to tunists also provides a sufficient bers of a small band. This home base food transfers in a turn-taking pattern explanation for the “underproduc- provided selection pressures for lan- that helps maintain equivalence in the tion” observed in foraging societies.52- guage and complex social skills; homi- general exchange of social benefits. An 54 The self-interest of each forager nids were considered to be unique generates an equilibrium mix of tac- indwidual that had been groomed was primates because they shared food in tics in which the average production more likely, later, to transfer food to its this c~ntext.~~-~~By the mid-l980s, of all is constrained. Further, be- benefactor. Individuals that hoarded however, this “consensus scenario” cause scrounging increases in fre- food were more likely to receive an ag- had been rejected.60In the new view, quency in larger groups, we perhaps gressive response when they sought early hominid found only should include it among the list of food. Chimps rarely offered food, but morsels too small to share, gleaned factors influencing band size.51 they often relinquished it selectively to from carcasses abandoned by other Small groups diminish the fre- animals seeking a portion of a bundle. . Only ‘‘justprior to the ap- quency of scrounging, moderating On balance, de Waa19 described his pearance of fully modern man”61(p. ARTICLES Evolutionary Anthropology 55

321; italics in original) did Homo sapi- tive hunting of intermediate and large pacity in the order, when evoked by ens begin hunting from a camp, trans- game may have become profitable appropriate evolutionary circum- porting back to it large packets for only after the realized value of such stances. redistribution. Sharing was very late prey to the hunter could be enhanced and did not influence hominid diver- through reciprocity-based exchange. CONCLUSIONS gence from other primates. The difficulty is one of two linked pre- These debates about hominid ori- conditions. Intermediate or large Intragroup resource transfers are gins are debates about behavioral packets are a condition for the evolu- much more common than has been ecology. When and with what effects tion of exchange, but they are more thought and may have a variety of did protohumans take up scavenging likely to enter the diet when they have causes. They take diverse forms. Sta- or hunting, the pursuit of intermedi- exchange value. As noted, radial forag- ble roosting groups of related and un- ate and larger packets, and passive or ing with scavenging offers one solu- related vampire bats share food, the active transfers of food? tion to this evolutionary dilemma. We supplicant signaling need by groom- Studies of the Hazda30 showed that need only envision that, once secured, ing the distended abdomen of an indi- scavenging can intermittently but sig- such low-cost but larger packets were vidual that was successful during the nificantly contribute to the diet of sa- reallocated through tolerated theft previous night’s foraging. This behav- vanna woodland hunter-gatherers if and by-product cooperation, thereby ior, which forestalls starvation in the the foragers are able to locate fresh catalyzing the evolution of reciproc- otherwise high-risk food quest of kills quickly and cooperatively dis- ity-based sharing. Once reciprocity these animals, is explained by reci- procity-based c~operation!~,~~Com- place larger carnivores such as . was active, encounter-contingent mon ravens use aggregation at Scavenging opportunities are eagerly hunting of medium and large game roosting sites to share information sought by the Hazda, presumably be- became profitable because the forager about the location of inh-equent but cause of their high return rates. was more certain of recouping the rich scavenging opportunities, which Blumenschine et a1.62have identified costs of pursuit and handling. then are cooperatively By- a variety of contexts in which car- It commonly is posited that techno- product cooperation is a good candi- casses still bearing flesh would be logical developments were necessary date for the origin of this behavior. available to hominids without compe- precursors to large game hunting. It is Social carnivores hunt cooperatively tition from other carnivorous scaven- also possible that social developments and engage in varying forms of re- gers. In a prehistoric setting, radial in the form of stabilized reciprocity source transfers6’ foraging by early hominids in search and exchange relationships were pre- Each of the cases cited here - of a variety of resources would multi- conditions of this activity. band-organized foragers, hunting ply their opportunities for finding This scenario illustrates one possi- chimps, and hominid scavengers, as fresh kills. The base camp population ble application of behavioral ecology well as bats, ravens, finches, rhesus could be readily mobilized for their models to questions of hominid ori- macaques, and social carnivores- collective seizure, whether it involved gins. In practice, local production and draws on a somewhat different com- active displacement of other species resource transfers would reflect a bal- bination of circumstances and mecha- or stealthy, passive appropriation. ance of by-product and reciprocity co- nisms. However, all of these cases have Winter-scavenging ravens,63 as well as operation, scrounging, nepotism (kin in common the asynchronous harvest the Hazda, may be an apt model of this selection), and sex-linked exchange of medium-sized or large packets by process. tactics. Nonetheless, the scenario de- social foragers. Behavioral ecology Resource items gain a place in the scribed is consistent with the apparent not only tells us to expect diversity, it encounter-contingent diet if they have overlap in the feeding patterns of gives us a means of predicting the vari- a sufficiently high net value (the value hominids and Bone ants. This capacity and the empirical after pursuit and handling costs). A fragments from multiple hearths that work it has stimulated have three partially consumed carcass, even one can be refit and reliably inferred to be broad implications for hunter-gath- that has been reduced to morsels by a from a single animal provide us one erer studies. I will use paleoanthropol- prior , may still be valuable means of gaining insight into prehis- ogy as an example, for it is here that if its cost is slight. Small or easily cap- toric sharing.65 inferences must be developed most di- tured game may also be profitable for The sharing hypothesis of the early rectly from theory and comparative a hominid. But intermediate and large 1970s faded, in part, because it be- materials. Analyses by Winterhalde~~~ game are costly to capture and may came tainted by charges of romanti- O’C0nnell,6~and Blumenschine et a1.62 offer the hunter a realized value that ~ism.~~Early hominid foragers develop implications for ethnography, is much less than their bulk suggests. seemed too nice and too much like ethnoarcheology, and paleoanthropol- By these observations, we would pre- their modern counterparts, extant ogy, respectively. dict hominid scavenging prior to ac- hunter-gatherers. However, evolution- Behavioral ecology establishes a tive hunting, and active hunting of ary ecology offers a supportive ac- theoretical basis for wide ranging com- small game prior to that of intermedi- count of sharing that is grounded in parative studies of behavior. Models of ate-sized game. unsentimental self-interest. Further, hominid behavior have been derived Larger game present a particular primate studies show reciprocity- from phylogenetic analogies (chim- difficulty. If pursuits were costly, ac- based sharing to be a widespread ca- panzees, baboons) or ecological 56 Evolutionary Anthropology ARTICLES analogies (social carnivores). Ethno- taxonomic lines throughout hominid suggestions about the ecology and evolution of sharing, hoarding and scrounging. SOCSci graphic analogy once flourished, but history, depending on ecological cir- Inf 26:31-54. now has declined. Behavioral ecology cumstances. Fearful of projecting con- 12 Axelrod R, Hamilton WD (1981) The evo- both broadens the range of such com- temporary ethnographic observations lution of cooperation. Science 211: 1390- parative possibilities and gives us onto early hominids, we inadvertently 1396. 13 Axelrod R, Dion D (1988) The further evo- theoretical rationales for selecting have robbed them of the capacities of lution of cooperation. Science 242: 1385-1 390. particular comparisons. By clear theo- bats. 14 Winterhalder B (1996) A marginal model retical orientation, it collapses the dis- Behavioral ecology shows the impor- of tolerated theft. Ethol Sociobiol 17:37-53. 15 Hawkes K (1992) On sharing and work. tinction between “conceptual” and tance of theory-driven studies. Recent Curr Anthropol33:404-407. “referential”models.69In principle, ra- field and laboratory observations of 16 Hawkes K (1993) Why hunter-gatherers vens may tell us much about hominid flexible, highly tactical behaviors in a work An ancient version of the problem of variety of species generally have been public goods. Curr Anthropol34:341-361. scavengers as will chimpanzees, wild 17 Stephens DW, Krebs JR (1986) Foraging dogs, or even modern foragers. When theory-driven. Behavioral ecology Theory. Princeton:Princeton University Press. this type of comparative analysis is models often have predicted the evo- 18 Giraldeau Luc-A, Hogan JA, Clinchy MJ lution of sophisticated means of self- (1990) The payoffs to producing and scroung- routine in paleoanthropology, the field ing: What happens when patches are divis- will have advanced significantly as an interest prior to the empirical work ible? 85:132-146. evolutionary science.70 revealing their presence. Risk-averse 19 Caraco T, Giraldeau Luc-A (1991) Social foraging choices72 and facultative foraging: Producing and scrounging in a sto- Behavioral ecologV provides an alter- chastic environment. J Theor Biol 153:559- native to unilinear evolutionism. Key shifts in scrounging behaviorZ1are ex- 583. issues in paleoanthropolgy have cen- amples. In like fashion, paleoanthro- 20 Vickery WL, Giraldeau Luc-A, Templeton pology surely will benefit from greater JJ, Kramer DL, Chapman CA (1991) Produc- tered on the acquisition of traits that ers, scroungers, and group foraging. Am Nat occur regularly in extant foraging so- attention to behavioral ecology theory 13 7847-863. cieties but presumably were absent in and observations. We might take a cue 21 Giraldeau Luc-A, Soos C, Beauchamp G from Darwin himself and try to imag- (1994) A test of the producer-scrounger forag- our last common ancestor with the ing game in captive flocks of spice finches, great apes. The issues have been those ine the Descent of Man, had it been Lonchura punctulata. Behav Ecol Sociobiol of ordination; sequence and timing written with full awareness of these 34:251-256. new models and studies. 22 Hart TB, Hart JA (1993) The ecological ba- motivate the central debates. Was sis of hunter-gatherer subsistence in African sharing (or for that matter, hunting, rain forests: The Mbuti of eastern Zaire. In ACKNOWLEDGMENTS Pulford MH (ed), Peoples ofthe Ituri, pp 36-61. base camps, pair bonding, etc.) a ca- Fort Worth Harcourt Brace. pacity that was acquired early or late, I thank Sheryl Gerety, Flora Lu, Cot- 23 Bailey RC, Peacock NR (1993) Efe Pygmies and with what consequences? Current tie Pasternak, the Evoh.itionary An- of northeast Zaire: Subsistence strategies in thropology referees, Eric Smith, Bram the Ituri Forest. In Pulford MH (ed),Peoples of thinking, leery of the tyranny of the the Ituri, pp 102-129. Fort Worth: Harcourt ethnographic record,” holds that Tucker, and Blake Weiler. Brace. these “advanced traits” were ex- 24 Hill K (1988) Macronutrient modifications REFERENCES of : An approach using tremely late. Thus, they were not for- indifference curves applied to some modern mative influences on the divergence 1 Borgerhoff Mulder M (1991) Human behav- foragers. Hum Ecol 16:157-197. from the pongids or panids, as was be- ioural ecology. In Krebs, JR,Davies NB (eds), 25 Kaplan H. Hill K (1985) Hunting ability Behavioural Ecology: An Evolutionary Ap- and among male Ache lieved in the 1970s. proach, 3rd ed. Pp 69-98. Oxford: Blackwell. foragers: Preliminary results. Curr Anthropol To a behavioral ecologist, this all 2 Cronk L (1991) Human behavioral ecology. 261131-33. looks uncomfortably like old debates Ann Rev Anthropol20t25-53. 26 Stanford CB (1995) Chimpanzee hunting 3 SmithEA( 1992) Human behavioral ecology: behavior and human evolution. Am Sci in social theory over the stages of uni- I. Evol hthropol1:20-25. 83 :256-26 1. linear evolutionism. Our models and 4 SmithEA( 1992) Human behavioral ecology: 27 Mesterton-Gibbons M, Dugatkin LA (1992) field work tell us that it is ecological 11. Evol Anthropol1:50-55. Cooperation among unrelated individuals: 5 Smith EA, WinterhalderB (1992) Evolution- Evolutionary factors. Q. Rev Biol67267-281. circumstance and evolutionary ary Ecology and Human Behavior. New York: 28 Connor RC (1995) Altruism among non- mechanism that allow us to explain Aldine de Gruyter. relatives: Alternatives to the ‘prisoner’s di- the presence and form of exchange 6 Kaplan H, Hill K (1985) Food sharingamong lemma.’ TREE 10184-86. Ache foragers: Tests of explanatory hypothe- 29 Brown JL (1983) Cooperation-A biolo- and other complicated behaviors. ses. Curr Anthropol26:223-246. gist’s dilemma. Adv Stud Behav 13:l-37. Taxonomic line or grade appear to 7SmithEA(1988)Riskanduncertaintyin the 30 O’Connell JF, Hawkes K, Blurton Jones N have weak predictive power in this ‘original affluent society’: Evolutionary ecol- (1988) Hadza scavenging: Implications for ogy of resource-sharing and land tenure. In Plio/Pleistocene hominid subsistence. Curr realm. This observation will be as true Ingold T,Riches D, Woodburn J (eds), Hunter- Anthropol29:356-363. of hominid species as it is of the Gatherers, Vol I: History, Evolution and Social 31 WinterhalderB (1990) Open field, common broader comparative materials that Change, pp 222-251. New York: Berg. pot: Harvest variability and risk avoidance in 8 Hawkes K (1992) Sharing and collective ac- agricultural and foraging societies. In Cash- stimulate it. Any or all of the types of tion. In Smith EA, Winterhalder B (ed), Evo- dan E (ed), Risk and Uncertainty in Tribal and resource transfers defined or exempli- lutionary Ecology and Human Behavior, pp Peasant Economies, pp. 67-87. Boulder, CO, fied here might have characterized 269-300. Hawthorne, NY Aldine de Gruyter. Westview Press. 9 de Waal FBM (1989) Food sharing and re- 32 WinterhalderB (1986) Diet choice, riskand any or all populations of hominids. ciprocal obligations among chimpanzees. J food sharing in a stochastic environment. J Forms of transfer presumably were lo- Hum Evol18:433-459. Anthropol Archaeol 5:369-392. cally specific. They may have ap- 10 Blurton Jones NG (1984) A selfish origin 33 Sigmund K (1993) Games ofLife: Explora- for human food sharing: Tolerated theft. Ethol tions in Ecology, Evolution and Behaviour. peared, developed, changed, and Sociobiol 5:l-3. London:Penguin. disappeared many times in many 11 Blurton Jones NG (1987) Tolerated theft, 34 Peterson N (1 993) Demand sharing: Reci- ARTICLES EvolutionaryAnthropology 57 procity and the pressure for generosity among 49 Soltis J, Boyd R, Richerson PJ (1 995) Can Changing views of their behavior. J Anthropol foragers. Am Anthropol95:860-874. group-functional behaviors evolve by cultural Archaeol4:292-327. 35 Hill K, Kaplan H (1993) On why male fora- ? An empirical test. Curr An- 62 Blumenschine RJ. Cavallo JA, Capaldo SD gers hunt and share food. Curr Anthropol34: thropol36:473494. (1994) for carcasses and early 70 1-7 10. 50 -David N (1990) The giving environ- hominid behavioral ecology: A case study and 36 Dyer PD, Minnegal M (1993) Are Kubo ment: Another perspective on the economic conceptual framework. J Hum Evol 27:197- hunters ‘show offs’?Ethol Sociobiol 14:53-70. system of hunter-gatherers. Cum Anthropol 213. 31:189-196. 37 Hawkes K (1993) Reply. Curr Anthropol 63 Heinrich B, Marzluff J (1995) Why ravens 34:706-710. 5 1 Kelly RL (1 995) The Foraging Spectrum: Di- share. Am Sci 83:342-349. 38 Smith EA (1993) Comment. Curr Anthro- versity in Hunter-Gatherer Lifeways. Washing- pol 34:356. ton: Smithsonian Institution Press. 64 Stiner MC (1993) Modern human origins 52 Sahlins M (1972) Stone Age Economics. - faunal perspectives. Ann Rev Anthropol 39 Winterhalder B (1997) Gifts given, gifts 22 55-82, taken: The behavioral ecology of intra-group, Chicago: Aldine de Gruyter. non- exchange. J Archaeol Res, in 53 Lee RB (1979) The !Kung Sun: Men, 65 Enloe JG, David F (1992) Food sharing in press. Women, and Work in a Foraging Society. Cam- the Paleolithic: Carcass refitting at Pincevent. 40Trivers RL (1971) Theevolutionof recipro- bridge: Cambridge University Press. In Hofrnan JL, Enloe JG (ed), Piecing Together cal altruism. Q Rev Biol46:35-57. 54Winterhalder B (1993) Work, resourtes and the Past: Applications of Refitting Studies in Ar- population in foraging societies. Man 28:321- chaeology, pp 296-3 15. Oxford: Tempvs Repa- 41 Boyd R (1988) Is the repeated prisoner’s ratvm. dilemma agood model of reciprocal altruism? 340. Ethol Sociobiol9211-222. 55 Hauser MD, Marler P (1993) Food-associ- 66 Wilkinson GS (1990) Food sharing in vam- 42 Trivers RL (1972) and ated calls in rhesus macaques (Macaca mu- pire bats. Sci Am 262:76-82 sexual selection. In Campbell B (ed), Sexual latta): I. Socioecological factors. Behav Ecol 67 Packer C, Ruttan L (1988) The evolution of Selection and the Descent of Man. pp 136-179. 4:194-205. cooperative hunting. Am Nat 132: 159-198. Chicago: Aldjne. 56 Hauser MD, Marler P (1993) Food-associ- 68 O’Connell JF (1 995) Ethnoarchaeology 43 Hamilton WD (1964) The genetical evolu- ated calls in rhesus macaques Macaca MU- needs a general theory of behavior. J Archaeol tion of social behaviour. I. J Theor Biol 7:17- latta): 11. Costs and benefits of call production Res 3:205-255. 51. and suppression. Behav Ecol4:206-212. 69 Stanford CB, Allen JS (1991) On strategic 44 Hamilton WD (1964) The genetical evolu- 57 Cartmill M, Pilbeam D, Isaac G (1986) One storytelling: Current Models of Human Behav- tion of social behaviour. 11. J Theor Biol 7:17- hundred years of paleoanthropology. Am Sci ioral Evolution. Curr Anthropol32: 58-61. 51. 7441 0-420. 70 Cartmill M (1990) Human uniqueness and 58 Isaac GL (1984) The archaeology of human 45 Feinman S (1979) An evolutionary theory theoretical content in paleoanthropology. Int of food sharing. SOCSci Inf 18:695-726. origins: Studies of the lower Pleistocene in J Primatol11:173-192. 46 Bliege Bird RL, Bird DW (1 996) Delayed East Africa 1971-1981. Adv World Archaeol reciprocity and tolerated theft: The behavioral 3:l-87. 71 Wobst HM (1978) The archaeo-ethnology ecology of food-sharing. Curr Anthropol, in 59 Toth N, Schick KD (1986) The first million of hunter-gatherers or the tyranny of the eth- press. years: The archaeology of protohuman cul- nographic record in archaeology. Am Antiq- 47 Wilkinson GS (1988) Reciprocal altruism ture. Adv Archaeol Method Theory 9: 1-96. uity 43:303-309. in bats and other . Ethol Sociobiol 60 Blumenschine RJ (1991) Breakfast at Olor- 72 CaracoT, Martindale S, Whittam TS( 1980) 9:85-100. gesailie: The natural history approach to Early An empirical demonstration of risk-sensitive 48 Wilson DS, Sober E (1994) Reintroducing Stone Age archaeology. J Hum Evol 21:307- foraging preferences. Anim Behav 282320- group selection to the human behavioral sci- 327. 830. ences. Brain Behav Sci 17585454. 61 Binford LR (1985) Human ancestors: 0 1997 Wiley-Liss, Inc.

Books Received

Byrne R (1995) The Thinking Ape: versity Press. xvii + 428 pp. ISBN York: Oxford University Press. vi + Evolutionary Origins ofIntelligence. 0-1 9-509597-9 (cloth). 339 pp. ISBN 0-19-509581-2. New York: Oxford University Press. Williams GC (1996) Adaptation and $24.95 (paper) ix + 266 pp. ISBN 019852188X. Natural Selection. Princeton: Prince- Hersey GL (1996) The Evolution of $52.50 (cloth). ton University Press. xv + 307 pp. Allure: Sexual Selection from the Primack RB, Lovejoy TE (eds) ISBN 0-691-02615-7.$14.95 (paper). Medici Venus to the Incredible Hulk. (1995) Ecology, Conservation, and Bonner JT (1 996) Sixty Years of Bi- Cambridge: The MIT Press. xvi + Management of Southeast Asian ology: Essays on Evolution and De- 219 pp. ISBN 0-262-08244-6. $30.00 Rainforests. New Haven: Yale Uni- velopment. Princeton: Princeton (cloth). versity Press. vii + 304 pp. ISBN 0- University Press. x + 143 pp. ISBN Hauser MD (1996) The Evolution of 300-06234-6. $35.00 (cloth). 0-691-02130-9. $24.95 (cloth). Communication. Cambridge: The Walker A, Shipman P (1996) The Markova I, Graumann CF, and K MIT Press. xii + 592 pp. 0-262- Wisdom of Bones: In Search of Hu- Foppa (eds) (1995) Mutualities in 08250-0. $55.00 (cloth). man Origins. New York: Alfred A. Dialogue. New York: Cambridge Vrba ES, Denton GH, Partridge TC, Knopf, Inc. x + 338 pp. ISBN 0-679- University Press. xvii + 282 pp. Burckle LH (eds) (1995) Paleocli- 42624-8. $26.00 (cloth). ISBN 0-521-49595-4.$54.95 (cloth). mate and Evolution with Emphasis on Shore B (1996) Culture in : Buss DM, Malamuth NM (1996) Human Origins. New Haven: Yale Cognition, Culture, and the Problem Sex, Power, Conflict: Evolutionary University Press. v + 547 pp. ISBN 0- of Meaning. New York: Oxford Uni- and Feminist Perspectives. New 300-06348-2. $85.00 (cloth).