Teoría de señales, interacción estratégica y capital simbólico

221

C u r r e n t A n t h ro p o l o g y Volume 46, Number 2, April 2005� 2005 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204/2005/4602-0003$10.00

Signaling Theory,Strategic Interaction,and Symbolic Capital1

by Rebecca Bliege Bird andEric Alden Smith

Signaling theory provides an opportunity to integrate an interac-tive theory of symbolic communication and social benefit withmaterialist theories of individual strategic action and adaptation.This article examines the potential explanatory value of signal-ing theory for a variety of anthropological topics, focusing onthree social arenas in which signaling might plausibly be impor-tant: unconditional generosity, “wasteful” subsistence behavior,and artistic or craft traditions. In each case, it outlines the waysin which the phenomena correspond with the expectations ofsignaling theory by showing how a given pattern of action mightsignal particular hidden attributes, provide benefits to both sig-naler and observers, and meet the conditions for honest commu-nication. The ethnographic evidence suggests that the fundamen-tal conditions for reliable signaling of condition-dependentqualities may exist in many social domains. It appears that sig-naling theory has considerable promise for generating novel andpowerful insights into the ethnographic realm.

r e b e c c a b l i e g e b i r d is Assistant Professor in the Depart-ment of Anthropological Sciences at Stanford University (Stan-ford, CA 94305-3052, U.S.A. [[email protected]]). Born in 1967,she was educated at Pomona College (B.A., 1990) and the Univer-sity of California, Davis (Ph.D., 1996). She is currently engagedin a long-term research project in the Western Desert of Austra-lia on men’s and women’s hunting and land management strate-gies. Her publications include (with D. W. Bird) “Delayed Reci-procity and Tolerated Theft: The Behavioral Ecology of FoodSharing Strategies” (current anthropology 38:49–78), (with E.A. Smith and D. W. Bird) “The Hunting Handicap: Costly Signal-ing in Male Foraging Strategies” (Behavioral Ecology and Socio-biology 50:9–19), and (with D. W. Bird, E. A. Smith, and G.Kushnick) “Risk and Reciprocity in Meriam Food Sharing” (Evo-lution and Human Behavior 23:297–321).

e r i c a l d e n s m i t h is Professor of Anthropology at the Uni-versity of Washington. He was born in 1949 and received hisB.A. (1972) from the University of California, Santa Barbara, andhis M.A. (1976) and Ph.D. (1980) from Cornell University. His re-search focuses on ecological, economic, and evolutionary aspectsof production, reproduction, and decision making. Among hispublications are Inujjuamiut Foraging Strategies (Hawthorne,N.Y.: Aldine, 1991), Evolutionary Ecology and Human Behavior(Hawthorne, N.Y.: Aldine, 1992), and numerous articles on hu-man behavioral ecology.

The present paper was submitted 7 xii 03 and accepted 4 viii 04.

1. For helpful discussion of the theory and cases in this paper, wethank Carl Bergstrom, Jim Boone, Monique Borgerhoff Mulder, SamBowles, Doug Bird, Herb Ginits, Craig Hadley, Stevan Harrell, Kris-ten Hawkes, Frank Marlowe, and Polly Wiessner. For comments

Materialist and evolutionary analyses of subsistence,economic exchange, mating and marriage, and parentinghave become increasingly sophisticated and detailed (seereviews in Smith and Winterhalder 1992, Wilk 1996, Vo-land 1998, Hill and Kaplan 1999, Cronk, Chagnon, andIrons 2000, Winterhalder and Smith 2000). The power ofthese analyses lies in the focus on individuals as strategicdecision-makers and on empirically tractable measuresof costs and outcomes. While this approach has hadmany successes, the more ritualized and communal as-pects of social behavior, those that appear to be drivenby cultural meaning and collective interest more thanby individual gain, have proven difficult to explain.

We wish to argue that the ostensible limitations of theindividual-strategist approach as currently conceived aredue in part to an overly narrow conception of motivationand social interaction. While evolutionary and ecologicalanthropologists have rightly emphasized the importanceof such components of adaptive success as food acqui-sition, mate choice, and resource competition, they havepaid insufficient attention to social, symbolic, and pres-tige-related aspects of individual strategizing. Thesecomponents of cultural behavior may appear to generatebenefits that are unmeasurable or that favor social groupsat the expense of their individual members, but recentempirical and theoretical work, as well as a long-stand-ing body of social theory, suggests an alternative inter-pretation. Using “costly-signaling” or “handicap” the-ory, anthropologists, ecologists, and biologists havebegun to integrate altruistic or seemingly irrational be-haviors into adaptationist and strategic analyses of de-cision making and social behavior.

There also exists a long-standing parallel set of theo-retical concepts in social theory dating back to ThorsteinVeblen and Marcel Mauss and more recently refined byPierre Bourdieu. For these theorists, individually costlybut collectively beneficial (or at least prestigious) behav-iors such as public generosity or extravagant piety are aform of social competition: the most generous or self-sacrificial individuals gain higher prestige, and the re-cipients or observers gain material benefit at the expenseof their own prestige (Veblen 1994 [1899], Mauss 1924,Fried 1967). In its various guises, this approach has beenknown as conspicuous consumption, wasteful advertis-ing, and the accumulation of symbolic capital. Our goalhere is to show how these various evolutionary, eco-nomic, and social theories about communication andstatus can be unified within a single theoretical frame-work: signaling theory.

We believe that signaling theory can offer not only a

on earlier versions of the manuscript (not always heeded), we aregrateful to Boone, Borgerhoff Mulder, Bowles, and Wiessner, as wellas Brenda Bowser, Ben Fitzhugh, Sarah Hrdy, Richard McElreath,Fraser Neiman, Paul Roscoe, and two anonymous reviewers. TheMeriam research was supported by the National Science Founda-tion (SBR-9616096 and SBR 9616887 and predoctoral awards to RBBand D. W. Bird), the Wenner-Gren Foundation, the L. S. B. LeakeyFoundation, and the Australian Institute for Aboriginal and TorresStrait Islander Studies and most importantly by the gracious hos-pitality and logistical support of the Meriam people.

222 F current anthropology Volume 46, Number 2, April 2005

rigorous and systematic framework for combining vari-ous convergent ideas about status competition but pro-vocative new insights into exactly those domains of hu-man behavior and culture that have appeared mostresistant to adaptationist analysis. Signaling theory pro-vides a way to articulate idealist notions of the intangiblesocial benefits that might be gained through symbolicrepresentations of self with more materialist notions ofindividuals as self-interested but socially embedded de-cision makers. By paying attention to the problem of howcredibility is maintained when individuals have to makeinterdependent decisions (about mates, alliances, con-flict, and trust) based on incomplete information, sig-naling theory provides a new interpretation of such sym-bolic performance as aesthetic elaboration, initiationrites, ethnic boundaries, ceremonial feasting, wealth cir-culation, conspicuous consumption, monumental archi-tecture, religious commitment, and the individuallycostly provisioning of collective goods.

The paper is organized as follows: We first discuss thehistorical parallels in social and biological theories ofsignaling and summarize the structure of costly-signal-ing theory and the conditions under which it may beapplied. We then turn to some ethnographic illustrationsof its potential explanatory power. We focus on threelikely arenas for signaling dynamics: unconditional gen-erosity, “wasteful” subsistence activities, and artisticelaboration. We conclude with a discussion of the com-plexities raised in applying signaling theory to the ex-planation of cultural variation.

Conspicuous Consumption and SymbolicCapital

Thorstein Veblen’s theory of the leisure class first drewattention to the notion that wasteful expenditures oftime and money and conspicuous displays of lack of in-terest in economic profit may function as a means ofgaining competitive advantages over others. For exam-ple, Veblen wrote: “If, in addition to showing that thewearer can afford to consume freely and uneconomically,it can also be shown in the same stroke that he or sheis not under the necessity of earning a livelihood, theevidence of social worth is enhanced in a very consid-erable degree” (Veblen 1994 [1899]:105). It is clear fromthis and numerous other passages that Veblen’s con-spicuous consumption is a form of costly signalingwherein the underlying attribute being signaled is ex-traordinary wealth. Veblen proposed that signaling inthis manner enhanced social status when knowledge ofothers’ qualities was not widely known or could be un-reliable, as in situations of high socioeconomic mobility.This provided a ready explanation of the distinction be-tween “old money”—the class of people who did notneed to engage in such conspicuous displays becausetheir wealth was already common knowledge (at leastamong those to whom they wanted it to be known)—

and the nouveau riche, who needed to advertise theirnewly acquired wealth.

Veblen also realized that variations in the cost or in-tensity of the display allowed fine distinctions to bemade between individuals who are competing withinspecific social arenas for status and its perquisites (suchas quality mates and powerful allies) (1994:32–33):

The greater the degree of proficiency and the morepatent the evidence of a high degree of habituationto observances which serve no lucrative or other di-rectly useful purpose, the greater the consumption oftime and substance impliedly involved in their ac-quisition, and the greater the resultant good repute.. . . Differences between one person and another inthe degree of conformity to the ideal in these re-spects can be compared, and persons may be gradedand scheduled with some accuracy and effect accord-ing to a progressive scale.

Veblen focused on such “uneconomical” displays in or-der to draw attention to the shortcomings of classicaleconomic theory, which presumed that what was goodfor the individual was also good for the group. His con-tribution to economic theory, while not widely recog-nized until quite recently, was to point out that the pur-suit of self-interest, rather than leading to collectivelybeneficial outcomes, merely led to the waste of time andresources in elaborate social competition.

While Veblen focused on the social waste of conspic-uous advertisement in capitalist society ca. 1900, MarcelMauss tackled the conspicuous generosity characteristicof precapitalist societies with an eye to criticizing thenotion that the benefits of such phenomena as the kulaand potlatch lay in narrow economic notions of profitmaximization. As Mauss was at pains to demonstrate, insuch gift-giving systems transfers for immediate economicgain were “viewed with the greatest disdain” (1924:36).Confronted with seemingly irrational exchange behaviorin which wealth transactions were immediately costlyand did not seem to produce material gain and in whichreceiving wealth had to be countered with equal orgreater giving in order to maintain status, Mauss con-cluded that conspicuous giving was based on a super-structural economy of prestige: “Nowhere else is theprestige of an individual as closely bound up with ex-penditure. . . . Consumption and destruction are virtuallyunlimited. . . . The rich man who shows his wealth byspending recklessly is the man who wins prestige.” Par-ticularly in cases of intensified social competition, achief could maintain political authority only throughdemonstrations of quality that put his rivals “in theshadow of his name.” Mauss, however, did not offer atheory of prestige at the level of the individual actor,instead claiming that the benefits of such costly gener-osity lay in the structural integration of culture. In thissense, he and Veblen offered very different theories ofsocial status and conspicuous display.

Pierre Bourdieu’s An Outline of a Theory of Practice(1977) has reintroduced the individual strategic actor

bliege bird and smith Signaling Theory, Strategic Interaction, and Symbolic Capital F 223

into a theory of prestige via his concept of “symboliccapital.” Bourdieu argues that the accumulation of sym-bolic capital is just as “rational” as the accumulation ofeconomic capital, particularly since such capital may befreely converted from one form to another, ultimatelyin order to gain advantages in the form of additionalwealth, power, allies, and marriage partners. His critiqueof rational action does not pit idealism against materi-alism but rather broadens the concept of rationality toinclude the pursuit of social gains, specifically symboliccapital. For Bourdieu, the highest profits in symbolic cap-ital can be attained when individuals act in ways thatreliably demonstrate lack of interest in material acqui-sition by engaging in conspicuous consumption or con-spicuous generosity. The value of the display in termsof its symbolic capital lies in the cost of the investmentin terms of time, energy, or wealth (Bourdieu 1977, 1990;Turner 1991). Bourdieu explains that behaviors that ap-pear at first glance to be economically “absurd,” such aspurchasing an ox simply to show that one has surplusresources, actually enhance a family’s symbolic capital,their “credit of renown,” at a time when marriage ne-gotiations are critical (1990:120). He argues that this “ex-hibition of symbolic capital (which is always very ex-pensive in economic terms)” increases the family’s socialstanding by displaying the quality of its goods, which isin turn linked to the attributes of its members. The re-sultant high social standing allows the family to, for ex-ample, acquire powerful affines through marriage andadditional material wealth and (in the Kabyle case atleast) to demonstrate the ability of its male members todefend and preserve its land and women.

Bourdieu further refines the link between the individ-ual attributes and the symbolic value of the display inDistinction (1984:281):

The objects endowed with the greatest distinctivepower are those which most clearly attest the qual-ity of the appropriation, and therefore the quality oftheir owner, because their possession requires timeand capacities which, requiring a long investment oftime, like pictorial or musical culture, cannot be ac-quired in haste or by proxy, and which therefore ap-pear as the surest indications of the quality of theperson. This explains the importance which the pur-suit of distinction attaches to all those activitieswhich, like artistic consumption, demand pure,pointless expenditure, especially of the rarest andmost precious thing of all . . . namely, time.

The convergence here with Veblen is remarkable (Trigg2001). For Bourdieu, as for Veblen, conspicuous gener-osity is no different from conspicuous expenditure: bothare ultimately strategic actions designed to accumulatesymbolic capital. Both involve significant costs whichfunction to attach value to symbolic capital. Both schol-ars, and Mauss to a lesser extent, seem to have recog-nized this solution to the paradox of economically ir-rational behavior. The costs of such strategies areoutweighed by the benefits gained through manipulating

social relationships with other individuals. The apparentparadox of wastefully expending time and wealth is dis-solved if the cost of the display functions to ensure thatonly high-quality individuals can afford them at all.Thus, the signal value of conspicuous consumption ismaintained by its costs; these costs in turn are the pricewealthy individuals pay for prestige.

However, there are two elements missing from theseearly formulations of signaling theory. First, costlinessis not always necessary to guarantee signal value. Somesignals maintain symbolic value because they are index-ically related to that which they signify: they are simplyimpossible to fake. For example, the head flattening prac-ticed by some Native American tribes to distinguish thefree-born from slaves was a reliable signal not becauseof its cost but because it could be performed only forchildren on the initiative of free-born parents. Second,while Bourdieu’s notion of symbolic capital explainshow costs ensure honesty, none of these accounts of sig-naling theory explain why an honest display confers pres-tige. Why would it be to the benefit of others to defer toone who signals his or her social or economic superi-ority—in other words, why tolerate inequality?

The answer to this comes from the evolutionary-eco-logical theory of costly signaling, which notes that thereare mutual benefits to be gained from truthful com-munication. Those seeking assurance that a given in-dividual has sufficient personal resources or belongs toa kin group of sufficient resource holdings or productiv-ity to qualify as either an equal or a social superior donot need to put their faith in words but can examine theevidence of deeds, such as displays of generosity or waste,that are too costly to be worth faking. More broadly,signal cost (actual or potential) can serve as a powerfulmeans of guaranteeing honesty and thus allow observersto gauge the relevant hidden qualities of potential allies,mates, or competitors. Inequality is tolerated when sig-nalers demonstrate their competitive superiority, anddeference (or interest in the signal) provides greater ben-efits than resistance (or ignoring the signal).

The Evolutionary Ecology of Signaling

There are several approaches to the evolution of signalswithin contemporary evolutionary biology: the “run-away” approach, the sensory-exploitation view, and thecostly-signaling or handicap approach. Although theseapproaches can be reconciled with a single broad theo-retical framework (Kokko et al. 2003), they are distinctenough that we feel obliged to describe the first twobefore concentrating on the third. In the runaway ap-proach, pioneered by Fisher (1930), relatively arbitraryobserver preferences (e.g., for mates with showy plum-age) that are correlated via genetic inheritance to signals(e.g., because offspring inherit both the preference andthe signal) can lead to the elaboration of signals overevolutionary time as preferences and signals coevolve.While formal models show that the Fisherian runawayprocess could favor the evolution of signals with no adap-


tive value other than their effect in attracting mates(Pomiankowski, Iwasa, and Nee 1991) and the runawaylogic can be extended to models of cultural evolution(Boyd and Richerson 1985), there is relatively little directempirical evidence for such processes. The sensory-ex-ploitation view of signal evolution shares with the run-away approach the assumption that observers have pre-existing preferences (perhaps evolved in other contexts)which lead them to be attracted to certain signals andthat signals which “exploit” these preferences by beingeven more elaborate (so-called supernormal stimuli) canbe favored even if they are relatively arbitrary or mala-daptive (Ryan 1998).2

We focus on the costly-signaling approach here be-cause its focus on communication, direct parallels withsocial theory, and attention to strategic action make itthe most clearly relevant to anthropological contexts. Aswith signaling in social theory, costly-signaling theoryinvolves the communication of attributes that are rela-tively difficult or expensive to perceive directly and thatvary in quality, intensity, or degree between signalers(either groups or individuals). The fundamental problemaddressed by costly-signaling theory is how individualswith partially competing interests may nevertheless mu-tually benefit from signaling these differences in quality.For this mutual gain to be realized, the two parties mustfind ways to guarantee at least partial honesty of theinformation they communicate. The fact that they dohave competing interests would seem to lead inevitablyto deceit, which in turn should cause signals to be un-reliable and eventually ignored, thus leading to completefailure of communication. But in the real world, bothroutinely misleading signals and unconditional honestyseem to be relatively uncommon. The costly-signalingapproach suggests that the persistence of reliable com-munication emerges from conditions ensuring that hon-esty is the best move in the game of communication:both players do better by cooperating (giving an honestsignal or paying attention to the information) than byattempting to defect (lying about one’s intentions orqualities or ignoring the signal).

This fundamental problem of reliable signaling can beillustrated with a classic example from the realm of non-human behavior. Many ungulates (deer, antelopes, etc.)react to a nearby predator by seemingly paradoxical be-havior—displaying a prominent rump patch, leapinghigh into the air rather than running straight away, andso on. These actions are paradoxical because they seemdesigned to attract the predator’s attention and act toslow the prey’s progress in escaping. The paradox di-minishes if these actions are viewed as signals from theprey that tell the predator “I have detected you, and Iam in excellent condition, so it is a waste of your time

2. Without the correlated inheritance element, however, there isno “runaway” process, and the evolutionary equilibrium is reachedwhen signals match the preexisting sensory bias; considerable ev-idence for signal evolution via sensory exploitation has been col-lected among animal populations (Endler and Basolo 1998), and itis easy to imagine how this analysis might be extended to culturalphenomena such as advertising, propaganda, and pornography.

to chase me.” If this is a truthful signal, the predatorbenefits by saving the energy expended in a chase thatwill likely be fruitless, and the prey obviously benefitsby avoiding the cost of a lengthy and tiring escape (andperhaps the even greater cost of becoming dinner).

But why should such a signal be credible? After all,prey are not interested in benefiting their predators (orvice versa). It would seem that an antelope in substan-dard condition (and hence a slower runner) would benefiteven more from signaling that it was in good condition,in which case misleading signals would proliferate in theprey population until predators learned (or were designedby natural selection) to ignore them. The standard so-lution to this dilemma offered in signaling theory is thatsignal honesty can be favored and thus an “honest sig-naling equilibrium” maintained if the costs of signalingare structured in such a way that dishonest signals donot pay. One way in which this condition might be metin the predator-prey case is if predators “tested” suchsignals (e.g., by pursuing prey for a short distance or onsome proportion of encounters). Then prey in poor con-dition would improve their chances of survival by for-going the signal and devoting all their time and energyto (a) remaining inconspicuous or (b) running away di-rectly once detected and predators would generally re-ceive truthful signals. Empirical tests of these hypoth-eses (using data on African ungulates) indicate that mostare well supported (Caro 1994, FitzGibbon and Fanshawe1988).

The conditions for the evolutionary or game-theoret-ical stability of costly signaling can be specified as fol-lows (for more technical analyses, see Grafen 1990, John-stone 1997, Gintis, Smith, and Bowles 2001):

1. Members of a social group vary in some underlyingattribute (physical condition, resource endowment, need,motivation) that is difficult or impossible to observe butcould in principle be reliably signaled.

2. Observers stand to gain from accurate informationabout this variation in attribute quality.

3. Signalers and recipients have conflicting interestsin the sense that successful deceit (e.g., appearing to bea fleeter prey, a more dangerous adversary, or a moreattractive mate than is in fact the case) would benefitsignalers at the expense of the recipients.

4. Signal cost or benefit (to the signaler) is quality-dependent (i.e., the marginal cost of the signal is nega-tively correlated with the signaler’s quality or the mar-ginal benefit is positively correlated with the same).

When these conditions are met, we can expect one ormore “design forces” (i.e., decision making, subcon-scious learning, natural selection, or adaptive culturaltransmission) to favor a system of communication con-forming to the costly-signaling framework.

Costly-signaling theory appears to provide a coherentframework that ties together a host of disparate socialphenomena and unifies some seemingly unrelatedstrands of social theory, but does it really have wide ex-planatory power in human affairs? While it is much tooearly to make a firm assessment, a preliminary surveysuggests that it probably does.


Signaling Theory and the EthnographicRecord

Signaling theory offers an intriguing opportunity for rec-onciling the strategic-actor and materialist approachesin the social sciences with approaches centered on mean-ing, social value, and ritual. While many social anthro-pologists have long recognized the role of symbolic cap-ital in motivating individual behavior, they have gener-ally been averse to both formal theory and quantitativeempirical tests (i.e., to conventional methods of science).In turn, economic, ecological, and evolutionary anthro-pologists have been reluctant to include such factors intheir analyses because of this very absence of a rigorous,testable theory describing how individuals might ma-terially benefit from socially meaningful but economi-cally costly behavior. Signaling theory allows us to ad-dress issues of symbolic value with rigorous empiricaldata and a set of testable predictions derived from a bodyof theory that is linked to individual strategizing andevolutionary dynamics.

To illustrate the utility of signaling theory, we ex-amine three social arenas in which signaling might plau-sibly be important: unconditional generosity, “wasteful”subsistence behavior, and artistic or craft traditions. Ineach case, we outline the ways in which the phenomenaplausibly correspond with the expectations of signalingtheory, noting how a given pattern of action might (1)signal a particular hidden attribute, (2) provide benefitsto both signaler and observers, and (3) demonstrate howsignals of attribute quality remain credible. While suchplausibility arguments cannot take the place of rigoroushypothesis testing, at this early stage of the applicationof signaling theory to ethnographic cases they are a val-uable first step.

signaling and unconditional generosity

Sahlins’s classic formulation of reciprocity modeled it asa continuum ranging from exchanges which were “al-truistic,” “disinterested,” and “unconditional” throughthose which were “mutualistic” and “balanced” to thosewhich were purely “self-interested” or “unsociable”(1972:195). Sahlins’s model proved to be the source of agreat deal of contention, particularly his notion of un-conditional exchange. Many suggested that there was nosuch thing as a purely altruistic, disinterested gift; giftswhich may appear unconditional and altruistic eitherprovide benefits in terms of symbolic capital or prestige(as argued by Bourdieu), create and maintain social re-lationships or social structure benefiting both parties (asargued by Levi-Strauss as well as structural Marxists), orare given under conditions of asymmetric need, with theexpectation that a corresponding return will occur atsome later date when the asymmetry is reversed.

This latter view was first modeled in evolutionary bi-ology as “reciprocal altruism” (Trivers 1971) and givenwide currency by computer simulation models showinga “tit-for-tat” strategy (cooperate if your partner coop-

erates, defect if your partner defects) to be quite robustcompared with less friendly alternatives (Axelrod andHamilton 1981, Axelrod 1984). Many researchers sub-sequently used reciprocal altruism as a way to explainthe widespread practice among hunter-gatherers (andothers) of widely sharing individually harvested re-sources, particularly meat from large animals. Sharingreciprocally is commonly explained in terms of the eco-logical benefits gained by reducing the risk associatedwith acquiring such productive but highly variable re-sources (Smith 1988). Pooling such resources smoothesout consumption variance for all participants, a consid-erable benefit when harvests are so unpredictable thatindividual hunters can expect many days or weeks toelapse between successes (Winterhalder 1990). Otherbenefits that might be obtained through the mechanismof reciprocal altruism include a sort of “health insurancepolicy”: food is shared with the expectation that in timesof illness, when foraging is impossible, one can dependon one’s previous sharing partners (Gurven et al. 2000a,Sugiyama and Scalise Sugiyama 2003).

While these risk-minimization effects of food sharingare plausible and can be demonstrated to exist (Cashdan1985, Kaplan, Hill, and Hurtado 1990), they may well bea welcome but unanticipated outcome of rather than amotive for food sharing. In any case, sharing in order toreduce risk or ensure future aid where no formalizedinsurance institutions exist poses a major dilemma forthe giver. He or she must discriminate among those whoreceive, withholding food from individuals who will notreturn the favor in the future on the basis of the oftennoisy and ambiguous evidence of the past. If there is nodiscrimination against nonreciprocators, sharing entailsa prisoner’s dilemma payoff structure (Smith and Boyd1990) in which there are incentives to be among therecipients and not among the givers—to free-ride on theefforts of others (Blurton Jones 1986, Hawkes 1993).

The evidence for such discrimination (sharing condi-tional on reciprocation) is inconsistent. Detailed analysisdoes reveal patterns of conditional reciprocity in inter-household food sharing among several Amazonian In-dian forager or forager-horticultural peoples (Ache livingin permanent settlements [Gurven et al. 2001], Hiwi[Gurven et al. 2000b], Yanomamo [Hames 2000]). Butequally detailed evidence suggests that lack of attentionto a recipient’s past history or future probability of re-ciprocating is characteristic of food-sharing patterns inother widely distributed cases, such as the Ache of theParaguayan forest while on trek (Kaplan and Hill 1985a),the Hadza of the East African savanna (Hawkes 1993,Hawkes, O’Connell, and Blurton Jones 2001), and theMeriam of Torres Strait, Australia (Bliege Bird and Bird1997, Bliege Bird et al. 2002). In these cases (and othersthat have been described in less detail), at least sometypes of harvested resources are shared unconditionallywith most or all members of the community, and somehunters consistently provide more than others whilesharing more or less equally in the catch. These “altru-istic” providers in fact enjoy higher social status andreproductive success than their less productive peers, de-


spite the absence of any conditional exchange of meatfor these social perquisites (Kaplan and Hill 1985b, Mar-lowe 2000, Bliege Bird, Smith, and Bird 2001). Enhancedsocial status and its subsequent political and reproduc-tive advantages could be the “selective incentive” (Olson1965) that motivates certain individuals to provide col-lective goods.

This insight has led to an active debate about how toexplain the undisputed fact of extensive food sharing insmall-scale societies (Winterhalder 1996, Wenzel, Hov-erlsrud-Broda, and Kishigami 2000). Two distinct typesof explanations have emerged as an alternative to recip-rocal altruism to account for unconditional generosity:(1) strong reciprocity, in which at least some individualsare motivated to punish free-riders irrespective of thecosts to themselves as a consequence of a history of cul-tural or genetic group selection (Bowles and Gintis 1998,2003; Boyd et al. 2003; Fehr and Gachter 2002; Gintis etal. 2003), and (2) “symbolic-capital” explanations, whichemphasize the social benefits that successful individualsobtain as a result of their generosity (Boone 1998, Gurvenet al. 2000a, Hawkes 1993, Smith and Bliege Bird 2000,Wiessner 2002, Bodenhorn 2000).

Costly signaling is a symbolic-capital explanation forsuch sharing that focuses on its benefits for both giversand receivers. It applies particularly to displays whichare characterized by (1) the extension of consumptionrights to multiple others regardless of their exchange re-lationship to the “giver,” (2) distribution or consumptionin a social arena in which knowledge of the distributionis transmitted to multiple others, and (3) the dependenceof the ability to produce the display upon some hiddenattribute of the donor in which observers (who may ormay not be recipients of the material donation) have asignificant interest. These conditions occur most readilyin contexts such as funerary rites (Smith and Bliege Bird2000), big-game meat distributions (Hawkes and BliegeBird 2002, Bodenhorn 2000), big-man feasting (Wiessnerand Schiefenhovel 1995), Northwest Coast Indian po-tlatching (Boone 2000), and charity galas in capitalistsociety (Veblen 1994 [1899]). We will expand on the firstof these, using ethnographic information from our ownfieldwork.

Meriam funerary feasts. Funeral ceremonies are someof the most elaborate and important rituals in Melane-sian societies. Among Meriam Islanders of Torres Strait,the occasion of any death engenders a series of expensivepublic ceremonies and feasting events. When a death isfirst announced, the deceased’s kin group (generally, apatrilineal clan) meets to select a leader for the eventsto follow and to elicit monetary contributions for mor-tuary expenses. The next day the bood, a public sittingnear the deceased’s home attended by kin and others,begins. The immediate family is responsible for feedingall who come to show respect for the deceased and thekin group. Donations of food are made by kinsmen andothers in the village, supplementing the hunting, fishing,and purchasing efforts of the immediate family. Thebood lasts for several days, as long as the immediatefamily wishes to mourn publicly and supply food to all

comers (a form of feasting termed bood lewer). A tem-porary cross and grave are constructed at the burial site(one of several elements reflecting Meriam colonial his-tory and conversion to Christianity in the late nine-teenth century). At the end of the bood, a public funeralfeast (izurwur lewer) announces the end of the formalperiod of mourning.

Roughly two to five years later, the family announcesthat the permanent tombstone is ready. These tomb-stones have developed into elaborate granite, tile, andconcrete structures with color photos, engravings,molded reliefs of totems and fruits, and other expensivedetails. The family hosts a series of public work-partyfeasts to engage helpers to ready the grave, and a finalpublic feast (kirim akos lewer) is held after the publicsetting of the headstone (kirim akos). Following thisevent, the date for the final “tombstone-opening” cere-mony is set (the “opening” being an unveiling of thetombstone, which has been covered from view since thesetting ceremony). This ceremony begins with publicwork-party feasts, which can start up to two monthsbefore the date of the penultimate feast. All those whocome to assist must be fed by the immediate family.Theoretically the entire island population (over 400)could attend, though attendance by half or less is typical.

The tombstone-opening feast (kirim auskir lewer) fea-tures tables laden with bowls of cooked fish, turtle, shell-fish, and other foraged, planted, and purchased items.Gifts intended for the guests such as sarongs, beads, andclothing are displayed publicly by tying them to thefeasting mood (shade structure) or placing them in largecontainers constructed for the purpose. Gardening menwho have managed to cultivate a particularly large (butwoody and mostly inedible) yam, sweet potato, or bunchof bananas donate their goods to be displayed on specialbamboo racks on the feasting grounds. Careful lists aremade to ensure that representatives of all family alliesare given gifts and seated at a special table. Feasts providea large audience for many other forms of competitionand display of skills, and contests involving top spinning,dart throwing, and especially dancing occur throughoutthe night.

The quantity and quality of resources displayed anddistributed at public feasts is a direct indication of thenumber of feast contributors, the time and energy theyhave invested in working for the feast giver, and theproduction skills of those who contribute big fish,hunted turtle, or display yams. A feast giver with manystrong allies is able to provide shade and comfortableconditions for the guests, to distribute huge amounts ofelaborately cooked food to them, and to give them ex-pensive material gifts. A feast giver gains in status whenhis allies provide prestige foods (mackerel, hunted turtle)or foods that take time to prepare (steamed manioc pud-dings, conch cooked in coconut milk). Men prefer to do-nate large packages of uncooked meat to public feasts,especially large pelagic fish, giant tridacnid clams, andsea turtles. Women also provide foraged foods for feasts,but their contributions tend to come in the form ofsmaller quantities of prized (often rare) foods that are


elaborately prepared. Some men do prepare feastingfoods, but women in particular gain some measure ofnotoriety for their cuisine and the time they invest inpreparation, while men tend to gain more from display-ing their skill in acquisition (Smith and Bliege Bird 2000,Bliege Bird, Smith, and Bird 2001).

The Meriam to whom we spoke clearly do not viewthese feasts as instances of conditional reciprocity, inwhich guests are expected to repay the giver with anotherfeast at a later date. Rather, they are viewed as expres-sions of the Meriam conception of generosity, which in-volves reference to debe tonar, a phrase meaning “thegood way.” Following debe tonar is said to provide long-term benefits in the form of an enhanced social repu-tation as a “good person.” Subsistence decisions inter-sect with the aspect of debe tonar that incorporatesgenerosity with food. Meriam do not approve of condi-tional sharing of “table food” (derapeili, “to share por-tions out”) between households following the harvestingof wild or cultivated foods. Likewise, there are socialnorms governing the provisioning of food to feasts: suchfoods are explicitly designated public goods (kies), opento unconditional consumption. To share contingentupon a return in kind is to share selfishly, and to recip-rocate a gift freely given is to imply that the giver hadselfish motives. Debe tonar states that such uncondi-tional sharing is its own reward. A Meriam adage goes,“When you expect payment back, you don’t make a last-ing impression.” Sharing selfishly is equivalent to notsharing at all; both are identified as gobar, “greedy withfood.” While derapeili and kies sharing are explicitly un-conditional, there are contexts within which conditionalreciprocal exchange is considered proper: markets(tama), labor exchange (irapu, work-party feasts), cere-monial exchange partners (wauri tebud), and repayments(bodomelam) for use-rights to certain individually or cor-porately owned items (land or foraging areas or boats andtools).

How do Meriam feasts articulate with costly-signalingtheory? Feasts involve a form of collaborative signalingwherein the members of a lineage, clan group, or othersocial collective cooperate to produce a display indicat-ing some socially important attribute of the social group.In the Meriam case, the final tombstone-opening feastis organized by a corporation composed of consanguinealrelatives who cooperate to elicit wealth contributionsfrom individuals more distantly related to the deceased.There are two classes of recipients or signal observersor, as the Meriam would describe it, two circles: the innercircle of families related through marriage or alliancewho are the intended recipients of the accumulatedwealth objects and the outer circle of more distantly re-lated people who receive only the accumulated bountyof feast food. As many Meriam describe it, the immediatefunction of honoring the inner circle is to reaffirm theirclose relationship even though the death of one of theirlineage members has the potential to change the rela-tions of power between them: “We honor a member ofeach family with gifts to show we have not forgottenthem.” In this sense, the Meriam tombstone-opening

ceremony has some fundamental similarities withKwakiutl potlatches as described by Boone (2000), but itdiffers in that it seems to be designed less to reinforcethe status of a particular leader than to cement alliancesbetween key patrilineages on the death of a crucial lin-eage member.3

We suggest that one function of investing in costlydisplays of wealth and resources through feasting is tosignal lineage “strength,” which may be defined as thequalities of and cohesion among individual members.Here, cohesion is measured by the strength of the alli-ances between its members, as indicated by the laborthey are willing to invest in building the feasting groundand the costs they are willing to pay to contribute boththe monetary cost of purchasing and donating the giftsof cloth and other material items and the time/energycost of acquiring or preparing particular foods. Signalingtheory suggests that such reliable displays of cohesionshould be particularly important following the death ofa powerful and charismatic lineage member (Boone, per-sonal communication). Indeed, among Meriam and, wesuspect, in most other social groups, funerary displaysare more expensive and extravagant for socially central,politically powerful, and highly prestigious elders thanfor youngsters or marginal adults. There are many at-tributes that lineage members might signal through feastcontributions, among them their ability to invest time,to acquire costly or rare foods, or to give away wealthitems. Elaborately well-prepared items indicate skilledfemale cooks with ample time, the provisioning of largefish and turtle and the display of large garden produceindicate the skills of male hunters and gardeners, andthe display of quantities of wealth items indicate theoverall wealth of the lineage. Signalers (feast givers) maygain enhanced political power through cementing alli-ances with more patrilineages the more wealth they giveaway and may increase their prestige through uncondi-tional distributions of food. These in turn may increasepurchasing power in the marriage market for unmarriedlineage members as others seek to marry into a patrilineassociated with higher prestige and political power.Guests benefit differently depending upon whether theyare members of the inner or the outer circle. Inner-circlemembers gain reliable information about the lineage’sintent to continue to ally itself with them (guaranteedby the cost of the wealth provided) and whether or nottheir children will make valuable marriage partners (asguaranteed by the quality and extent of both wealth andfood items provided). Members of the outer circle may

3. This is not to say that such signaling is the only function of atombstone opening: every death requires such ceremony in orderto appease the spirit of the deceased. The tombstone is considereda “house” for the deceased, a place where he or she may be kepthappy so as not to wander about among the living causing havoc.A more elaborate tomb is considered to provide greater assurancethat the deceased will remain in one place, where he or she maybe consulted from time to time for intercession on family matters.In fact, delays in construction of the final tomb often necessitatean intermediary feast wherein the family members publicly assurethe deceased’s spirit that they are working on its “house” and thatit will be ready by a certain date.


benefit by gaining information about the current stateof political alliances in the community and may also usethe honest information conveyed by the quality and ex-tent of food provided to assess the hunting, fishing, orcooking skills of a lineage’s marriageable members.

While we view the above account as plausible, we re-iterate that more rigorous formulation of signaling ar-guments and empirical tests thereof are needed. We alsomake no claim that signaling theory is likely to explainall acts of unconditional generosity; in particular, actscarried out anonymously or secretly (such as shelteringJews during the Holocaust) would be hard to explain asforms of signaling, calculated or otherwise.

signaling and “wasteful” subsistenceactivity

Apart from the benefits of hosting a public feast, thereis the distinct question of why individuals choose toproduce goods for display or consumption in this or otherpublic contexts. More specifically, why is it that mendisplay their individual attributes by growing large yamsor hunting turtles, for example? Materialist-oriented re-searchers have been reluctant to recognize that such ac-tivities are actually “wasteful” in a narrow economicsense. Especially in the case of the hunting of large an-imal prey, it is generally assumed that the main benefitaccruing to the hunters is nutritional: either the huntersacquire more meat with less cost than with alternativeprey or they benefit in the long term through conditionalsharing arrangements. Here we discuss two examples ofsubsistence activities that do not appear to fit these as-sumptions and may be better explained using signalingtheory.

Yam growing. Among many Melanesian societies,Dioscorea yams are the focus of men’s gardening effort(Malinowski 1935, Lea 1964, Forge 1966, Kaberry 1971,Weiner 1976, Beckett 1988, Scaglion 1999). While menmay sometimes compete to grow greater quantities ofyams, they often concentrate on growing a few yams thatare as large as possible. At lengths of up to 3 m, dependingon the depth and quality of the soil, such yams are gen-erally woody and inedible, suitable only as propagulesfor more yams and for display.

Among the Central Abelam of Papua New Guinea, thepolitical nature of yam-growing competitions is overt(Forge 1970, Kaberry 1971). The Abelam plant two majortypes: long yam (wabi, D. alata) and hairy yam (ka, D.esculenta/nummularia). D. alata is the only variety ofyam that tends to grow longer with additional care, at-tention, nutrition, and good soil, and wabi gardens arekept separate from ka gardens. A typical wabi garden,tended entirely by men observing taboos on contact withwomen, may contain around 30 long yams, whereas aka garden, tended by both men and women, growsaround 50 small wabi, 600 ka, and 2,000 taro plants (Lea1964). The ka garden produce is destined for householdconsumption, whereas the large wabi in a man’s gardenare destined for display and ceremonial prestige ex-changes. Wabi gardens contribute very little to daily sub-

sistence, even though they require more hours of laborper kilogram of yams produced: Lea provides data show-ing that in one village, planting long wabi yielded 4 lb.of yams per hour of labor while ka provided 37 lb. perhour. On average, men who planted wabi expendednearly twice as much time and effort (460 hours) onplanting wabi as on planting ka (260 hours). A very long(2–4-m) wabi hole required on average 25–35 hours oflabor to construct.

Each Abelam hamlet has a big-man who excels in thegrowing of long yams and knows the magic needed togrow them and who may or may not be its most seniorelder. At harvest, the big-man assumes ownership of allthe yams for which he has performed his magic: “Thereis a close identification between a man and his finestyam: it is a symbol of his manhood and his industry.Many of the longest yams are not eaten: they are dis-played at harvest, stored, distributed, stored again, andeventually planted. . . . When a man dies, some of hisyams are lashed to a mortuary frame by his grave andallowed to rot” (Kaberry 1971:41). Long yams are des-tined for use in ceremonial yam exchanges (Lea 1964,1969) in which the best and longest yams are decoratedwith elaborate painted wooden masks, baskets, shellrings, feathers, and other accouterments and given to theexchange partner of the visiting hamlet. The exchangeis accompanied by a day-long feast, displays of dancing,and “bombastic” oratory. By accepting the yams, the re-cipient obligates himself to match the display or else losestatus, and the largest, finest yams may be rejected bythe recipient, fearing that he may not be able to matchtheir size or quality. Thus, although there is an elementof reciprocity in relations with exchange partners, thefundamental dynamic is one of competitive display, asunderlined by the fact that if the yam is rejected, thegrower leaves the yam on display to rot rather than usingit in another exchange and yet gains immediate prestige(Kaberry 1971:52):

Over a period of years he has established his reputa-tion—one that is acknowledged by members of hishamlet, by the clan of his ceremonial partner, andby the village at large. . . . He is described as onewho has a name: “This name man, man of renown,his wabi and ka are good. When he plants themthey are abundant”; or “he is one who has harvestedbig yams.” He has many garden plots and store-houses; he and his wife or wives produce a surplusof food for lavish distribution at feasts.

Differential success in yam growing is reflected not onlyin the length and weight of the yam but in multiple otherattributes, including shape, surface quality, and hairi-ness. The length of the yam itself is partially a reflectionof the skill of the grower but also may be a product ofthe genetics of the propagule (Richard Scaglion, personalcommunication). Men who are able to obtain pieces ofa very long yam may have a greater chance to producea long yam of their own; thus men seek to cultivatealliances with those who grow the longest yams in an


attempt to gain access to those propagules. It is the menwho possess both skills and wide trade networks whoare able to produce the longest yams. Big-yam men be-come high-status political entrepreneurs, using theirtrade contacts to thwart their rivals and pursue theirpolitical ambitions (Forge 1962, 1970).

Meriam gardening shows a similar pattern. Gardensare distinguished by their purposes: some are planted fordisplay to gain prestige (kies gedub, “common goods gar-den”) and others for provisioning (eroli gedub, “eatinggarden”). Meriam men tend to plant primarily for pres-tige. Men’s gardens feature the same yam species as inAbelam (D. alata). Yams are grown primarily for displayat feasts, for gift giving, or for trade. The labor a manputs into his display garden is substantial: each yamsucker is planted with great care and ritual attention,and each must have a custom-crafted hole as deep aspossible (up to 1 m or more) so that the yam will en-counter minimal resistance to its growth. In hard soil,each such hole takes several hours to dig and anotherhour to pound the soil fine so that the hole can be refilledand appropriately fertilized. Men subsequently devoteconsiderable time and attention to their yams over thenine months they take to reach maximum size.

Like Abelam women, Meriam women plant primarilyfor “table” (household provisioning). They also plant D.alata yams in addition to hairy yams, but their holes areonly 10 cm deep, they may plant 20 for every 1 that aman plants, and theirs reach an optimal eating size infour to six months. Such yams may sometimes be do-nated by women to public feasts as consumables but notas display items. Meriam gardeners speak explicitly ofthe tradeoffs between planting for prestige and plantingfor table. Aside from the fact that taking so much timeto plant a single yam prevents one from planting manymore small yams, large prestige yams are not only toowoody to be eaten but too valuable to be eaten; thus onegardener was cautioned that the prohibition againstwomen in a prestige garden was a practical one, else shewould eat all the food in it before it had gotten big enoughto display. Men also should not have sex with womenwhile planting yams, but, as one gardener put it, “a manmight be able to make up for his lack of women oncehe displayed his yams.”

But why does the ability to grow such large and une-conomical yams confer status?4 Meriam explain that onegains prestige not from a big harvest of table yams butfrom growing and displaying a single huge yam. For gar-deners (and foragers), large harvests are primarily a prod-

4. Boyd and Richerson (1985:269) proposed an answer based on acultural analogue to Fisher’s runaway model in which the abilityto grow larger yams served as an indicator trait of farming skills,leading others to grant these individuals prestige and copy thisindicator trait. They argue that since those who most admire big-yam growers should be more likely to imitate the indicator trait,a positive-feedback runaway dynamic leading to even larger yamsand even more prestige could occur. While this hypothesis has somesimilarities to the costly-signaling explanation proposed here, itdoes not explain why individuals who have no interest in growinglarge yams (e.g., women) would grant prestige or other social ben-efits to big-yam growers.

uct of production time, whereas harvests of rare and bigthings take both additional time and skill. A Meriamman’s foraging and gardening effort is not a good strategyfor putting food on the table but rather a strategy in-volving a display of skill and command of esotericknowledge. The more skilled and knowledgeable a manis, the bigger and better his yams. As a result, he gainsa reputation as a wise man (lug asmer le) or an expert(luzap le), and he demonstrates his store of supernaturalpower, zogo (benevolent) or maid (malevolent). The pro-cess of gaining and displaying knowledge (saserim) is themajor pathway by which men gain social dominance(Beckett 1988). A man who demonstrates his commandof ritual knowledge without divulging the nature of thatknowledge gains much social attention and a measureof influence in public decision-making processes.

In big-yam competitions such as these, the attributesbeing signaled are esoteric or ritual knowledge and/orthe ability to devote time to activities that are not di-rectly productive (but in fact compete with food acqui-sition). The benefit to the signaler is enhanced prestigerelative to other growers; qualitative evidence indicatesthat this enhances the signaler’s social influence. Thebenefit to observers of the yam competition is the ac-quisition of knowledge that will help them to make bet-ter decisions about whom to support in political ma-neuverings. Finally, the guarantee of reliability of thissignal is the dependence of the size/quality of the yamon the time, knowledge, and political connections pos-sessed by the gardener.

Turtle hunting. Meriam turtle hunting has been theobject of a detailed examination of the ability of signalingtheory to explain seemingly inefficient (and costly) for-aging and food distribution patterns (Smith and BliegeBird 2000, Bliege Bird, Smith, and Bird 2001, Smith,Bliege Bird, and Bird 2003). There are two primary typesof marine turtle acquisition on Mer, hunting and col-lecting. Turtle hunting occurs primarily in the contextof public feasting events: hunters choose to hunt in re-sponse to a request from feast organizers to provide tur-tles for consumption at a previously announced feast.Upon accepting such a request, a hunt leader will assem-ble a crew of several “jumpers” (young men who divefor the turtles and wrestle them live into the boat) anda boat driver (who takes commands from the leader) andspend a long day traveling far from Mer in search of free-swimming turtles. In contrast, turtles are collected pri-marily in the context of household provisioning but alsofor feasts by men of all ages, as well as women and chil-dren. Collecting occurs only during the nesting season(October through April), when egg-laying turtles can beharvested on local beaches with minimal effort and risk.

Turtle hunting is a competitive pursuit, with a verydifferent complement of participants from collecting.Compared with collecting, hunting is more costly (intime, energy, and risk), provides meat less efficiently (be-cause of higher travel, search, and pursuit costs), and isassociated with wider distributions of meat (table 1).Hunters keep no meat for themselves unless (quiterarely) hunting for household consumption, in which


table 1Meriam Turtle Hunting and Collecting

VariableTurtle

HuntingTurtle

Collecting

Return rate, before sharing(kcal/forager/hour)

8,110 19,390

Return rate, after sharing(kcal/forager/hour)

�630 3,200

Consumer households perturtle

26.7 18.7

Portion kept by producer (kcal) 7,780 14,390

source: Bliege Bird, Smith, and Bird (2001). Figures are for nest-ing season only, as collecting is not possible in the non-nestingseason; all differences are statistically significant (p ! 0.05) ex-cept portion kept by producer.

table 2Demographic Measures of Meriam Men (Turtle Hunt-ers versus Nonhunters)

VariableTurtle Hunters

(n p 53)Nonhunters

(n p 61)

Age (study sample) 35.2 36.2Age at first reproduction 23.9 25.3Surviving offspring by age 50 4.0 1.7Mean number of mates 0.76 0.46Proportion of “hardworking”

mates25.0 9.9

Age differential with coresi-dent partner (man’s ageminus woman’s age)

3.6 0.1

source: Smith, Bliege Bird, and Bird (2003). All differences arestatistically significant except age.

case they still keep less and share more than turtle col-lectors. Hunters take on a variety of costs for which theyare not materially compensated, including the time andenergy for the hunt (averaging about 40 man-hours),money for fuel, and time organizing and preparing thehunting crew and its equipment prior to the hunt. Theability to bear such costs appears to be linked to hunterquality: because a hunt leader is an organizer and deci-sion maker, his abilities peak as he gains skill and ex-perience. The signals sent by hunting are also efficientlybroadcast: hunts are associated with larger numbers ofconsumers and thus a broader audience overall than col-lections, even during the nesting season and duringhousehold consumption events (Bliege Bird, Smith, andBird 2001).

The costs that hunters incur for providing turtle sogenerously for public consumption do not appear to comein the form of increased shares of collected turtle or otherfoods, as we might predict if risk-reduction reciprocitywere structuring the payoffs for hunting. Those who ac-quire turtle (both hunted and collected) more frequentlyand share more widely or in greater quantity do not re-ceive turtle more frequently than those who share lessor not at all (Bliege Bird and Bird 1997). In addition, suchgenerosity with turtle does not appear to be repaid withshares of fish or other foods (Bliege Bird et al. 2002). Thus,households of turtle hunters generate sustained net out-flows of subsistence resources to nonhunter households.

The costly-signaling explanation of collective goodsprovisioning as applied to the Meriam turtle huntingcase proposes that turtle hunters benefit from uncon-ditional sharing because their harvesting success sendsreliable signals about their quality to the community inwhich they will play out their lives as mates, allies, andcompetitors. Paying attention to such signals can benefitobservers because the costs and potential for completefailure inherent in the signal guarantee that it is a truth-ful measure of the underlying qualities at issue: onlythose endowed with the skills necessary will reliablysucceed and therefore be asked repeatedly to serve oncrews or as hunt leaders. The benefits accruing to sig-nalers (hunters) will depend upon the specific signal and

audience, but for hunt leaders they appear to consist ofbeing deferred to by elders and gaining the benefits of ahard-working wife’s labor (Smith, Bliege Bird, and Bird2003), while for jumpers it is a means of establishingsocial dominance among peers and hence preferentialaccess to various social resources, including enhancedmating success (table 2).

Additional ethnographic cases in which “wasteful”subsistence practices appear to yield signaling advan-tages have been documented. For example, Sosis (2000)has used signaling theory to explain why men on theMicronesian island of Ifaluk engage in torch fishing fordog-toothed tuna even when other fishing patches, par-ticularly yellowfin tuna, are more productive on thosedays. His results suggest (p. 243) that torch fishing is areliable signal of work effort widely observed by mem-bers of the community and that the benefits gained bysignalers are likely related to increasing the status of thematrilineage as a whole. Sosis suggests (p. 293) that sig-naling displays of quality may be increasingly importantwhere ecological constraints limit the amount of honestinformation others can obtain about potential mates, al-lies, and competitors.

artistic elaboration

Artistic expression is a human universal dating back tothe origin of our species. While signaling theory can befruitfully applied to the full range of artistic production—music, dance, oral and written literature, and the plasticarts—we focus on visual artistic or craft traditions. Innearly all human societies, people devote a considerableamount of time and effort to decorating otherwise util-itarian objects (clothing, pots, baskets, tools, dwellings,etc.), as well as their own bodies. They usually also pro-duce nonutilitarian objects used purely for decoration oras wealth objects, as well as aesthetically elaboratedmagico-religious paraphernalia. Anthropologists and bi-ologists have begun to use signaling theory to explainthe adaptive significance of this artistic elaboration (e.g.,


Kohn and Mithen 1999, Miller 1999, Hagen and Bryant2003), though rigorous empirical tests are still lacking.

Bowser (2000) has extensively analyzed ceramics cur-rently produced by a mixed Achuar-Quichua communityin Ecuador. Efforts at fine craftsmanship and aestheticelaboration are concentrated on bowls used for servingvisitors, especially those for chicha (fermented maniocbeer). Achuar and Quichua women “are respected fortheir excellence in making and painting chicha bowls,”and “both men and women say that making chichabowls is one of the most important aspects of a woman’srole” (p. 227). While it is impolite to do so publicly, inprivate potters readily criticize others in terms of thethinness of the bowl walls, the fineness of the painting,and the aesthetics of design elements. For marriedwomen particularly, bowl quality thus indicates the skillof the potter.

A woman who makes finer pots may gain social ben-efits in two ways. First, skilled potters may be able tosecure better marriages—marriages to men who are po-litically well connected. Indeed, at least some level ofability in crafting locally acceptable pots is a prerequisitefor marriage, since “competence in painting potterybowls is a marker of a girl’s transition to marriageableage” (Bowser 2000:227). Second, married women con-tinue to produce pottery that signals both their skill andtheir political alliances. This is because fine bowls areadmired by men and women desiring to impress impor-tant visitors and thus expand their network of politicalalliances. The design of chicha bowls is shaped by thetwin, often competing demands to maintain a style thatindicates group membership and political alliances anddemonstrate enough individuality to gain and maintaina reputation as a skilled potter.

We propose that the audience and the nature of thesignal sent by the potter through the aesthetic qualityof the pottery change over a woman’s life course. Shemay begin with signals of skill: symmetry (challengingin these hand-coiled pots), wall thinness, delicacy of dec-oration, and the like. These may indicate intrinsic qual-ities (fine motor control, cognitive abilities, etc.) or signaltime devoted to developing these skills (itself a measureof time available for investment in nonsubsistence ac-tivities). Ceramic quality is a credible signal of the pot-ter’s ability, since the relevant attributes of symmetryand fineness vary quantitatively as a function of skill.More skilled potters make more attractive marriage part-ners and thus have increased opportunity to belong to astronger political faction and have better alliance part-ners than less skilled potters. After marriage, womencontinue to demonstrate their skill to potential allies inorder to strengthen their existing political alliances andcreate new ones. It is in the best interest of others toally themselves with a skilled potter because she alreadyhas a superior alliance through marriage.

Many other cases show similar features that suggestthe broad applicability of signaling theory to the arts andaesthetic design. Wiessner’s studies of beaded headbandsamong the Kalahari San is a good example. Wiessnerreports that “of 48 women who gave a reason for in-

vesting stylistic effort in beadwork, 42 (88%) mentioneda desire to impress the opposite sex, 37 (77%) to promotereciprocal relations” (1984:204). As in Bowser’s potterystudy, we see here that women, not just men, activelysignal to potential mates and allies and that artistic effortand skill appear to be important means by which indi-viduals attempt to distinguish their qualities from thoseof others.

Signaling theory suggests that stylistic traditions inthe arts exist in part to highlight individual variation inskill, much as conformity to routines in gymnastics orother sports facilitates observer discrimination of skilldifferences between competitors (Zahavi and Zahavi1997). Thus, “decorative styles might converge on com-mon standards in order to facilitate invidious compari-son within the group” and, ironically, conformity mightactually be a result of individuals’ attempting to distin-guish themselves from each other (Boone, personal com-munication).

other domains

Signaling theory has explanatory promise in a variety ofother ethnographic domains that, for lack of space, wecan only briefly mention here.

Religious ritual and commitment. Costly-signaling the-ory is increasingly applied toward an understanding ofreligious ritual (Boone 1998, 2000; Irons 2001; Sosis 2003;Sosis and Alcorta 2003; Sosis and Bressler 2003). Partici-pation in religious ritual can serve as a costly signal ofcommitment to provide a collective good to members ofone’s religious group under circumstances in which in-dividuals seeking short-term gains would be tempted todefect. The costs of ritual observances include time (e.g.,attending services), resources (e.g., financing communalrituals), and, in more intensive forms (handling venomoussnakes, walking on coals, engaging in prolonged fasts, par-ticipating in communal violence, and the like), risks tothe life or well-being of the participant. Signaling theoryprompts the hypothesis that by paying these costs indi-viduals signal to others that they are indeed committedto long-term collective action in a social group. Signal costhelps secure this commitment if a defector would not findit worthwhile to pay it, whereas a committed individualwill recoup it through group membership over the longrun. Alternatively, costly ritual may demonstrate the sig-naler’s ability to “waste” time or resources, as in otherforms of conspicuous expenditure (Boone 1998, 2000).This explanation may best fit cases in which individualschoose to sponsor particularly expensive rituals or reli-gious displays (e.g., the ostentatious churches of Renais-sance Italy).

Monumental architecture. One of the first anthropo-logical applications of costly-signaling theory analyzedmonumental architecture in Mesoamerica as a costly ad-vertisement of political quality by elites attempting tointimidate their rivals (Neiman 1997). Neiman hypoth-esized that Classic Mayan monument size and com-plexity should scale with the intensity of payoffs to ad-vertisement and therefore factors such as recency of


migration into an area (which predicts less knowledgeof relative competitive ability of neighboring polities)should spur intensified monument construction. Hisanalysis suggests that variability in signaling intensityis related to ecological conditions that spur migration,concentrate audiences, and enhance differences in qual-ity between individuals, here primarily elites controllingresource catchment areas (see Boone 2000 for a similarargument applied to Kwakiutl potlatching). The spatialscale at which cycles of monument construction andabandonment occur then provides an estimate of thescale of political competition, and the figure Neimanderives from the costly-signaling analysis matches theone that epigraphers have derived independently fromemblem glyphs. In general, given its obvious utility inreliably advertising wealth, control over labor, and hencemilitary and economic power (Trigger 1990, Kohler1998), monumental architecture seems likely to providefertile ground for application of signaling theory (for acounterargument, see Aranyosi 1999).

Embodied handicaps. Finally, there are plausiblegrounds for expecting signaling theory to illuminatecostly and often seemingly irrational forms of bodilymodification, ranging from tattoos and scarification togenital cutting (male and female) and footbinding. Somepreliminary research has explored the hypothesis thatbloodletting and intensive forms of body modificationmay serve to test nutritional status or immunocompe-tence in environments with high rates of pathogens (Lud-vico and Kurland 1995, Neiman 1997, Singh and Bron-stad 1997). Alternative hypotheses exist for thesepractices, of course, ranging from providing irreversiblemarkers of ethnic affiliation to heightening the emo-tional impact of initiation rites, and we do not neces-sarily expect signaling explanations to be the sole or evenpredominant explanation for ritualized bodily modifi-cation. Here we simply argue that some of these practicesdo plausibly fit the criteria expected for costly signalingas strikingly embodied handicaps that signal socially rel-evant aspects of the bearer’s quality. In some instances,such as Chinese footbinding, historical evidence indi-cates that the practice began as a way for high-statusindividuals (in this case, elite concubines) to differentiatethemselves from social competitors as women of leisure.It then spread via competitive emulation to all but thepoorest peasant classes, who could not afford the reduc-tion in female mobility (and thus field labor) it entailed(Stevan Harrell, personal communication; Gates 2001).As in other domains, much further research is needed tosubject these plausible scenarios to rigorous empiricaltest.

Broader Implications

Having surveyed a number of heuristic ethnographic ex-amples in which signaling theory appears to have poten-tial explanatory value, we now turn to a set of broaderissues raised by this exercise. First, we discuss some is-sues raised when one applies a theory such as costly

signaling to cultural phenomena and to processes thatinvolve individual agency as well as cultural or geneticinheritance. Second, we briefly consider what factorsmight shape variation in both the form of signaling dis-plays and the intensity of signaling that might occur inany particular case. We then consider the contributionsof signaling theory to an understanding of prestige-basedhierarchy, group-level signaling, and gender differencesin signaling. Finally, we discuss some common miscon-ceptions about costly-signaling theory.

culture and signal evolution

The first set of issues is in part a question of causalmechanisms and should be familiar to all evolutionaryanthropologists. What does it mean to say that the sametheory can explain genetically evolved morphologicalcharacters such as peacocks’ tails or infant body fat (Hrdy1999), contingent behavior such as gazelle stotting orhosting a feast, and ritualized behavior such as the court-ship behavior of a waterfowl or the display of a giantyam? These putative examples come equally from hu-man and nonhuman realms; the issue of “genetic” versus“nongenetic” signal origins does not demarcate human-ity from other species. Nevertheless, even a casual sur-vey suggests that the displays produced by the humananimal arise from individual agency and cultural inher-itance to a vastly greater degree than is the case for otherspecies. What is the explanatory significance of this dif-ference?

One issue concerns the role of inheritance (genetic orcultural) versus contingent behavior. When a peacockmatures and grows an extravagant tail, there would notappear to be much room for individual strategizing. Forhumans, however, it appears that strategizing, context-dependent behavior, and other forms of contingent sig-naling are common. We find it remarkable that the rulesthat govern the (genetic) evolution of signals such aspeacocks’ tails and gazelle stotting seem to have suchpromise in explaining human signaling behavior whichvaries independently of genetic change. We propose thatthis convergence is due to one or both of the followingforces: (1) the role of genetically evolved preferences orpredispositions in guiding human decision making, asargued by a variety of theorists (e.g., Irons 1979, Durham1990, Tooby and Cosmides 1992), and (2) adaptive cul-tural evolution (Boyd and Richerson 1985, 1995; Durham1990). In other words, signaling systems that are adaptive(fitness-enhancing) may arise either through the actionof evolved psychological mechanisms or via a process ofcultural evolution. There is, of course, no guarantee thateither process will always produce adaptation, but thereare ample reasons (discussed in the cited literature andelsewhere) for supposing that often or even usually thiswill be the case (though this must remain a primarilyempirical question).

The relation of cultural evolution to other forms ofbehavioral adaptation is the subject of an ongoing debate(recently reviewed by Laland and Brown 2002). We canoffer two brief contributions to this debate here. First,


we expect the very nature of signaling to entail a largerrole for cultural evolutionary dynamics than might bethe case in some other domains of human behavior. Sincesignaling is inherently social, requiring coordination ofsenders and receivers, any one individual has relativelylimited ability to impose novel signals. This set of con-ditions may encourage the cultural evolution of signalsbased on conventional meanings that are relatively in-dependent of genetically evolved constraints. If we mayparaphrase Marx, while people creatively fashion theirown signals, they do not do so with materials of theirown free choosing.

Second, while we have no doubt that culturaltraditions shape signal evolution, people seem to be ableto innovate new signals rapidly, without too much cul-tural inertia. Some of this innovation seems clearly tiedto changes in costs (and hence signal value), as in therapid demise of lace making as a signal of high statuswith the advent of machine-stitched lace (Zahavi andZahavi 1997) or the reversal of the status value of sun-tanned skin following the shift from agrarian to factory/office labor in Europe and North America. Other inno-vations seem to arise and spread for relatively arbitraryreasons, as seems to be true for much of fashion (be ittribal or postmodern), but further research is needed todetermine what relation to signal cost, condition-depen-dent quality, and observer benefit such seemingly arbi-trary shifts may involve. In sum, the cultural evolutionor epidemiology of signals is an undeveloped area of em-pirical research and promises to be a complex and fruitfultopic.

Ultimately, the evolution of signals is a matter of thedifferential replication of information (cultural and ge-netic), as determined by the effect of this information inthe material world of living things and their environ-ments. That humans have added cultural information tothe genetic information replicated in the lineages ofother species is a strange and wonderful thing. That sig-naling theory can illuminate the evolution of signals inboth realms is both surprising and exciting.

cross-cultural variation

What determines cross-cultural variation in the arenasin which signaling is important? What determines cross-cultural variation in the amount of signaling in any onearena? There is no doubt that investment in costly dis-play varies cross-culturally, in intensity as well as inform. The received wisdom is that richer environmentscreate less pressure for strict economic utility maximi-zation, allowing more time for symbolic or “wasteful”pursuits such as art, ceremony, and pursuit of prestige.Yet in one of the harshest environments inhabited byhumans, the Great Sandy Desert of Western Australia,hunter-gatherers invest huge amounts of time and energyin seemingly “wasteful” religious ritual (Tonkinson1991). While such ritual no doubt has a practical aspect—that of facilitating spatial memory of important watersources, foraging patches, tool sites, and the paths link-ing these economically important sites—the sheer elab-

oration of initiation ritual, including bodily mutilationsand rigorous physical and social tests, suggests that thesymbolic display value is equally if not more important.Likewise, it would seem that if hunting large, evasiveanimals is a display of skill, environments lacking suchanimals should promote a more narrowly efficient hunt-ing strategy. Yet even here, where kangaroo and emu areextremely rare, men still prefer to hunt them rather thanthe more plentiful lizards and small mammals (BliegeBird and Bird n.d.).

Rather than looking for broad correlates between in-tensity and form of signaling and some ecological param-eter, it may be more fruitful to examine how the socialenvironment creates or constrains competition in spe-cific social domains. Signaling that is primarily aboutattracting mates should covary with the social and eco-logical factors shaping mating opportunities: where ei-ther serial monogamy or polygyny is common, signaling(by both males and females) should extend over a longerportion of the reproductive life span than where prohi-bitions on divorce or polygyny constrain opportunitiesfor remarriage. When endogamy rates in a local com-munity are high, information about the qualities of pro-spective mates should be relatively abundant and sig-naling therefore less intensive than when mates comefrom other communities. These and similar hypothesesappear eminently testable.

Signaling that primarily concerns competition for po-litical influence should similarly covary with the op-portunity for mobility within the local or regional so-ciopolitical system (greater mobility leading to moreintensive signaling). One of the more important factorshere may be the benefits and constraints involving thecreation and perpetuation of prestige hierarchies.

a signaling theory of prestige

Boone (1998) was one of the first to propose that signalingtheory could explain why costly displays establish orreinforce social status. By using conventional signalssuch as ceremonial sponsorship or magnanimous behav-ior to advertise their resource-holding potential, signal-ers can resolve conflicts before they rise to costlier levelsinvolving direct conflict. Boone and Kessler (1999; Boone2000) further refined the argument to show that the long-term benefits of achieving high status may be realizedby lineages that accumulate this status and exploit theprestige deference they achieve from it to gain priorityof access to resources during periodic downturns in theeconomy (e.g., due to drought).

Prestige is sought through competition, but not justany type of competition will do. Prestige competitionoften seems to focus on activities with unambiguouslyranked outcomes that depend upon participant skill orother hidden qualities. Men gain prestige and other ad-vantages through killing more rivals than other men (Go-delier 1986), growing more (Weiner 1988) or bigger (Ka-berry 1971) yams, giving away more food or goods atfeasts or engaging in ostentatious displays of wealth(Kroeber and Gifford 1949, Codere 1950), gaining more


powerful ritual knowledge (Godelier 1986, Stanner 1966),or hunting more successfully or productively (Kaplan andHill 1985a, Godelier 1986, Altman 1987, Wiessner 1996,Bliege Bird, Smith, and Bird 2001, Henrich and Gil-White2001). Competitive efforts of these sorts contribute tothe prestige quest because they provide effective signal-ing media, that is, they allow reliable communicationregarding differences in intrinsic quality between com-petitors. In each case, certain characteristics of the pres-tige-generating activity affect relative success so as toreveal the competitors’ levels of intrinsic skill or othersocially valued attributes. Activities with high “signal-ing potential”—those suited to the broadcasting of reli-able signals that benefit both observer and signaler—should be preferred. Such activities permit observers todiscriminate skill levels among competitors and drawthe attention of competitors (and potential mates orallies).

These “games of skill” provide arenas for honest sig-naling that “games of chance” lack. The outcome of atop-spinning game played by Meriam men depends im-portantly on the top-making abilities of each competitor.Men make their own tops from round bean pods or grinda sandstone top from local stone (Haddon 1935). Thesignal sent by Meriam top spinners concerns their abilityto construct a perfectly balanced top, a feat that requirescognitive and manufacturing skills to transform an ir-regularly shaped object into a perfectly symmetrical oneand then locate its exact center for placing the spindle.It may also (or alternatively) signal the spinner’s abilityto generate sufficient surplus as to devote many hoursto the construction of and practice with stone tops. Suchskill or surplus is not revealed if top spinners merelycompete to spin manufactured tops, for there is then nointrinsic link between the length of the spin and thestone-working ability of the spinner. Likewise, top spin-ners maximize the information-carrying capacity of thesignal by eliminating as many randomly generated ex-ternalities as possible: competitors create wind-hoods tokeep a sudden gust from interfering with their top’s spin,use a broken piece of glass as a spinning surface to re-move any differences due to friction, and damp downvibrations from overly enthusiastic onlookers with a bedof sand.

Competition over food production is the most ubiq-uitous form of status competition, occurring amongthose who hunt and gather as well as those who garden.Of the many characteristics of food production thatmight provide reliable measures of producer quality, wepropose two highly observable indices: failure rate andharvest size. These can be reliable signals if they (1) aresignificantly linked to intrinsic qualities of the producerrelative to extrinsic producer-independent forces or sim-ply producer effort and (2) vary significantly among pro-ducers in such a way that observers can reliably distin-guish those who average fewer production failures orlarger harvests. In cases in which failure-rate or harvest-size differences among producers show little variation,observers may be able to discriminate only broad skilllevels, and prestige-based hierarchies will tend to be shal-

low and relatively egalitarian. If between-individual dif-ferences in failure rate or harvest size are great, observersmay be able to rank producers ordinally. If skill levels(and associated production outcomes) remain consistentover time, we might expect strong, stable prestige hi-erarchies to develop, with proportionally less investmentin signaling.

Our argument suggests that the evolution of hierarchyis constrained by the nature of the production system,specifically the relation between productive resourcesand producers’ skill differentials. The fairly egalitarianpolitical organization observed in most hunter-gatherersocieties may thus be due in part to the very shallowand inconsistent or noisy (due to extrinsic factors) dif-ferences in foraging success between individuals(Hawkes 2000). When resources become more predicta-ble in space and time, making differences in harvestmethods and in resource defense prime determinants ofvariation in productive output, stable prestige differencesshould appear. As status competition intensifies in suchsystems, competition may shift to control over labor andexchange goods, and signaling games may be elaboratedto display differences in these attributes.

A clear example of these trends occurred in NorthwestCoast Indian societies (Ames 2001, Bishop 1987, Donald1997). Here, each house group was headed by a chief,who (along with his close relatives, members of the titlednobility) claimed inherited and supernaturally sanc-tioned rights to prime fishing and gathering sites (par-ticularly salmon streams) and directed the harvest anddisposition of these resources. Northwest Coast chiefsand other nobles generated wealth and associated powerand prestige in three main ways: by controlling the pro-duction of the commoners and of slaves (the latter ownedby nobles) residing in their house group through near-monopoly of trade with other groups and by supportingcraftsmen to produce wealth objects and ceremonialitems (weavings, carvings, totem poles, dance costumes,etc.). But the Northwest Coast nobility had very limitedcoercive power over commoners and craft specialists andindeed had to compete with other house groups to attractthe most industrious and skilled of these. This was doneby signaling house productivity and resource holdingsvia bestowal of wealth and food gifts in often vast quan-tities, typically as part of marital alliances or at the pot-latches held to mark major rites of passage of nobles(such as births, deaths, and marriages). While one couldnot generally become a chief or other title-holder simplyby engaging in such wealth displays, one had to engagein them to validate noble status. Those who had the rightmix of productive fishing sites and trade networks, suf-ficient labor, and political acumen could produce morefrequent and elaborate displays and thereby attract evenmore industrious commoners, warriors to capture slavesand booty, and strategic allies, thereby increasing theirregional standing in a positive feedback cycle. Althoughslaves had no say in whom they belonged to, most of theother participants had some choice in where they livedand whom they supported, and signaling via wealth dis-


plays and transfers played a crucial role in shaping suchchoices.

groups and individuals

Signaling theory in the social realm has tended to focuson the symbolic capital accruing to groups as a result ofsignaling by members, while theory in evolutionary ecol-ogy tends to focus on the benefits to individuals. Thereis no reason to suspect that one or the other must takeprecedence: individuals may often signal competitivelyin order to gain benefits for themselves and may alsocooperate with other members of a coalition or socialgroup to signal group attributes to other such groups. Inmany cases, the two levels of signaling may be inter-twined (as in the Meriam feasting complex describedabove, for example). Although group-level signaling is inprinciple amenable to the same sorts of analyses as in-dividual signaling, it raises several important issues.

Collective action problems. To the extent that mem-bers of a group have different interests, there will beconflict over who will pay the cost of group-level sig-naling. In the case of feasting or some other signal ofcoalition strength or cohesion, the signal intensity willbe maximized if all members of the lineage contributeto the production of the signal; but if the signal will beproduced even if some fail to contribute, there will betemptation to free-ride on the signaling efforts of others.Undoubtedly, if many individuals choose not to contrib-ute to the signal, the feast will be noticeably poorer andthe status of the lineage will suffer. This collective actionproblem (Olson 1965) must be solved to avoid failure.One solution would be a system of monitoring and pun-ishment of slackers (though this itself raises a second-order collective action problem of who will carry thisout). Alternatively, if there is heterogeneity among groupmembers in the costs and benefits of signaling, thosewho have most to gain from signaling group quality willhave greater incentives to pay the costs of signal pro-duction and/or enforcement. In any case, groups that cansolve collective action problems and motivate membersto contribute at high rates will send stronger signals andgain competitive advantages over other groups.

Piggybacking. Allowing individuals to gain additionalselfish benefits from contributing to the group-signalingeffort will ameliorate any collective action problems thatarise. In this case, individuals have clear incentives tocontribute to the collective display because they gainindividually by signaling their own hidden attributes (asin the case of Meriam turtle hunters discussed above).Allowing individuals to showcase their skills will ac-complish this as well; thus it is quite understandablethat feasts and other ceremonial events are often accom-panied by dancing and other competitive displays or thatartisans are motivated to display their skills in such con-texts (e.g., carved masks in Northwest Coast Indianceremonies).

Collective benefits. The greatly expanded scale and“redistributive” nature of much group-level signaling(particularly feasting) has important social implications.

If through collective displays a group demonstrates to itsown members that it can overcome any collective actionproblems associated with signal production, it may bemore likely to be able to solve such problems in otherarenas, such as warfare and defense, thereby deterringpotential competitors. Furthermore, signals in the formof public goods sent by large numbers of cooperatingindividuals do end up providing large material benefitsto recipients. For example, frequent feasts by competinglineages spur production surpluses, which tend to feedmore people, a perhaps subsidiary but beneficial out-come. The complex interplay between individual andcollective benefits to signaling is discussed further bySmith and Bliege Bird (2000, n.d.) and Gintis et al. (2001).

gender and signaling games

In our previous publications applying signaling theory tothe explanation of certain subsistence activities, we haveconcluded that in most of these cases subsistence-basedsignaling appears to be a male specialty. In many soci-eties, women’s ability to enter these signaling contestsis limited by their own opportunity costs and theirchoices about whether to participate; in others, men im-pose rules that limit female participation. However, itwould be a major mistake to conclude that women donot compete in signaling games. While women (and fe-male animals in general) have been portrayed as rela-tively passive, selfless mothers, more modern views havetended to show the opposite (Hrdy 1999). Females com-pete, often directly, with other females not only for ac-cess to resources needed to support their families butalso for high-quality mates and allies.

Yet there is also no doubt that women often competevery differently from men. One striking example con-cerns the Mendi, residents of the central Highlands ofNew Guinea. Both men and women build up networksof exchange partners, twem (Lederman 1990). Women’sand men’s twem relationships are similar in form butdiffer in scope, intensity, and goals. Men tend to havemore partners at one time and more exchange debts andcredits and to spend more time in exchange-related ac-tivities. Women conduct transactions mainly for mar-riage and funeral feasts sponsored by individuals, an ac-tivity that involves giving away smaller portions to moreindividuals. Men focus on acquiring large amounts ofwealth for periodic display in clan-sponsored ceremoniessuch as warfare compensations that contribute towardtheir status as “big-men.” While some may interpretsuch differences as evidence of male coercion, others seethem as expressing differences between men and womenin goals or interests. Exploring such issues may be oneof the most significant arenas for future ethnographicapplication of signaling theory.

misunderstandings of costly signaling

Costly-signaling theory has provided a very robust andelegant framework for analyzing the problem of reliablecommunication when signalers and recipients lack per-


fectly coincident interests. Since partially conflicting in-terests can be predicted in a large variety of interactions(even including parents and offspring, mates, and allies),the costly-signaling solution is potentially of wide ap-plicability. However, it is easy to interpret the theory inan overly simplistic or misleading fashion.

Cost is not enough. Since virtually all behavior in-volves some cost, signal cost itself cannot be diagnosticof a costly-signaling dynamic or adaptation. At mini-mum, a costly-signaling analysis must show that signalcost (or benefit) is quality-dependent as defined above.Even quality-dependent cost does not guarantee honesty;for example, only wealthy suitors can afford diamondengagement rings, but this does not guarantee that suchsuitors will follow through with marriage, only that theyare willing to pay a significant cost to keep that optionopen.

Signals need not impose handicaps. Despite the ter-minology utilized by many proponents of costly-signal-ing theory (e.g., Zahavi and Zahavi 1997), signal costneed not be so great as to impose a “handicap” upon thesignaler; in fact, the logic of costly-signaling theory pre-dicts that a signal that would handicap a low-qualitysignaler may be cheap for a high-quality signaler (Getty1998). The theory requires that one distinguish betweenthe observable or objective cost of a signal (e.g., themoney a suitor spends on an engagement ring, the energya peacock expends in dragging around a large tail) andthe fitness or utility deficit the signal imposes on thesignaler. If cost in the latter sense is quality-dependent,then the objective cost (signal intensity) will be a goodindex of ability to pay and hence of some underlyingdimension or quality of interest to recipients.

Costs have different sources. Confusion about signalcost can also arise through failure to distinguish differenttypes of cost. Some signals are intrinsically costly toproduce (e.g., stotting, monumental architecture) or tomaintain (e.g., a peacock’s tail, familiarity with the latestcinema or musical styles). Other signals are cheap toproduce but entail costs through their potential conse-quences—especially social consequences (e.g., “statusbadges” in the plumage of certain bird species or boastingabout one’s fighting ability in a bar full of rough char-acters). It is easy to see how the signal production costsfit into the costly-signaling framework; if such costs arequality-dependent (higher-quality individuals pay lowermarginal costs), then the conditions for honest signalingare at least partially met. In the second case, when signalproduction is cheap but has potentially high conse-quence costs, it becomes somewhat harder to distinguishcostly signaling from other means of ensuring honestcommunication. But the quality-dependence require-ment remains the same: if higher-quality individuals facelower consequence costs for a given level of signal in-tensity, then honest signals of quality may be favored.This sort of dynamic is relevant to the problem of howhuman linguistic communication can be kept relativelyhonest: those who exaggerate their abilities or accom-plishments (signal a higher quality than they actuallypossess) will be punished (pay a disproportionate con-

sequence cost) if their exaggerations are discovered(Lachmann, Szamado, and Bergstrom 2001).

Honesty need not be costly. If signaler and recipienthave sufficiently coincident interests because of high re-latedness or other factors, honest communication can bestable without signal cost (Bergstrom and Lachmann1998, but cf. Brilot and Johnstone 2003). In addition, an-atomical constraints may guarantee accuracy in somecases (Maynard Smith and Harper 1995). For example, inthe competitive roaring of red deer stags, larger and heav-ier deer have lower-frequency roars, and this is an un-fakeable (but not costly) signal (Reby and McComb 2003).

Costs can dwindle at equilibrium. Even under acostly-signaling dynamic, cost may be very low or evenabsent at equilibrium (Lachmann, Szamado, and Berg-strom 2001). For example, suppose that the relevant qual-ity is dichotomous (i.e., there are only two types of sig-nalers, high-quality and low-quality) and the equilibriumis such that only high-quality individuals signal (a stan-dard result in dichotomous-quality signaling models).Such an equilibrium can occur even if signal cost (forhigh-quality individuals) is zero or negative—that is,even if the signaling behavior is intrinsically beneficialto signalers, not including the benefits obtained whenreceivers respond (Gintis et al. 2001). This is not as re-mote a possibility as it might seem. For example, supposethat some individuals are superior foragers and can ef-ficiently pursue and harvest certain prey types thatwould be very inefficient for low-quality foragers to ob-tain. In that case, high-quality foragers may benefit bothfrom the foraging returns and from the signal value ofharvesting such prey (e.g., they may be preferred as matesbecause their success in capturing the prey in questionserves as a reliable signal of more general foraging abil-ities). There will be no signaling cost at equilibrium,even though this equilibrium is maintained by cost func-tions (i.e., the cost of taking these prey would be so highfor low-quality foragers that even the addition of receiverbenefits would leave them with a net deficit in fitness).A similar result can apply to continuous distributions ofquality and signaling intensity if costs are due to signalconsequence rather than production (Lachmann, Sza-mado, and Bergstrom 2001). In both cases, signaling equi-libria are stabilized by the costs that nonoptimal behav-ior would entail (e.g., signaling more intensively thanone’s quality would warrant) rather than by the costs ofobserved (equilibrium) signaling. In an insightful dis-cussion of these various issues, Lachmann et al. (p.13193) conclude that low- or zero-cost signals—so-calledconventional signals—can be expected when signalersand recipients have sufficiently coincident interests orsignal claims are independently verifiable. Conversely,costly signals are predicted (1) in one-time interactions,(2) when communicating about otherwise unverifiableproperties, or (3) when the expected gains from deceptionexceed the expected costs from sanctions or punish-ments imposed by deceived observers.

Signaling frequency may not decline with familiarity.One might expect that once a set of individuals had es-tablished differences in quality among its members, sig-


table 3Applications of Signaling Theory to Ethnographic Examples

Signal TypeUnderlying

QualityBenefit

to SignalerBenefit

to RecipientCondition-

Dependence

Redistributive feast Alliance (kin group)size and cohesion

Enhanced politicalpower and mar-riage-market status

Reliable informationon whom to allyoneself with, marry,or defer to

Amount of feastfood provided di-rectly dependenton size and indus-triousness of hostgroup

Big-yam display Esoteric knowl-edge, surplus la-bor for nonpro-ductive work

Increased prestige,political power,mating success

Reliable informationon whom to allyoneself with, marry,or defer to

Yam size dependenton skill, knowl-edge, and laborinvestment

Hunting difficult prey Physical vigor, cog-nitive ability, for-aging ability

Deter competitorsand attract potentialmates

Avoidance of costlycompetition withbetter competitor,choice of highest-quality mate

Failure rate andmean return ratesdependent uponskill and cognitiveability

Artistic elaboration Cognitive and mo-tor skills, timeavailable for non-subsistence labor

Attract better matesand allies

Reliable informationon whom tomarry or allyoneself with

Quality of artisticproduction di-rectly dependenton skills and/ortime investment

naling would cease. More precisely, the more informa-tion observers have on signaler quality, the lower themarginal payoff from signaling, and hence the lower theequilibrium amount of signaler investment in signalingwe should observe (Neiman l997, Lachmann, Szamado,and Bergstrom 2001). Yet further thought suggests sev-eral reasons that individuals might continue to gain ben-efits from signaling and observers from attending to sig-nals even if the audience is quite familiar with thesignaler. First, when signal meaning is noisy because ofvariability unrelated to the relevant intrinsic quality ofthe individual, signalers may need to signal repeatedlyin order to ensure that receivers can interpret the signal(Boone 1998). Secondly, repeated signaling is necessaryif past and present quality are poorly correlated; for ex-ample, signals of health and vigor may be subject to rapidshort-term changes (e.g., due to infectious disease), andindividuals or coalitions may need to advertise their cur-rent economic condition if fortunes rise and fall or mem-bership changes over time (Boone 2000). Finally, whensignalers are competing with others in their social groupregarding their commitment to delivering collectivegoods in order to retain political privileges, they mayneed to continue signaling even if others are quite fa-miliar with the competitors. To cease signaling wouldin effect signal inability or unwillingness to continue inthe role of political patron, thereby yielding the perks ofpolitical leadership to competitors. This sort of dynamicmay characterize situations ranging from big-man sys-tems to potlatching by rival chiefs to “pork-barrel” pol-itics in representative democracies.

Costly-signaling theory is a complex and subtle frame-work that applies only to certain situations in whichspecified conditions hold. To employ it properly, wemust carefully consider types and sources of costs, theirrelation to signaling equilibria versus disequilibria, al-

ternative means of ensuring credibility, and the condi-tions governing variation in signal meaning over time.

Conclusions and Prospects

Our primary aim in this paper is to examine the potentialexplanatory value of signaling theory for a variety of an-thropological topics. We want to reiterate that applica-tion of this theory to human social behavior is at an earlystage and judgment of its value must await rigorous em-pirical evaluation. Our discussion of ethnographic phe-nomena was designed to illustrate what signaling expla-nations might have to offer; we did not attempt to subjectthese explanations to adequate tests—indeed, the req-uisite evidence to do so is generally lacking (althoughsome exceptions, including Meriam turtle hunting, arediscussed above). In addition, even if signaling theoryproves as powerful an explanatory framework as we sug-gest it may, it will elucidate only one aspect of whyhuman behavior and social institutions (even the onesdiscussed in this paper) take the forms they do.

The available evidence suggests that signaling theorydoes have considerable promise for generating novel andpowerful insights into the ethnographic domain. For avariety of phenomena ranging from subsistence produc-tion to feasting to artistic elaboration, the fundamentalconditions for honest signaling of condition-dependentqualities appear to be met (table 3). The underlying qual-ities being signaled range from physical vigor to cogni-tive skills to coalition size and cohesion. While the pre-cise benefits to signalers and receivers have not beenmeasured in most cases, the leading contenders includeobtaining better mates, forming valuable alliances, andavoiding the costs of violent competition.

We have argued here that signaling theory offers a re-


markable opportunity to bridge semiotic and materialistframeworks by integrating an interactive theory of sym-bolic communication and social benefit with materialisttheories of individual strategic action and adaptation.One of the classic lessons of anthropology has been thateconomic activity is not just about making a living butabout communicating values and expressing meaning.This insight formalizes what many different scholars (be-ginning with Veblen) have realized: signaling theory of-fers a deep understanding of why material goods and thetime and resource costs they represent are such goodmedia for communication. As Rappaport (1979:181) haswritten,

When a Goodenough Islander (Young 1971) or aSiuai (Oliver 1955) transmits the message that he isa man of importance, influence, or prestige by givingaway large numbers of yams and substantial num-bers of pigs he is not simply claiming to be a bigman. He is displaying the fact that he is. Theamount that he gives away is an index of his “big-ness” because it is “really affected by” that which itsignifies—his influence, prestige, authority . . .

But signaling theory also proposes that communica-tion is aimed at securing social benefits with real ma-terial consequences and that these consequences are cru-cial to an understanding of the form, dynamics, and fateof communicative and symbolic behavior. This propo-sition is thus far mostly untested, but the opportunitiesfor testing and refining signaling-based explanations inethnographic contexts are rich. We hope that the sketchwe have provided will generate further interest and sub-stantive research.

Comments

michael alvardDepartment of Anthropology, Texas A&M University,4352 TAMU, College Station, TX 77843-4352, U.S.A.([email protected]). 9 xi 04

One has to be encouraged by an idea that can incorporatethe thoughts of Bourdieu, Veblen, and Darwin into a the-oretical foundation that has promise for explaining bothpeacock tails and potlatches. To appreciate the contri-bution of the ideas presented by Bliege Bird and Smithone has to be attentive to the fact that signaling theorywas developed, in part, to help understand how honestcommunication between organisms could evolve (Daw-kins and Krebs 1978). Signaling theory is a theory ofcommunication, and in the context of humans it is fun-damentally a theory of culture. As a practitioner of be-havioral ecology, I have had nagging misgivings aboutits apparent lack of tools for adequately incorporatingculture into its purview (Alvard 2003). I am encouragedof late, however. Bliege Bird and Smith correctly tell usthat signaling theory has the power to explain, within

the larger umbrella of evolutionary theory, much sym-bolic and ritual behavior that has otherwise been putdown by many anthropologists to the capriciousness ofculture. It therefore has the potential to provide the the-oretical tools for linking traditional cultural anthropol-ogy, behavioral ecology, and evolutionary approaches toculture.

One advantage of cultural learning is that it providesbenefits in contexts in which individual learning is error-prone or costly in time and/or effort (Boyd and Richerson1995). A forager deciding whether to hunt or fish mayfind it best in some situations simply to copy his neigh-bor’s strategy rather than spend time and effort in costlytrial-and-error learning. Cultural organisms can engagein individual learning if costs are low and success likely;otherwise, they can imitate others. This process parallelsan important aspect of signaling as presented in the tar-get article: “costly-signaling theory involves the com-munication of attributes that are relatively difficult orexpensive to perceive directly.” It seems to me that oneway to look at signaling is as essentially a cultural pro-cess in which the hidden attributes are social ones.

A useful way to think about signaling and culture to-gether is in terms of the economic concept of transactioncost. For example, finding a suitable spouse involvestime and effort, especially if it is difficult to discern thequalities of the various choices. In addition, it is difficultto learn from direct personal experience because choos-ing a mate occurs too infrequently during one person’slifetime to produce a sufficiently large sample (Thaler1992). The time and effort spent in acquiring adequateinformation to make an informed choice are transactioncosts that can be mediated by attention to signals thatcorrelate with the signaler’s quality of interest. Onecould interpret Bliege Bird and Smith as arguing thatsignals allow social partners to reduce the transactioncosts of political, sexual, or economic exchange.

Exchange is the most basic form of cooperation, and,in fact, signaling theory rests on the assumption thatthere are mutual benefits to be gained from truthful com-munication. In these mutualistic contests, signals allowpeople to solve what game theoreticians call coordina-tion or assurance games (Binmore 1994). One can con-ceive of coordination as an exchange in which it paysfor players to cooperate (i.e., there are mutual benefitsto be gained, in contrast to a prisoner’s dilemma, inwhich it always pays to defect). The key for the partic-ipants in such an exchange is to increase the likelihood,or assurance, that their partners will “play by the samerules” (i.e., follow the same postmarriage residence rules,drive on the same side of the road, butcher and distributeprey in the same way [Alvard and Nolin 2002]).

But how do we know whom to trust? With whom dowe share these arbitrary details of social life? These ques-tions concern identifying attributes of others that arerelatively difficult or expensive to perceive directly.Among other advantages, the cultural mechanisms in-volved in signaling provide people the ability to infereach other’s mental states, to assort preferentially withothers who have similar (or complementary) intentions


or capabilities, and to reap the advantages of coordinatedactivities (Tomasello 1999). It is to the understanding ofthis process that signaling theory has the opportunity tomake its greatest contribution.

michael chibnikDepartment of Anthropology, University of Iowa,Iowa City, IA 52242, U.S.A. ([email protected]). 7 xi 04

This lucid and ambitious paper persuasively argues thatsignaling theory enables us to understand disparate phe-nomena in which individuals or groups make great ef-forts to provide information to others about underlyingattributes that are difficult or impossible to observe di-rectly. The authors carefully specify the conditions underwhich such costly signaling is likely to occur. They haveconvinced me that a signaling framework helps explaincertain otherwise puzzling aspects of human behavior.

The paper could have been more explicit about theextent to which the emitters and interpreters of costlysignals understand their actions. Such signals have oftenbeen analyzed from the perspective of semiotics (e.g.,Peirce 1932–58, Sebeok 2001); the assumption here isthat senders and receivers know what is being com-municated. The authors, however, do not emphasize thepresupposed cultural knowledge entailed in signaling.Their focus instead is on functional analyses in whichsignaling is seen as a way for individuals and groups toacquire prestige, power, and mates. Relatively little at-tention is given to the cultural evolution of signaling.The authors tentatively say that “design forces” such asdecision making, subconscious learning, natural selec-tion, and adaptive cultural transmission should favor “asystem of communication conforming to the costly-sig-naling framework.” This formulation leaves open thequestion whether the users of such systems of com-munication are aware of their alleged functions.

As Bliege Bird and Smith acknowledge, many of their“signaling” explanations of particular cultural phenom-ena are not new. They derive inspiration from the workof Veblen and Bourdieu on conspicuous consumption andsymbolic capital and cite diverse other writers taking asignaling approach to topics such as the potlatch, mon-umental architecture, and Melanesian feasts. One couldadd related (although perhaps less directly relevant) anal-yses of multivalent ritual symbols (Turner 1967) and hi-erarchies of responses to environmental perturbations(Vayda and McCay 1975). Nonetheless, a number of theirsignaling-oriented explanations are new to me. Some areconvincing (e.g., those concerning seemingly wastefulsubsistence activities); others (e.g., those related to ar-tistic elaboration) seem less compelling.

The underlying assumptions of this paper are clearlythose of evolutionary ecology. In case after case, the func-tion of costly signaling is said to be “enhanced socialstatus and its subsequent political and reproductive ad-vantages.” What marks this as evolutionary ecology isof course the emphasis on “reproductive advantages.”

The evidence that costly signalers have greater “repro-ductive success” than others is not presented in any de-tail. I would think that desires for increased social status(often including the acquisition of prestigious mates) andpower would be sufficient motives for costly signalingamong humans even if reproductive success is relevantto analogous phenomena among other species.

Perhaps the paper’s most debatable assertion is that“signaling theory allows us to address issues of symbolicvalue with rigorous empirical data and a set of testablepredictions derived from a body of theory that is linkedto individual strategizing and evolutionary dynamics.”Little supporting evidence is presented for this bold state-ment. The paper includes assertions such as “More rig-orous formulation of signaling arguments and empiricaltests thereof are needed,” “As in other domains, muchfurther research is needed to subject these plausible sce-narios to rigorous empirical test,” “Application of thistheory to human social behavior is at an early stage andjudgment of its value must await empirical evaluation,”and “The precise benefits to signalers and receivers havenot been measured in most cases.” One of the majorcritiques of functionalist explanations is their nonfalsi-fiability; I see no reason to think that the explanationsof signals offered here are any different.

A notable feature of this paper is its reexamination ofconventional explanations for food sharing among for-agers. Most anthropologists explain such generosity (of-ten “unconditional”) with respect to uncertain returnsto subsistence activities. From this perspective, foodsharing is a risk-reducing activity that enables membersof foraging groups to be fed even on days on which theirsubsistence efforts are unsuccessful. Food sharing alsohelps those who are ill or are working on non-subsis-tence-related tasks. Bliege Bird and Smith suggest thatrisk minimization is “a welcome but unanticipated out-come of rather than a motive for food sharing.” Theyemphasize instead the social benefits that accrue tothose who signal their survival skills by being able toshare some of the meat and plant foods they obtain.These two explanations of food sharing do not strike meas contradictory; both may help explain unconditionalgenerosity among foragers.

lee cronkDepartment of Anthropology, Rutgers University, NewBrunswick, NJ 08901-1414, U.S.A. ([email protected]). 29 x 04

Human behavioral ecologists have been promoting sig-naling theory for the study of cultural phenomena fornearly two decades (e.g., Harpending, Draper, and Rogers1987; Kaplan 1987; Cronk 1991, 1994a, b, 1995, 1999,2001, n.d.; Irons 1996). Bliege Bird and Smith report pro-gress on this front, identify some parallels with otherapproaches, and provide ethnographic examples thatcostly-signaling theory can elucidate. I am already onrecord as sharing their general position and will limit mycomment to a few points.


Bliege Bird and Smith identify parallels betweencostly-signaling theory and the insights of Veblen,Mauss, and Bourdieu. While these theorists are likely tobe familiar to cultural anthropologists, other social sci-entists have independently developed similar ideas.Schelling (1960) provided insights about signals of com-mitment, Spence (1973) developed a model of job-marketsignaling, Frank (1988) applied the same logic to signalsof moral commitment, and political scientists (e.g.,Baumgartner and Leech 1998, Kollman 1998) have madesimilar arguments about the lobbying efforts of interestgroups. Even Theodore Geisel, better known as Dr. Seuss(1961), explored indexical signals (Maynard Smith andHarper (2003) and the circumstances in which theymight be undermined by innovation.

Costly signals’ ubiquity and prominence make theman appropriate choice for an article of this kind. However,as Bliege Bird and Smith point out, signaling theory alsooffers other insights. Among the most promising is theexamination of the role of receiver psychology in signaldesign (Guilford and Dawkins 1991, Rowe 1999). BliegeBird and Smith refer to this briefly in their discussion ofsensory exploitation, suggesting that a consideration ofthis factor might shed light on such phenomena as ad-vertising, propaganda, and pornography. Other examplesalready in the literature of signals with design featuresthat reflect the importance of receiver psychology in-clude “motherese” (Fernald 1992), kin terms in politicalrhetoric (Salmon 1998), kin terms and mating competi-tion (Chagnon 1988, 2000), derogation of same-sex mat-ing competitors (Buss and Dedden 1990), features of re-ligious concepts (Boyer 1999) and rituals (Sosis andAlcorta 2003:265), symmetry in ceramic designs (Wash-burn 1999), and neoteny in toy design (Hinde and Barden1985).

Signaling theory’s breadth has implications for the wayit is best approached by anthropologists. Bliege Bird andSmith’s method of first identifying behaviors that areremarkable for their costliness is representative of muchof the existing literature. This approach is vulnerable tothe criticism that it stacks the deck in favor of costly-signaling theory. An alternative is to identify a signalwithout regard to its apparent costliness and then explorea variety of explanations, including but not limited tocostly-signaling theory, for its design features. For ex-ample, some colleagues and I chose to study cosmeticsnot because of their costs but because of their popularity,their importance in the lives of many people, and theirrelevance to the study of facial attractiveness. Whilewomen may spend a great deal of time and money oncosmetics, such costs are not related to their underlyingqualities in the way predicted by costly-signaling theory.Our preliminary findings indicate that receiver psychol-ogy is more important than cost in understanding cos-metics use, with women using them primarily to maketheir faces more noticeable and memorable (Cronk et al.2002, Milroy et al. 2002, Wigington et al. 2004).

As Bliege Bird and Smith explain, the common andconflicting interests of signalers and receivers are keysto understanding signal design. In evolutionary terms, a

conflict of interests exists between two parties when se-lection would favor a different outcome for their inter-action if it were determined solely by selection on genesin one party or the other (Maynard Smith 1991, Trivers1974). Two parties have common interests when selec-tion acting on genes in both of them would favor thesame outcome from their interaction. Bliege Bird andSmith state that costly signaling is favored when sig-nalers and receivers have “partially competing inter-ests.” It is possible to describe this situation more pre-cisely. Specifically, costly-signaling theory is relevantwhere there are confluences of interest between individ-ual signalers and individual receivers in contexts wherethe broader categories to which signalers and receiversbelong have conflicting interests. For example, predatorsand prey as categories have conflicting interests, but in-dividual predators and individual prey, like the ungu-lates mentioned by Bliege Bird and Smith, experience aconfluence of interests if the prey are truly capable ofmaking a pursuit not worth the predator’s bother. Theconflict of interest in that case is not between the alert,physically fit individual of the prey species and the pred-ator but rather between different members of the preyspecies. Similar reasoning can be applied to other situ-ations where costly and indexical signals are common,such as courtship displays.

lourdes giordaniDepartment of Anthropology, 124 WSB, StateUniversity of New York at New Paltz, New Paltz, NY12561, U.S.A. ([email protected]). 1 xi 04

Bliege Bird and Smith offer us a clear and well-structuredarticle that highlights potential applications of signalingtheory for the study of human behavior and culture. Thistheory stems from biology’s interest in the role that in-formation plays in the evolution and functioning of bi-ological systems. In particular, it owes much to the workof Amotz Zahavi, an Israeli evolutionary biologist whohas sought to explain the evolution of structures andbehaviors that appear to endanger organisms and, thus,to decrease their fitness. As is the case in sociobiologyand related fields of inquiry (e.g., evolutionary psychol-ogy and behavioral ecology), sexual selection as a sign offitness and cost/benefit estimates are key to signalingtheory.

The theory, in addition, favors visible behaviors inwhich sender(s) and receiver(s) of information interactand in the process appraise each other and gain somebenefit. Moreover, it assumes that cost helps ensure thereliability of the signal (Zahavi 1979, 1993). This is prob-lematic when it comes to humans because they engagein a variety of concealed and ambiguous actions that areoften costly. Bliege Bird and Smith, to their credit, at-tempt to present a nuanced argument and concede thatthis exposes one of the theory’s limitations, specifically,its inability to account for costly unconditional acts ofgenerosity that are secret or anonymous.

However, it seems to me that Bliege Bird and Smith


must scrutinize further the cost-reliability component ofthe theory, since the ethnographic record reveals manyinstances in which humans invest a great deal in the cre-ation and propagation of unreliable and ambiguous sig-nals. This is particularly the case in the realm of politics,where bluffing and deception are rather common (Bailey1988). In fact, among the members of any human societyone will commonly find multiple and conflicting versionsof what constitutes appropriate action, with the ambiguitythus generated often leading to change and innovation. Ifsignaling theory is to have the broad applications thatBliege Bird and Smith hope for, it must account for bothbehavioral and cultural continuity and change and do sowithout promoting reductionisms that do violence to thegamut of human possibilities. For instance, they suggestthat stylistic traditions in art may facilitate within-groupcomparisons. Though this may be the case in some in-stances, stylistic traditions may also persist because it isthe process of re-creating/copying traditional motifs (in-cluding the psychological states associated with it) thatis valued rather than the final product that may enter therealm of invidious comparison.

Advocates of signaling theory should also explore thefact that humans generate a plethora of costly nonutil-itarian signals, often creating (materially and symboli-cally) and embellishing more than is necessary. This isclearly seen in the evolution of various graphic arts,where earlier naturalistic forms through time becomemore elaborate and stylized to the point that their mean-ing is accessible to a select few. Though the drawing ofsocial boundaries frequently underlies such develop-ments, I contend that there is more to this, since sociallines can be drawn via less elaborate and costly means.It seems that the human mind generates a multitude ofsignals to satisfy itself and in the process generates“waste.” While Bliege Bird and Smith will turn wasteinto nonwaste and explain the expenditures involved byappealing to the “manipulation of social relations” and“status building” (and their ultimate reproductive ben-efits), I suggest that it may be a by-product of the humanmind that may serve no purpose at all. Moreover,waste—even if costly—can be fun; the mind can derivepleasure from wasteful acts (signaling in this case) thatare costly because the mind also “plays.” While somemay argue that it actually needs to play, I will go fartherand claim that it does so even when there is no apparentneed as such.

Lest I be misunderstood, let me emphasize that I do notconsider status and manipulation of social relations un-important. To me they are part of a much larger storyabout the nature of human beings. For instance, uncon-ditional sharing while trekking can promote bonding anda feeling of security among members of an Ache trekkingparty, all of whom may face potential dangers (e.g., hostileencounters with Paraguayan nationals and physical in-juries). The selective incentive may not be social status(with its additional political and reproductive gains) butthe survival of the group regardless of individual personalstatus. Indeed, the fact that not all humans are equallymotivated to reproduce and that a significant portion of

our lives is not directly concerned with mate selection forthe purpose of reproduction should encourage us to ask ifsignaling theory places inordinate emphasis on sexual se-lection. I hope that scholars of gender and abstemiousreligious communities will join this interesting discus-sion.

edward h. hagen and peter hammersteinInstitute for Theoretical Biology, HumboldtUniversity, Berlin, Germany ([email protected]). 15 xi 04

The essence of signaling theory, as Bliege Bird and Smithexplain, is what is variously termed private, incomplete,or asymmetrical information—some participants in a so-cial interaction have difficult-to-observe qualities thatare critical to the decision making of other participants.Bliege Bird and Smith insightfully apply signaling theoryto phenomena frequently observed in nonmarket econ-omies. This theory has also been widely applied to phe-nomena in market economies. One of the first and mostinfluential of such papers was a formal model of the roleof education in the job market (Spence 1973) that wasessentially identical to biologists’ arguments about therole of the peacock’s tail in the mating market (seeBowles and Hammerstein 2003). The productivity of jobapplicants is a difficult-to-observe quality. Educationdoes not necessarily improve job applicants’ productivity(learning Latin has little application to business), but ifthe time cost of education is negatively correlated withproductivity (because, for example, high-productivity in-dividuals learn more quickly), then achieved educationlevel is an honest indicator of productivity and individ-uals with more education will be hired at higher wages.Signaling theory has also been applied to high borrowingrates on local Third World markets, discriminationagainst minorities in labor markets, unemployment,credit rationing, and bargaining. The work of Spence andothers in the 1970s laid the foundation for a general the-ory of the role of asymmetric information in markets,recently earning them Nobel Prizes in economics.

We agree with Bliege Bird and Smith that signalingtheory is likely to be broadly useful in the social sciences.It has been argued, for example, that honest signalingmay play an important role in powerful psychologicalstates such as depression and suicide (e.g., Hagen 2002,2003; Watson and Andrews 2002). Hagen and Bryant(2003) argue that group displays of music and dance andother forms of elaborate cultural production may be cred-ible signals of coalition quality because they require thelong-term association of the participants as well as thesurplus production necessary to support the cultural pro-ducers. This approach may illuminate two features ofhuman socioecology that distinguish it from that of otherprimates. First, reliable signals of coalition quality per-mit the evolution of group-level alliances resulting inuniquely human regionwide social structures (in con-trast, the segmentary lineage system of hamadryas ba-boons requires consanguineal ties [Rodseth et al. 1991]).


Second, these signals entail cultural transmission. Giventhat song learning in birds is one of the few clear casesof cultural transmission in nonhuman animals and thatsinging is common in apes (i.e., gibbons and siamangs),coalition signaling may be important to the evolution ofhuman cultural transmission as, for example, a preadap-tation for other forms of culture.

Although we find many of Bliege Bird and Smith’s ex-amples compelling, we are less convinced that signalingtheory adequately explains the frequent and widespreadfood sharing commonly found in small-scale societies.They cogently argue that because individual quality canchange over time, signals of quality must be repeatedlysent. But do they have to be sent almost every day? AsBliege Bird and Smith note, the number of models of foodsharing is surprisingly small. Because we believe that con-sideration of a broader range of models would be useful,we outline one that has been relatively ignored by evo-lutionary ecologists. Corporate groups, including hunter-gatherer bands and lineages, typically defend a territory,and this defense entails political and military costs thatare distributed across group members. Corporate groupmembers pay these costs because the territory is valuableto them and they therefore consider themselves its “own-ers.” A corporate group that “owns” a territory also“owns” whatever valuables that territory contains. If thevalue of the territory is that it has wild game, then—andhere is our main point—that game is “owned” by thegroup and not by the particular hunter who kills the game.Hunters are “employees” of the corporate group, and theirproduction (hunted game) belongs to the corporation andmust be distributed among its members. Among the Ache,for example, meat is widely distributed and hunters veryrarely eat from their own kills (Hill and Hurtado 1996:65). Hunters are then “paid” in a variety of “wages” thatare culturally specific and (in contrast to Ache practice)may include rights to distribute the meat, to receivechoice parts of it, or to consume some fraction of the killprior to distribution and, as Bliege Bird and Smith suggest,increased status with all its perks.

We have merely sketched this group-ownership modelof food sharing, and it confronts the many challenges typ-ically associated with larger-scale collective actions, in-cluding free riding, division of labor, management, anddistribution of benefits. Despite this one point of partialdisagreement, we consider Bliege Bird and Smith’s articlea powerful and welcome blow to the wall that has dividedmaterialist and symbolic theories in the social sciences.

fraser d. neimanDepartment of Archaeology, Monticello, Box 316,Charlottesville, VA 22902, U.S.A. ([email protected]). 12 xi 04

Models of costly signaling are among the most excitingdevelopments in evolutionary theory in the past severaldecades. Bliege Bird and Smith brilliantly highlight theirrevolutionary relevance for anthropology and archae-ology. Archaeologists have long relied on monumental

architecture, fancy portable artifacts, and, most recently,evidence of feasting to construct narratives about socialcomplexity and power (e.g., Dietler and Hayden 2001,Trigger 1990), but they have lacked a theoretical accountof the causal dynamics that link extravagant artifacts tothe social and cultural dynamics that are alleged to ac-count for them. As Bliege Bird and Smith show, signalingtheory offers this missing link and thereby promises toconvert what has been a largely free-form interpretiveenterprise into a model-based analytical one. I hope toadvance their project by sketching the potential of sig-naling theory in historical archaeology.

Over 30 years ago James Deetz recognized, in the ar-chaeological record of seventeenth- and eighteenth-cen-tury British North America, important changes in theconsumption of objects at a variety of scales from diningparaphernalia to houses. He interpreted these changes assymptoms of a cognitive shift from a medieval to a Geor-gian mind-set typified by a novel emphasis on the sep-arateness of the individual and the orderly partitioningof daily life (1972, 1977). Since then social and economichistorians, using additional evidence from tax records,probate inventories, and merchant accounts, have rec-ognized an increase in the salience of fashion and therate of consumption across the Atlantic world after 1600.They have dubbed the trend “the consumer revolution”(e.g., Carson 2003). Their work offers independent evi-dence for the idea that much of what archaeologiststhink of as “Georgian” is actually an increase in theamount of effort devoted to social display by individualsat all wealth levels (Carson 1994, de Vries 1993, Pogue2001).

The period witnessed a dramatic increase in the fre-quency of imported, refined ceramic vessels. Some vesselforms were associated with the consumption of exoticand costly beverages such as punch, tea, and coffee.Other forms—mugs, cups, and later plates—replacedfunctional equivalents that had traditionally been madein wood and pewter (Smart 1989, 1994). Increasingly ves-sels came in matching sets. These changes are linkedbecause they represent increases in the costs paid byceramics consumers. Tea, coffee, and punch required notonly exotic and costly ingredients but also resource out-lays for novel vessel forms. Replacing traditional pewtervessels with ceramic ones involved additional costs be-cause ceramics had shorter use lives and old pewter couldbe profitably recycled (Neiman, McFaden, and Wheeler2000). Finally, the emphasis on matched sets meant thatpurchase costs could not be amortized over a long period.

A similar case can be made for architectural change.For example, the central passages that typified whatDeetz called the “Georgian” house plan were far largerthan the lobby entries that they replaced in New Eng-land, offering more scope for costly display in the firstroom visitors entered. Increasingly, the stylistic detailsof the most prestigious displays were drawn from clas-sical Roman and, later, Greek sources (Sweeney 1994).Consumers bore the increased labor costs required fortheir correct production. They also had to acquire theknowledge to discriminate correct from incorrect from


the study of expensive folios. Novel symmetrical facadesattested to their builders’ ability to build an entire struc-ture at once.

Signaling theory points to three factors that are likelyto account for these changes: (1) the increase in geograph-ical mobility that accompanied European colonial expan-sion, bringing together strangers who relied on signals inassessing one another’s competitive abilities (Carson1994); (2) the increasing proportions of the populationwhose livings were made in nonagricultural niches inwhich the determinants of economic and social successwere not available for public inspection, as a farmer’sfields were; and (3) the resulting increased rates of socialmobility, both up and down, which placed a premium onsignals engineered to be up-to-date indicators of signalers’qualities.

Signaling theory promises interpretations of cultural dy-namics that span oceans and centuries and helped createour modern world, but so did the Georgian mind-set andits teleological, neo-Marxist cousin the Georgian order ofmerchant capitalism (Leone 1988). Why should we bemore sanguine about the long-term credibility of accountsbased in signaling theory? First, the link between artifactsand their meaning is causal, not symbolic. Signaling mod-els are explicit enough to deliver detailed predictionsabout the design and distribution of variation relative tohistorically specific social and environmental contexts(e.g., Neiman 1997). Moreover, as Bliege Bird and Smithshow, they provide the best accounts we have of signalingin the historical trajectories of human and nonhuman spe-cies. The diversity of successful applications would beunlikely if signaling theory did not capture an importantprocess in the history of all living things.

Many of the theoretical notions that have dominatedhistorical—and much prehistoric—archaeology inevita-bly produce one-off results that are of marginal useful-ness to colleagues studying other times and places, de-spite their individual brilliance. In contrast, signalingtheory will allow archaeologists to build and evaluatemodels of the historical dynamics that gave us Chacoangreat houses, Chesapeake mansions, and Mayan stelae.Signaling models may mitigate the balkanization thatcharacterizes our postmodern professional lives.

Reply

rebecca bliege bird and eric aldensmithPalo Alto, Calif., and Seattle, Wash., U.S.A. 1 xii 04

We thank the commentators for their thoughtful andwell-reasoned replies. We are particularly grateful toCronk and to Hagen and Hammerstein for drawing at-tention to the development of signaling theory in eco-nomics, which we gave short shrift in the interest ofmore ethnographic details. Cronk is quite correct inpointing out that an approach which focuses on only

obviously costly signals “stacks the deck” in any anal-ysis. Our point in this paper was not to evaluate alter-native explanations for the design of signals but simplyto show how costly-signaling theory in particular canprovide adaptive explanations for apparently “wasteful”traits or behaviors. We did take pains to note that ques-tions about the function and evolution of design featuresin any particular signal should consider a variety of ex-planations and design forces, including receiver psy-chology.

As Chibnik points out, our notion of “design forces”shaping adaptations (including signaling strategies) is ag-nostic on the question of whether actors (signalers andreceivers) are consciously pursuing strategic goals. Wethink that the fact that the costly-signaling frameworkapplies to both intended and unintended aspects of com-munication and strategic interaction is a strength, not aflaw. But there are some tricky issues raised by function-alist analyses. Following Elster (1983), we define a func-tionalist explanation as one in which some action or in-stitution A is explained by its beneficial consequences B.The immediate problem is that a consequence (B) is usedto explain the action (A) that produces it. This seemingparadox can be resolved if there is a causal feedback loopsuch that the presence of B increases the production of A(Elster 1983). Prominent candidates for such a feedbackloop include learning (positive reinforcement and its var-iants) and natural selection, which also appear on our listof adaptive design forces. The poor reputation of func-tionalist analysis comes from the fact that many analysesfail to suggest any feedback loop linking consequence andaction, thus leaving the causal connection mysterious.When the hypothesized benefits are not ones that wouldbe favored by either natural selection (on genes or culture)or rational choice, the mysterious becomes dubious(Smith and Winterhalder 1992b:42). However, the signal-ing explanations reviewed in our article avoid both ofthese problems. They also are exempt from the problemof “nonfalsifiability” raised by Chibnik, as indicated bythe empirical studies we review.

Giordani states that signaling theory stems primarilyfrom “biology’s interest” in communication and, morespecifically, from the work of Amotz Zahavi. BecauseZahavian “handicap theory” was developed to explainhow seemingly costly and maladaptive male ornamen-tation could actually benefit male reproductive success,she suggests that signaling theory “places inordinate em-phasis on sexual selection” and, by association, repro-duction. As we (and nearly all commentators) point out,signaling theory owes as much to economics, decisiontheory, and social theory as it does to evolutionary bi-ology. Sexual selection is not an inherent component ofall signaling contexts, and, as we point out numeroustimes, signaling is not just about males attempting toimpress potential mating partners with honest signals ofhidden genetic quality. Rather, signaling games can in-volve women signaling the ability to invest time andeffort in the cooking of elaborate feasting dishes as partof a collective signal of lineage quality or men sendinghonest signals of coalition commitment to political sup-


porters. Such signals are not necessarily designed to ac-quire immediate reproductive advantages but do func-tion as a way to acquire social, symbolic, or materialbenefits. These benefits may translate into competitiveadvantages in acquiring strong marriage alliances or ahigh-quality partner, in reducing infant mortality, or inachieving access to resources during periods of scarcity—all of which could have reproductive benefits.

Giordani’s skepticism about the value of using repro-ductive consequences as a currency for measuring stra-tegic outcomes is shared by many social scientists. Ifdesire for increased power and social status is enough tomotivate signaling strategies (Chibnik), why bother withreproductive advantages? The problem here is the failureto distinguish between proximate and ultimate (evolu-tionary) causality (Mayr 1965). Within limits, people pre-fer more status or power to less, wealth to poverty, andso on. The only nonteleological and nonmystical expla-nation for these preferences (and the innate and learnedcognitive processes which underlie them) is the theoryof natural selection. Natural selection favors cognitivemechanisms that motivate strategic behavior correlatedwith reproductive success, whether or not this is some-thing the actors are consciously aware of. Standardcausal explanation of human behavior focuses on theworkings of these innate, learned, and/or consciouschoices, which collectively constitute proximate mech-anisms; ultimate or evolutionary analysis focuses onwhy these mechanisms exist and are designed the waythey are. These two modes of causal analysis are com-plementary, one essentially addressing how questionsand the other why (Mayr 1965), and each is a valid frame-work for analysis of human behavior. However, evolu-tion has endowed humans with unique proximate mech-anisms: capabilities for cultural transmission of beliefsand preferences and for the construction of symbolicworlds. In principle, these cognitive or cultural evolu-tionary processes may decouple choices from adaptivedesign (Boyd and Richerson 1985), but the extent towhich this is the case is an empirical issue that must bedecided by careful tests of evolutionary hypotheses(Smith and Winterhalder 1992b).

Giordani suggests that the prevalence of bluffing, am-biguity, and deception in human communication is prob-lematic for signaling theory. Indeed, signaling theory isdesigned to explain why and under what conditions hon-est communication might be in both parties’ interest—it does not predict that communication must always behonest. If the fundamental conditions for costly or hon-est signaling are not met, we should expect to see a greatdeal of lying, deceit, and misdirection. It is worth point-ing out here that signaling theory does not preclude othertheories of communication, nor does it explain all con-textually patterned signals. However, any theory of com-munication must explain why a high frequency of dis-honesty would not destabilize the communicationsystem altogether by motivating observers to ignore sig-nals. Giordani does not offer any suggestions in thisregard.

Chibnik proposes that risk-reduction explanations of

unconditional food sharing are complementary to sig-naling explanations, whereas Hagen and Hammersteinare skeptical of the latter. First, we should stress that wenever proposed that all instances of generalized foodsharing can be explained by a single model, costly sig-naling or any other. Second, our problem with risk-re-duction explanations is not that they contradict signal-ing ones (as Chibnik portrays our view) but that theexisting evidence does not consistently demonstrate thekinds of safeguards against free riding that would be re-quired to make such a system stable. But (as we alsonoted) costly signaling is only one possible model to ex-plain how unconditional generosity could be evolution-arily stable, and we welcome testable alternative expla-nations. Hagen and Hammerstein propose a “corporate-ownership” account that views food shares as the earn-ings of members of the corporate group (in this case, aresidential band or a lineage) that owns the territory fromwhich the food is harvested. Although we would notdismiss this alternative out of hand, we have some sig-nificant reservations. Two are empirical: most small-scale societies lack the corporate structure or decision-making processes that could allocate shares in themanner envisaged by Hagen and Hammerstein, andmany (particularly hunter-gatherer band societies suchas the Ache, whose generalized food sharing is so prom-inent) are nonterritorial. Finally, there are serious col-lective-action problems raised by the corporate-owner-ship account (as Hagen and Hammerstein acknowledge).

As if anticipating Chibnik’s skepticism about the roleof signaling theory in artistic traditions, Neiman pro-vides a very nice illustration of how it can account forstylistic changes in colonial American vessel forms andarchitecture. In this case, rather than competition amongthe artisans themselves to produce finer products (as wesuggested), it is competition among consumers to displaywealth and good taste that drives the production of in-creasingly expensive and high-quality items. This neatlyshows how costly-signaling theory can be used to explainboth conformity and change.

In closing, we want to reiterate that application of sig-naling theory to anthropological phenomena is at a veryearly stage. This is why we used such cautious and pro-spective language in our article when describing this areaof research. Chibnik seizes on these statements to ques-tion our assertion that “signaling theory allows us toaddress issues of symbolic value with rigorous empiricaldata and testable predictions” derived from evolutionaryand strategic-actor theory. Yet, as is exemplified by anumber of studies discussed in our article (and otherscited by Cronk, Hagen and Hammerstein, and Neiman),signaling theory has clearly begun to deliver on thesepromises. In any case, our primary goal was to explainthe promise of signaling theory clearly enough to mo-tivate other researchers to conduct research on patternsof human social behavior that have otherwise remainedmysterious or resistant to convincing explanation. Wejudge the responses of most commentators (includingChibnik) as indicating that our goal may well be attained.


References Cited

a l t m a n , j o n c . 1987. Hunter-gatherers today: An aboriginaleconomy in North Australia. Canberra: Australian Institute ofAboriginal Studies.

a l v a r d , m . 2003. The adaptive nature of culture. Evolution-ary Anthropology 12:136–49. [ma]

a l v a r d , m . , a n d d . n o l i n . 2002. Rousseau’s whale hunt?Coordination among big-game hunters. current anthropol-ogy 43:533–59. [ma]

a m e s , k e n n e t h m . 2001. Slaves, chiefs, and labour on thenorthern Northwest Coast. World Archaeology 33:1–17.

a r a n y o s i , e . f . 1999. Wasteful advertising and variance re-duction: Darwinian models for the significance of nonutilitar-ian architecture. Journal of Anthropological Archaeology 18:356–75.

a x e l ro d , ro b e r t . 1984. The evolution of cooperation. NewYork: Basic Books.

a x e l ro d , ro b e r t , a n d w i l l i a m d . h a m i l t o n . 1981.The evolution of cooperation. Science 211:1390–96.

b a i l e y, f . g . 1998. Humbuggery and manipulation: The artof leadership. Ithaca: Cornell University Press. [lg]

b a u m g a r t n e r , f r a n k r . , a n d b e t h l . l e e c h . 1998.Basic interests: The importance of groups in politics and polit-ical science. Princeton: Princeton University Press. [lc]

b e c k e t t , j e r e m y r . 1988. The Torres Strait Islanders: Cus-tom and colonialism. Cambridge: Cambridge University Press.

b e r g s t ro m , c a r l t . , a n d m i c h a e l l a c h m a n n .1998. Signaling among relatives. 3. Talk is cheap. Proceedingsof the National Academy of Sciences. U.S.A. 95:5100–105.

b i n m o r e , k . 1994. Game theory and the social contract.Cambridge: MIT Press. [ma]

b i s h o p , c h a r l e s a . 1987. Coast-interior exchange: The ori-gins of stratification in northwestern North America. ArcticAnthropology 24:72–83.

b l i e g e b i r d , r e b e c c a l . , a n d d o u g l a s w. b i r d .1997. Delayed reciprocity and tolerated theft: The behavioralecology of food sharing strategies. current anthropology 38:49–78.

———. n.d. “Human hunting seasonality,” in Primate seasonal-ity: Implications for human evolution. Edited by D. K. Brock-man and C. E. van Schaik. Cambridge: Cambridge UniversityPress.

b l i e g e b i r d , r e b e c c a , l . , d o u g l a s w. b i r d , e r i ca . s m i t h , a n d g e o f f r e y k u s h n i c k . 2002. Risk andreciprocity in Meriam food sharing. Evolution and Human Be-havior 23:297–321.

b l i e g e b i r d , r e b e c c a l . , e r i c a . s m i t h , a n dd o u g l a s w. b i r d . 2001. The hunting handicap: Costlysignaling in human foraging strategies. Behavioral Ecology andSociobiology 50:9–19.

b l u r t o n j o n e s , n i c h o l a s g . 1986. Natural selection andhuman birth weight. Annals of Human Biology 5:487–89.

b o d e n h o r n , b a r b a r a . 2000. “It’s good to know who yourrelatives are but we were taught to share with everybody:Shares and sharing among Inupiaq households,” in The socialeconomy of sharing: Resource allocation and modern hunter-gatherers. Edited by George W. Wenzel, Grete Hoverlsrud-Broda, and Nabuhiro Kishigami, pp. 27–60. Senri EthnologicalStudies 53. Osaka: National Museum of Ethnology.

b o o n e , j a m e s l . 1998. The evolution of magnanimity: Whenis it better to give than to receive? Human Nature 9(1):1–21.

———. 2000. “Social power, status, and lineage survival,” in Hi-erarchies in action: Cui bono? Edited by Michael W. Diehl, pp.84–110. Carbondale: Southern Illinois University Press.

b o o n e , j a m e s l . , a n d k a r e n l . k e s s l e r 1999. Morestatus or more children? Social status, fertility reduction, andlong-term fitness. Evolution and Human Behavior 20:257–77.

b o u r d i e u , p i e r r e . 1977. Outline of a theory of practice.Cambridge: Cambridge University Press.

———. 1984. Distinction: A social critique of the judgement of

taste. Translated by Richard Nice. Cambridge: Harvard Univer-sity Press.

———. 1990. The logic of practice. Stanford: Stanford UniversityPress.

b o w l e s , s a m u e l , a n d h e r b e r t g i n t i s . 1998. Themoral economy of communities: Structured populations andthe evolution of pro-social norms. Evolution and Human Be-havior 19:3–25.

———. 2003. The evolution of strong reciprocity: Cooperation inheterogeneous populations. Theoretical Population Biology 17:17–28.

b o w l e s , s . , a n d p . h a m m e r s t e i n . 2003. “Does markettheory apply to biology?” in Genetic and cultural evolution ofcooperation. Edited by P. Hammerstein, pp. 153–65. Cam-bridge: MIT Press. [ehh, ph]

b o w s e r , b r e n d a j . 2000. From pottery to politics: An eth-noarchaeological study of political factionalism, ethnicity, anddomestic pottery style in the Ecuadorian Amazon. Journal ofArchaeological Method and Theory 7:219–48.

b o y d , ro b e r t , h e r b e r t g i n t i s , s a m u e l b o w l e s ,a n d p e t e r j . r i c h e r s o n . 2003. The evolution of altruis-tic punishment. Proceedings of the National Academy of Sci-ences, U.S.A. 100:3531–35.

b o y d , ro b e r t , a n d p e t e r j . r i c h e r s o n . 1985. Cul-ture and the evolutionary process. Chicago: University of Chi-cago Press.

———. 1995. Why does culture increase human adaptability?Ethology and Sociobiology 16:125–43.

b o y e r , p a s c a l . 1999. Cognitive tracks of cultural inheri-tance: How evolved intuitive ontology governs cultural trans-mission. American Anthropologist 100:876–89. [lc]

b r i l o t , b . o . , a n d r . a . j o h n s t o n e . 2003. The limitsto cost-free signalling of need between relatives. Proceedings ofthe Royal Society of London B 270(1519):1055–60.

b u s s , d a v i d m . , a n d l i s a a . d e d d e n . 1990. Deroga-tion of competitors. Journal of Social and Personal Relation-ships 7:395–422. [lc]

c a ro , t i m m . 1994. Ungulate antipredator behaviour: Prelimi-nary and comparative evidence from African bovids. Behaviour128:189–228.

c a r s o n , c a ry. 1994. “The consumer revolution in colonialBritish America: Why demand?” in Of consuming interests:The style of life in the eighteenth century. Edited by CaryCarson, Ronald Hoffman, and Peter J. Albert, pp. 483–697.Charlottesville: University Press of Virginia. [fdn]

———. 2003. “Consumption,” in A companion to colonialAmerica. Edited by Daniel Vickers, pp. 334–65. Malden, Mass:Blackwell. [fdn]

c a s h d a n , e l i z a b e t h a . 1985. Coping with risk: Reciproc-ity among the Basarwa of Northern Botswana. Man 20:454–74.

c h a g n o n , n a p o l e o n a . 1988. “Male Yanomamo manipu-lations of kinship classifications of female kin for reproductiveadvantage,” in Human reproductive behavior: A Darwinianperspective. Edited by L. Betzig, M. Borgerhoff Mulder, and P.Turke, pp. 23–48. Cambridge: Cambridge University Press. [lc]

———. 2000. “Manipulating kinship rules: A form of male Yano-mamo reproductive competition,” in Adaptation and humanbehavior: An anthropological perspective. Edited by L. Cronk,W. Irons, and N. A. Chagnon, pp. 115–31. Hawthorne, N.Y.:Aldine de Gruyter. [lc]

c o d e r e , h e l e n . 1950. Fighting with property: A study ofKwakiutl potlatching and warfare, 1791–1930. Monographs ofthe American Ethnological Society 18.

c ro n k , l e e . 1991. Human behavioral ecology. Annual Reviewof Anthropology 20:25–53. [lc]

———. 1994a. Evolutionary theories of morality and the manipu-lative use of signals. Zygon 29(1):81–101. [lc]

———. 1994b.The use of moralistic statements in social manipu-lation : A reply to Roy A. Rappaport. Zygon 29:351–55. [lc]

———. 1995. Is there a role for culture in human behavioralecology? Evolution and Human Behavior 16:181–205. [lc]

———. 1999. That complex whole: Culture and the evolution ofhuman behavior. Boulder: Westview Press. [lc]


———. 2001. Anthropology and the evolutionary study of humanbehavior. Research in Biopolitics 8:1–30.[lc]

———. n.d. “Behavioral ecology and the social sciences,” inMissing the revolution: Evolutionary psychology for social sci-entists. Edited by J. Barkow. Oxford: Oxford University Press.[lc]

c ro n k , l e e , l o r n e c a m p b e l l , a l i s o n m i l ro y, a n dj e f f ry a . s i m p s o n . 2002. Cosmetics as a signaling sys-tem. Paper presented at the annual meeting of the AmericanAnthropological Association, New Orleans, La. [lc]

c ro n k , l e e , n a p o l e o n c h a g n o n , a n d w i l l i a m g .i ro n s . Editors. 2000. Adaptation and human behavior: Ananthropological perspective. Hawthorne, N.Y.: Aldine deGruyter.

d a w k i n s , r . , a n d j . k r e b s . 1978. “Animal signals: Infor-mation or manipulation?” in Behavioural ecology. Edited by J.Krebs and N. Davies, pp. 282–309. Oxford: Blackwell ScientificPublishers.[ma]

d e e t z , j a m e s . 1972. “Ceramics from Plymouth, 1620–1835:The archaeological evidence,” in Ceramics in America, pp.15–40. Charlottesville: University Press of Virginia. [fdn]

———. 1977. In small things forgotten: An archaeology of earlyAmerican life. New York: Doubleday. [fdn]

d e v r i e s , j a n . 1993. “Between purchasing power and theworld of goods: Understanding the household economy in earlymodern Europe,” in Consumption and the world of goods. Ed-ited by John Brewer and Roy Porter, pp. 85–132. London: Rout-ledge. [fdn]

d i e t l e r , m i c h a e l , a n d b r i a n h a y d e n . Editors. 2001.Feasts: Archaeological and ethnographic perspectives on food,politics, and power. Washington, D.C.: Smithsonian InstitutionPress. [fdn]

d o n a l d , l e l a n d , 1997. Aboriginal slavery on the North-west Coast of North America. Berkeley: University of Califor-nia Press.

d u r h a m , w i l l i a m h . 1990. Advances in evolutionary cul-ture theory. Annual Review of Anthropology 19:187–210.

e l s t e r , j o n . 1983. Explaining technical change: A case studyin the philosophy of science. Cambridge: Cambridge Univer-sity Press.

e n d l e r , j . a . , a n d a . b a s o l o . 1998. Sensory ecology, re-ceiver biases, and sexual selection. Trends in Ecology and Evo-lution 13:415–20.

f e h r , e r n s t , a n d s i m o n g a c h t e r . 2002. Altruisticpunishment in humans. Nature 415 (January 10):137–40.

f e r n a l d , a . 1992. “Human maternal vocalizations to infantsas biologically relevant signals: An evolutionary perspective,”in The adapted mind. Edited by J. Barkow, L. Cosmides, and J.Tooby. Oxford: Oxford University Press. [lc]

f i s h e r , ro n a l d a . 1930. The genetical theory of natural se-lection. Oxford: Clarendon Press.

f i t z g i b b o n , c l a i r e d . , a n d j . h . f a n s h a w e . 1988.Stotting in Thomson’s gazelles: An honest signal of condition.Behavioral Ecology and Sociobiology 23:69–74.

f o r g e , j . a n t h o n y. 1962. Paint—a magical substance. Pal-ette 9:4–16.

———. 1966. Art and environment in the Sepik. Proceedings ofthe Royal Anthropological Institute 1965, pp.23–32.

———. 1970. “Prestige, influence, and sorcery: A New Guineaexample,” in Witchcraft confessions and accusations. Editedby Mary Douglas, pp. 257–75. Association of Social Anthropol-ogists Monographs. London: Tavistock.

f r a n k , ro b e r t h . 1988. Passions within reason: The strate-gic role of emotions. New York: W. W. Norton. [lc]

f r i e d , m o r t o n h . 1967. The evolution of political society.New York: Random House.

g a t e s , h i l l . 2001. Footloose in Fujian: Economic correlatesof footbinding. Comparative Studies in Society and History 43:130–48.

g e t t y, t h o m a s . 1998. Reliable signalling need not be ahandicap. Animal Behaviour 56:253–55.

g i n t i s , h e r b e r t , s a m u e l b o w l e s , ro b e r t b o y d ,

a n d e r n s t f e h r . 2003. Explaining altruistic behavior inhumans. Evolution and Human Behavior 24:153–72.

g i n t i s , h e r b e r t , e r i c a l d e n s m i t h , a n d s a m u e ll . b o w l e s . 2001. Cooperation and costly signaling. Journalof Theoretical Biology 213:103–19.

g o d e l i e r , m a u r i c e . 1986. The making of great men: Maledomination and power among the New Guinea Baruya. Cam-bridge: Cambridge University Press.

g r a f e n , a l a n . 1990. Biological signals as handicaps. Journalof Theoretical Biology 144:517–46.

g u i l f o r d , t i m , a n d m a r i a n s t a m p d a w k i n s . 1991.Receiver psychology and the evolution of animal signals. Ani-mal Behaviour 42(1):1–14. [lc]

g u rv e n , m i c h a e l , w e s l e y a l l e n - a r a v e , k i m h i l l ,a n d m a g d a l e n a h u r t a d o . 2000a. “It’s a wonderfullife”: Signaling generosity among the Ache of Paraguay. Evolu-tion and Human Behavior 21:263–82.

———. 2001. Reservation food sharing among the Ache of Para-guay. Human Nature 12:273–97.

g u rv e n , m i c h a e l , k i m h i l l , h i l l a r d k a p l a n , a n ah u r t a d o , a n d r i c h a r d l y l e s . 2000b. Food transfersamong Hiwi foragers of Venezuela: Tests of reciprocity. HumanEcology 28:171–218.

h a d d o n , a l f r e d c . 1935. General ethnography. Reports ofthe Cambridge Anthropological Expedition to Torres Straits,vol. 1. Cambridge: Cambridge University Press.

h a g e n , e . h . 2002. Depression as bargaining: The case post-partum. Evolution and Human Behavior 23:323–36. [ehh, ph]

———. 2003. “The bargaining model of depression,” in Geneticand cultural evolution of cooperation. Edited by P. Hammer-stein, pp. 95–123. Cambridge: MIT Press. [ehh, ph]

h a g e n , e d w a r d h . , a n d g r e g o ry a . b ry a n t . 2003.Music and dance as a coalition signaling system. Human Na-ture 14:21–51.

h a m e s , r a y m o n d b . 2000. “Reciprocal altruism in Yano-mamo food exchange,” in Adaptation and human behavior:An anthropological perspective. Edited by Lee Cronk, Napo-leon Chagnon, and William Irons, pp. 397–416. Hawthorne,N.Y.: Aldine de Gruyter.

h a r p e n d i n g , h e n ry, p a t r i c i a d r a p e r , a n d a l a nro g e r s . 1987. Human sociobiology. Yearbook of PhysicalAnthropology 30:127–50. [lc]

h a w k e s , k r i s t e n . 1993. Why hunter-gatherers work. cur-rent anthropology 34:341–61.

———. 2000. “Hunting and the evolution of egalitarian societies:Lessons from the Hadza,”in Hierarchies in action: Cui bono?Edited by Michael W. Diehl, pp. 59–83. Carbondale: SouthernIllinois University Press.

h a w k e s , k r i s t e n , a n d r e b e c c a b l i e g e b i r d . 2002.Showing off, handicap signaling, and the evolution of men’swork. Evolutionary Anthropology 11:58–67.

h a w k e s , k r i s t e n , j a m e s f . o ’ c o n n e l l , a n d n i -c h o l a s g . b l u r t o n j o n e s . 2001. Hadza meat sharing.Evolution and Human Behavior 22:113–42.

h e n r i c h , j o s e p h , a n d f r a n c i s c o j . g i l - w h i t e .2001. The evolution of prestige: Freely conferred deference as amechanism for enhancing the benefits of cultural transmis-sion. Evolution and Human Behavior 22:165–96.

h i l l , k . , a n d m . h u r t a d o . 1996. Ache life history. NewYork: Aldine de Gruyter. [ehh, ph]

h i l l , k i m , a n d h i l l a r d k a p l a n . 1999. Life historytraits in humans: Theory and empirical studies. Annual Re-view of Anthropology 28:397–430.

h i n d e , ro b e r t a . , a n d l . a . b a r d e n . 1985. The evolu-tion of the teddy bear. Animal Behaviour 33:1371–73. [lc]

h r d y, s a r a h b l a f f e r . 1999. Mother nature: A history ofmothers, infants, and natural selection. New York: PantheonBooks.

i ro n s , w i l l i a m g . 1979. “Cultural and biological success,”in Evolutionary biology and human social behavior: An evolu-tionary perspective, pp. 257–72. North Scituate: Duxbury.

———. 1996. In our own self-image: The evolution of morality,deception, and religion. Skeptic 4(2):50–61. [lc]


———. 2001. “Religion as a hard-to-fake sign of commitment,”in Evolution and the capacity for commitment. Edited by Ran-dolph Nesse, pp. 292–309. New York: Russell Sage Foundation.

j o h n s t o n e , ru f u s a . 1997. “The evolution of animal sig-nals,” in Behavioural ecology: An evolutionary approach. Ed-ited by John R. Krebs and Nicholas B. Davies, pp. 155–78. Ox-ford: Blackwell.

k a b e r ry, p h y l l i s m . 1971. “Political organization amongthe northern Abelam,” in Politics in New Guinea. Edited byRonald M. Berndt and Peter Lawrence, pp. 35–73. Nedlands:University of Western Australia Press.

k a p l a n , h i l l a r d . 1987. Human communication and con-temporary evolutionary theory. Research on Language and So-cial Interaction 20:79–139. [lc]

k a p l a n , h i l l a r d , a n d k i m h i l l . 1985a. Food sharingamong Ache foragers: Tests of explanatory hypotheses. cur-rent anthropology 26:223–46.

———. 1985b. Hunting ability and reproductive success amongmale Ache foragers: Preliminary results. current anthropol-ogy 26:131–33.

k a p l a n , h i l l a r d , k i m h i l l , a n d a . m a g d e l e n ah u r t a d o . 1990. “Fitness, foraging, and food sharing amongthe Ache,” in Risk and uncertainty in tribal and peasanteconomies. Edited by Elizabeth Cashdan, pp. 107–43. Boulder:Westview Press.

k o h n , m a r e k , a n d s t e v e n m i t h e n . 1999. Handaxes:Products of sexual selection? Antiquity 73:518–26.

k o h l e r , t i m o t h y a . 1998. “Public architecture and powerin pre-Columbian North America,” in Power, monuments, andcivilization. London: University College London Press.

k o k k o , h a n n a , ro b e r t b ro o k s , m i c h a e l d . j e n -n i o n s , a n d j o s e p h i n e m o r l e y. 2003. The evolution ofmate choice and mating biases. Proceedings of the Royal Soci-ety of London B 270:653–64.

k o l l m a n , k e n n e t h . 1998. Outside lobbying: Public opinionand interest group strategies. Princeton: Princeton UniversityPress. [lc]

k ro e b e r , a l f r e d l . , a n d e . w. g i f f o r d . 1949. Worldrenewal: A cult system of Native Northwest California. Uni-versity of California Anthropological Records 13(1).

l a c h m a n n , m i c h a e l , s z a b o l c s s z a m a d o , a n dc a r l t . b e r g s t ro m . 2001. Cost and conflict in animalsignals and human language. Proceedings of the NationalAcademy of Sciences, U.S.A. 98:13189–94.

l a l a n d , k e v i n n . , a n d g i l l i a n b ro w n . 2002. Senseand nonsense: Evolutionary perspectives on human behavior.Oxford: Oxford University Press.

l e a , d a v i d a . m . 1964. Abelam land and sustenance: Swid-den horticulture in an area of high population density, Maprik,New Guinea. Ph.D. diss., Australian National University, Can-berra, Australia.

———. 1969. “Some non-nutritive functions of food in NewGuinea,” in Settlement and encounter: Geographical studiespresented to Sir Grenfell Price. Edited by A. Grenfell Price, FayGale, and Graham Lawton, pp. 173–84. Melbourne: OxfordUniversity Press.

l e d e r m a n , r e n a . 1990. Big men, large and small: Towards acomparative perspective. Ethnology 29:3–15.

l e o n e , m a r k . 1988. “The Georgian order as the order of mer-chant capitalism in Annapolis, Maryland,” in Recovery ofmeaning, pp. 235–61. Washington, D.C.: Smithsonian Institu-tion Press. [fdn]

l u d v i c o , l . r . , a n d j . a . k u r l a n d . 1995. Symbolic ornot-so-symbolic wounds: The behavioral ecology of humanscarification. Evolution and Human Behavior 16:155–72.

m a l i n o w s k i , b ro n i s l a w. 1935. Coral gardens and theirmagic. Bloomington: Indiana University Press.

m a r l o w e , f r a n k . 2000. The patriarch hypothesis: An alter-native explanation of menopause. Human Nature 11:27–42.

m a u s s , m a r c e l . 1924. The gift: Forms and functions ofexchange in archaic societies. London: Cohen and West.

m a y n a r d s m i t h , j o h n . 1991. Honest signalling: The SirPhilip Sidney game. Animal Behaviour 47:1115–20. [lc]

m a y n a r d s m i t h , j o h n , a n d d . g . c . h a r p e r . 1995.Animal signals: Models and terminology. Journal of Theoreti-cal Biology 177:305–11.

———. 2003. Animal signals. Oxford: Oxford University Press.[lc]

m a y r , e r n s t . 1965. “Cause and effect in biology,” in Causeand effect. Edited by Daniel Lerner. Cambridge: MIT Press.

m i l l e r , g e o f f r e y f . 1999. “Sexual selection for culturaldisplays,” in The evolution of culture. Edited by Robin Dun-bar, Chris Knight, and Camilla Power, pp. 71–91. Edinburgh:Edinburgh University Press/New Brunswick: Rutgers Univer-sity Press.

m i l ro y, a l i s o n , l e e c ro n k , l o r n e c a m p b e l l , a n dj e f f ry a . s i m p s o n . 2002. Health and cosmetics: Can fe-male signals of health and beauty be faked by the use of cos-metics? Paper presented at the annual meeting of the HumanBehavior and Evolution Society, Rutgers University, NewBrunswick, N.J. [lc]

n e i m a n , f r a s e r d . 1997. “Conspicuous consumption aswasteful advertising: A Darwinian perspective on spatial pat-terns in Classic Maya terminal monument dates,” in Redis-covering Darwin: Evolutionary theory and archeological expla-nation. Edited by C. Michael Burton, Geoffrey Clark, andDouglas Bamforth, pp. 267–90. Archeological Papers of theAmerican Anthropological Association 7.

n e i m a n , f r a s e r d . , l e s l i e m c f a d e n , a n d d e r e kw h e e l e r . 2000. Archaeological investigation of the Eliza-beth Hemings site (44AB438). Monticello Department of Ar-chaeology Technical Report Series 2. http://www.monticello.org/archaeology/publications/hemings.pdf. [fdn]

o l i v e r , d o u g l a s . 1955. A Solomon Island society. Cam-bridge: Harvard University Press.

o l s o n , m a n c u r . 1965. The logic of collective action: Publicgoods and the theory of groups. Cambridge: Harvard Univer-sity Press.

p e i r c e , c h a r l e s . 1932–58. Collected papers. 8 vols. Cam-bridge: Harvard University Press. [mc]

p o g u e , d e n n i s . 2001. The transformation of America: Geor-gian sensibility, capitalist conspiracy, or consumer revolution?Historical Archaeology 35(2):41–57. [fdn]

p o m i a n k o w s k i , a . , y. i w a s a , a n d s . n e e . 1991. Theevolution of costly mate preferences. 1. Fisher and biased mu-tation. Evolution 45:1422–30.

r a p p a p o r t , ro y a . 1979. Ecology, meaning, and religion.Richmond, Calif.: North Atlantic Books.

r e b y, d . , a n d k . m c c o m b . 2003. Anatomical constraintsgenerate honesty: Acoustic cues to age and weight in the roarsof red deer stags. Animal Behaviour 65:519–30.

ro d s e t h , l . , r . w. w r a n g h a m , a . m . h a r r i g a n ,a n d b . b . s m u t s . 1991. The human community as a pri-mate society. current anthropology 32:221–54. [ehh, ph]

ro w e , c . 1999. Receiver psychology and the evolution of mul-ticomponent signals. Animal Behaviour 58:921–31. [lc]

ry a n , m . j . 1998. Sexual selection, receiver biases, and theevolution of sex differences. Science 281:1999–2003.

s a h l i n s , m a r s h a l l d . 1972. Stone Age economics. Chi-cago: Aldine.

s a l m o n , c a t h e r i n e . 1998. The evocative nature of kin ter-minology in political rhetoric. Politics and the Life Sciences17(1):51–57. [lc]

s c a g l i o n , r i c h a r d . 1999. Yam cycles and timeless time inMelanesia. Ethnology 38:211–25.

s c h e l l i n g , t h o m a s c . 1960. The strategy of conflict. Cam-bridge: Harvard University Press. [lc]

s e b e o k , t h o m a s . 2001. 2d edition. Signs: An introduction tosemiotics. Toronto: University of Toronto Press. [mc]

s e u s s , d r . 1961. The Sneetches and other stories. New York:Random House. [lc]

s i n g h , d . , a n d p . m . b ro n s t a d . 1997. Sex differences inthe anatomical locations of human body scarification and tat-tooing as a function of pathogen prevalence. Evolution andHuman Behavior 18:403–16.

s m a r t , a n n . 1989. The role of pewter as missing artifact:


Consumer attitudes toward tablewares in late eighteenth-cen-tury Virginia. Historical Archaeology 23(2):1–27. [fdn]

———. 1994. “Fashionable sugar dishes, latest fashion wares:The creamware revolution in the eighteenth-century Chesa-peake,” in The historic archaeology of the Chesapeake. Editedby Paul A. Shackel and Barbara J. Little. Washington, D.C.:Smithsonian Institution Press. [fdn]

s m i t h , e r i c a l d e n . 1988. “Risk and uncertainty in the‘original affluent society’: Evolutionary ecology of resourcesharing and land tenure,” in Hunters and gatherers: History,evolution, and social change. Edited by Tim Ingold, DavidRiches, and James Woodburn, pp. 222–25. Oxford: Berg.

s m i t h , e r i c a l d e n , a n d r e b e c c a l . b l i e g e b i r d .2000. Turtle hunting and tombstone opening: Public generosityas costly signaling. Evolution and Human Behavior 21:245–61.

———. n.d. “Costly signaling and cooperative behavior,” in Mo-ral sentiments and material interests: On the foundations ofcooperation in economic life. Edited by S. Bowles, R. Boyd, E.Fehr, and H. Gintis. Cambridge: MIT Press.

s m i t h , e r i c a l d e n , r e b e c c a l . b l i e g e b i r d , a n dd o u g l a s w. b i r d . 2003. The benefits of costly signaling:Meriam turtle hunters. Behavioral Ecology 14:116–26.

s m i t h , e r i c a l d e n , a n d ro b e r t b o y d . 1990. “Riskand reciprocity: Hunter-gatherer socioecology and the problemof collective action,” in Risk and uncertainty in tribal andpeasant economies. Edited by Elizabeth Cashdan, pp. 167–91.Boulder: Westview Press.

s m i t h , e r i c a l d e n , a n d b ru c e w i n t e r h a l d e r . Edi-tors. 1992a. Evolutionary ecology and human behavior. Haw-thorne, N.Y.: Aldine de Gruyter.

———. 1992b. “Natural selection and decision-making: Somefundamental principles,” in Evolutionary ecology and humanbehavior. Edited by E. A. Smith and B. Winterhalder, pp.25–60. Hawthorne, N.Y.: Aldine de Gruyter.

s o s i s , r i c h a r d . 2000. Costly signaling and torch fishing onIfaluk Atoll. Evolution and Human Behavior 21:223–44.

———. 2003. Why aren’t we all Hutterites? Costly signaling the-ory and religious behavior. Human Nature 14:91–127.

s o s i s , r i c h a r d , a n d c a n d a c e a l c o r t a . 2003. Signal-ing, solidarity, and the sacred: The evolution of religious be-havior. Evolutionary Anthropology 12:264–74.

s o s i s , r i c h a r d , a n d e . b r e s s l e r . 2003. Cooperationand commune longevity: A test of the costly signaling theoryof religion. Cross-Cultural Research 37:211–39.

s p e n c e , m i c h a e l . 1973. Job market signaling. QuarterlyJournal of Economics 87:355–74.[lc, ehh, ph]

s t a n n e r , w. e . h . 1966. On Aboriginal religion. Sydney:Oceania Publications.

s u g i y a m a , l a w r e n c e , a n d m i c h e l l e s c a l i s e s u -g i y a m a . 2003. Social roles, prestige, and health risk: Socialniche specialization as a risk-buffering strategy. Human Nature14:165–90.

s w e e n e y, k e v i n . 1994. “High-style vernacular: Lifestyles ofthe colonial elite,” in Of consuming interests: The style of lifein the eighteenth century. Edited by Cary Carson, Ronald Hoff-man, and Peter J. Albert, pp. 1–58. Charlottesville: UniversityPress of Virginia. [fdn]

t h a l e r , r . 1992. The winner’s curse: Paradoxes and anoma-lies of economic life. New York: Free Press. [ma]

t o m a s e l l o , m . 1999. The cultural origins of human cogni-tion. Oxford: Harvard University Press. [ma]

t o n k i n s o n , ro b e r t . 1991. 2d edition. The Mardu Aborigi-nes: Living the dream in Australia’s desert. Fort Worth: Holt,Rinehart and Winston.

t o o b y, j . , a n d l . c o s m i d e s . 1992. “The psychologicalfoundation of culture,” in The adapted mind; Evolutionarypsychology and the generation of culture. Edited by JeromeBarkow and Leda Cosmides, pp. 19–136. Oxford: Oxford Uni-versity Press.

t r i g g , a n d r e w b . 2001. Veblen, Bourdieu, and conspicuousconsumption. Journal of Economic Issues 35:99–115.

t r i g g e r , b ru c e . 1990. Monumental architecture: A thermo-

dynamic explanation of symbolic behavior. World Archaeology22:119–32.

t r i v e r s , ro b e r t l . 1971. The evolution of reciprocal altru-ism. Quarterly Review of Biology 46:35–57.

———. 1974. Parent-offspring conflict. American Zoologist 14:249–64. [lc]

t u r n e r , j o n a t h a n . 1991. The structure of sociological the-ory. Belmont, Calif.: Wadsworth.

t u r n e r , v i c t o r . 1967. The forest of symbols. Ithaca: Cor-nell University Press. [mc]

v a y d a , a n d r e w, a n d b o n n i e m c c a y. 1975. New di-rections in ecology and ecological anthropology. Annual Re-view of Anthropology 4:293–306. [mc]

v e b l e n , t h o r s t e i n . 1994 (1899). The theory of the leisureclass. New York: Dover.

v o l a n d , e c k a r t . 1998. Evolutionary ecology of human re-production. Annual Review of Anthropology 27:347–74.

w a s h b u r n , d . 1999. Perceptual anthropology: The culturalsalience of symmetry. American Anthropologist 101:547–62.[lc]

w a t s o n , p . j . , a n d p . w. a n d r e w s . 2002. Toward a re-vised evolutionary adaptationist analysis of depression: The so-cial navigation hypothesis. Journal of Affective Disorders 72:1–14.

w e i n e r , a n n e t t e . 1976. Women of value, men of renown:New perspectives in Trobriand exchange. Austin: University ofTexas Press.

———. 1988. The Trobrianders of Papua New Guinea. NewYork: Holt, Rinehart and Winston.

w e n z e l , g e o r g e w. , g r e t e h o v e r l s ru d - b ro d a ,a n d n o b u h i ro k i s h i g a m i . Editors. 2000. The socialeconomy of sharing: Resource allocation and modern hunter-gatherers. Senri Ethnological Studies 53. Osaka: National Mu-seum of Ethnology.

w i e s s n e r , p o l l y. 1984. Reconsidering the behavioral basisfor style: A case study among the Kalahari San. Journal of An-thropological Archaeology 3:190–234.

———. 1996. “Leveling the hunter: Constraints on the statusquest in foraging societies,” in Food and the status quest: Aninterdisciplinary perspective. Edited by Polly Wiessner andWulf Schiefenhovel, pp. 171–91. Oxford: Berghahn Books.

———. 2002. Hunting, healing, and hxaro exchange: A long-termperspective on !Kung (Ju/’hoansi) large-game hunting. Evolu-tion and Human Behavior 23:407–36.

w i e s s n e r , p o l l y, a n d w u l f s c h i e f e n h o v e l . Editors.1995. Food and the status quest: An interdisciplinary perspec-tive. Oxford: Berghahn Books.

w i g i n g t o n , t a r a , l o r n e c a m p b e l l , l e e c ro n k ,j e f f ry a . s i m p s o n , a l i s o n m i l ro y, a n d c a ro l l .w i l s o n . 2004. Do women’s faces honestly signal long-termmate quality? Paper presented at the annual meeting of theHuman Behavior and Evolution Society, Free University of Ber-lin, Germany. [lc]

w i l k , r i c h a r d r . 1996. Economies and cultures: Founda-tions of economic anthropology. Boulder: Westview Press.

w i n t e r h a l d e r , b ru c e . 1990. “Open field, common pot:Harvest variability and risk avoidance in agricultural and forag-ing societies,” in Risk and uncertainty in tribal and peasanteconomies. Edited by Elizabeth Cashdan, pp. 67–87. Boulder:Westview Press.

———. 1996. Social foraging and the behavioral ecology of intra-group resource transfers. Evolutionary Anthropology 5:46–57.

w i n t e r h a l d e r , b ru c e , a n d e r i c a l d e n s m i t h .2000. Analyzing adaptive strategies: Human behavioral ecologyat twenty-five. Evolutionary Anthropology 9:51–72.

y o u n g , m i c h a e l . 1971. Fighting with food. Cambridge:Cambridge University Press.

z a h a v i , a m o t z . 1979. Why shouting? American Naturalist113:155–56. [lg]

———. 1993. The fallacy of conventional signaling. PhilosophicalTransactions: Biological Sciences 340:227–30. [lg]

z a h a v i , a m o t z , a n d a v i s h a g z a h a v i . 1997. Thehandicap principle: A missing piece of Darwin’s puzzle. NewYork: Oxford University Press.

Documents

Teoría de señales, interacción estratégica y capital simbólico