Original Article
Costly signaling, ritual and cooperation: evidence from Candomblé, an Afro-Brazilian religion

https://doi.org/10.1016/j.evolhumbehav.2011.11.004Get rights and content

Abstract

The apparent wastefulness of religious ritual represents a puzzle for rational choice theorists and evolutionary scholars. In recent years, it has been proposed that such rituals represent costly signals that promote intragroup cooperation precisely because of the effort and resources they require. This hypothesis was tested over the course of a 14-month long ethnographic study in the northeast of Brazil. The research focused on adherents of Candomblé, an African diasporic religion organized in autonomous congregations primarily located in low-income urban areas. Individuals who reported higher levels of religious commitment behaved more generously in a public goods economic game and revealed more instances of provided and received cooperation within their religious community. This suggests that ritual as a costly signaling may effectively predict willingness to cooperate with other group members and that the signaler may accrue benefits in the form of received cooperation. Socioeconomic variables are also shown to mediate religious signaling. This raises the possibility that signalers strategically alter their expressions of commitment as their needs and circumstances change.

Introduction

Recently, the study of religion has been revitalized by a surge of interdisciplinary work (see Barrett, 2000, Bulbulia et al., 2008, Bulbulia, 2004a, Feierman, 2009, Voland & Schiefenhövel, 2009). This emerging field can be roughly divided into two main areas: cognitive studies and behavior-centered approaches. The former have focused on the acquisition and transmission of religious concepts as by-products of other cognitive capacities (e.g., Atran, 2002, Barrett, 2004, Boyer, 1994, Boyer, 2001, Cohen, 2007, Dawkins, 2006, Dennett, 2007, Guthrie, 1993, Kirkpatrick, 2004, Lawson & McCauley, 1990, McCauley & Lawson, 2002, Shariff and Norenzayan, 2007, Tremlin, 2006, Whitehouse, 2004; but see Bering & Bjorklund, 2004, Bering, 2002), while the latter have been more concerned with adaptationist explanations that center on the relationship between ritual and prosociality (e.g., Alcorta & Sosis, 2005, Bering & Johnson, 2005, Bulbulia, 2004b, Cronk, 1994, Irons, 1996, Irons, 2001, Johnson & Bering, 2006, Sosis, 2003, Sosis & Bressler, 2003, Sosis & Ruffle, 2003, Sosis and Ruffle, 2004, Wilson, 2002; see also Pyysiäinen and Hauser, 2010, Sosis, 2001).

While theory has rapidly advanced and diversified, there is comparatively little empirical research. Behavioral approaches in particular are rich in theoretical material but suffer from a lack of data. In this area, costly signaling has become an important framework for understanding the role of religious ritual. Costly signaling or ‘handicap’ theory emerged in biology (Grafen, 1990, Zahavi & Zahavi, 1997, Zahavi, 1975), although parallel ideas have a long history in the social sciences (Mauss, 2000, Spence, 1973, Veblen, 1994 ). In the context of sexual selection, where it originated, costly signaling refers to ‘wasteful’ traits and behaviors that do not appear to aid survival. Such traits serve as advertisements of high genetic quality precisely because they are expensive to produce and maintain. Since genetic quality cannot be directly assessed, a costly trait or behavior can serve as a signal to females of the condition of possible mates. Because mating interactions have the potential for a conflict of interest between males and females, males may benefit from signaling high quality regardless of their condition. The honesty of the system is maintained because only individuals in truly good condition will be able to bear the costs of the signal without sacrificing other aspects of their fitness. Signalers benefit because their advertisements make them attractive as mates, while receivers stand to gain reliable information about the sender's quality.

While all signals aimed at obtaining a response from a receiver carry inherent costs to the sender, these can be characterized as efficacy and strategic costs (Dawkins, 1993, Guilford & Dawkins, 1991). The former refer to the energetic and opportunity costs endured by the sender to produce a signal that is discernible to a potential receiver and the increased predation risk associated with being more noticeable. Every signal carries some efficacy costs. Costly signals are different because they exceed this minimum threshold and require an additional investment (i.e., strategic cost) from the sender in order to produce an extravagant trait or behavior that is hard to fake. In a given population, individuals will be present in a continuum of lowest to highest quality signalers whose ability to afford strategic costs will result in a range of signal intensities. While these hard-to-fake signals can potentially be faked, the high costs involved it their production will deter free-riders from advertising falsely (this is in contrast to ‘indices’, signals which cannot be faked because they are the direct result of an underlying attribute; Maynard-Smith & Harper, 2003).

The logic of costly signaling theory has been extended to various aspects of human behavior, including cooperation (Gintis, Smith & Bowles, 2001), conspicuous consumption (e.g., Griskevicius et al., 2007, Miller, 2009, Nelissen & Meijers, 2011), hunting (Smith & Bliege-Bird, 2000) and fishing (Sosis, 2000; see also Bliege-Bird & Smith, 2005). In the case of religion, Irons, 1996, Irons, 2001) has proposed that costly rituals may be understood as hard-to-fake signals that reliably indicate commitment to the group. Irons argues that religion represents a “cultural commitment” that binds participants to a set of ideas related to a community of believers. Since group members cannot directly evaluate each other's commitment, rituals that involve extensive time, effort or resources ensure that only those individuals who are truly devoted to the community and its shared ideas (that is, high-quality signalers) will find that benefits of group-belonging outweigh the costs of participation. Irons argues that throughout human evolution, intragroup cooperation represented an important selective advantage. Individuals in groups who were able to work together more efficiently had a better chance of survival than those who did not. Hard-to-fake rituals that advertised commitment to the religious group increased trust among members, facilitating the emergence of cooperation and collective action.

An analogous theory has been advanced by economist Larry Iannaccone, who argues that “stricter” churches are more successful precisely because they weed out free-riders not willing to comply with the sacrifices that are required to be part of the community (Iannaccone, 1992, Iannaccone, 1994). Other theorists have advanced similar formulations (e.g., Alcorta & Sosis, 2005, Atran, 2002, Boyer, 2001, Bulbulia, 2004b, Cronk, 1994, Henrich, 2009, Sosis & Alcorta, 2003, Sosis, 2003) and provided empirical support (e.g., Ruffle & Sosis, 2007, Sosis & Bressler, 2003, Sosis et al., 2007). Like Irons, these researchers argue that ritual increases cooperation by providing an advertisement system that communicates the intensity of members' faith and thus solidarity toward the group. Since group enterprises are vulnerable to free-riding, professions of commitment can be reliable only if they are accompanied by “credible signals of cooperative intentions that are costly-to-fake” (Sosis, 2003, p. 93).

However, in order for ritual (or any other trait or behavior) to fit the characteristics of a costly signal that is evolutionary stable, certain conditions must to be satisfied. Bliege-Bird & Smith (2005; see also Gintis et al., 2001, Grafen, 1990, Johnstone, 1997) advance the following four criteria: one, members of a population vary with regard to an attribute that cannot be directly assessed but can be otherwise signaled; two, it is advantageous for others to obtain reliable information about variation in attribute quality; three, there is a conflict of interest between signalers and receivers so that deception (i.e., signaling higher-than-actual quality) benefits signalers at the expense of receivers; and four, the costs of producing or performing the signal are condition-dependent (low-quality individuals pay higher costs to produce the signal, or, ceteris paribus, the benefits of signaling are greater for high-quality individuals).

Sosis et al. (Sosis et al., 2007, Sosis, 2003, Sosis, 2005) have argued that ritual meets these conditions. One, intensity of religious faith is a variable trait among members of a population that cannot be directly assessed but can be potentially signaled through ritual participation; two, observers benefit from gaining information about this variation because religious faith is tied to group commitment and thus willingness to participate in cooperative exchanges with other group members; three, religious groups offer benefits in the form of intragroup cooperation which can be exploited by nonbelievers (i.e., low-quality signalers) who deceptively advertise religious faith; and four, the opportunity costs of religious participation will be lower for believers (i.e., high-quality signalers) than for nonbelievers because the former have willingly changed or abandoned behaviors in order to comply with the religious doctrine, while the latter have not. Alternatively, if the benefits of group belonging are delayed, only true believers will be willing to engage in ritual participation long enough to receive them (see Sosis et al., 2007).

Thus, the costly signaling theory of religion proposes that costly rituals function as hard-to-fake signals of commitment to the group and the ideas it represents and that such advertisements promote trust and solidarity among group members. While it is not possible to directly measure feelings of commitment or solidarity, it is possible to assess levels of cooperation among group members. Empirical evidence supports the link between hard-to-fake ritual and intragroup cooperation. Sosis and Bressler (2003; see also Sosis, 2000) compared the longevity of religious and secular communes that existed in North America in the 19th and early 20th century. Their analyses show that religious communes lasted longer and were less likely to dissolve than secular ones due to collective action failures (i.e., internal disputes or economic collapse). In addition, religious communes imposed almost twice the number of costly requirements on members than secular ones, and increased number of costly requirements was positively correlated with commune longevity among religious communities. Sosis and Ruffle (2003) further tested the relationship between hard-to-fake ritual and intragroup cooperation among members of religious and secular Israeli kibbutzim. They found that religious males, the only group required to engage in public ritual, cooperated more in an economic game than religious females and secular participants of both genders. Moreover, only males who attended synagogue daily cooperated more than religious females and no differences were found between religious males who did not attend synagogue daily and religious females nor between secular males and females. The authors conclude that cooperation levels are related to frequency of ritual participation rather than to gender.

Although costly signaling has become a central tenet of neo-Darwinian accounts of religion, with the exception of the work of Sosis et al. cited above (Sosis, 200; Sosis & Bressler, 2003, Sosis & Ruffle, 2003), empirical studies are scarce. Here, I present results from a 14-month long ethnographic investigation of costly signaling. The work was carried out in the city of Salvador da Bahia in northeastern Brazil from July 2005 to October 2006 (Salvador and Bahia are both terms used to refer to the city; here, Salvador will be used). The study focused on communities of Candomblé, an Afro-Brazilian religion that originated in the region in the 19th century and which has been the subject of considerable ethnographic work (e.g., Bastide, 2001, Capone, 2004, Landes, 1994, Lima, 2003, Matory, 2005, Prandi, 1996, Rodrigues, 1935, Wafer, 1991). Salvador and its immediate surroundings remain the center of Candomblé, but the religion has spread to other major cities like São Paulo and Rio de Janeiro (see Prandi, 1991).

Section snippets

Research setting

Salvador has a population of almost 3 million, making it the third largest metropolitan area in Brazil. Traces of its past as a main port of entry for the slave trade remain evident in its population, which is approximately 75% Afro-Brazilian. The urban landscape is a mosaic of commercial thoroughfares, luxurious high rises and modest residential neighborhoods, all interspersed with numerous slums and shantytowns. Candomblé religious centers, most commonly called terreiros (other terms are axé,

Hypotheses

Costly rituals can be characterized as behaviors and behavioral restrictions required by the religious group that represent “time, energetic and/or financial costs that are not directed toward accomplishing somatic or reproductive goals efficiently or that limit an individual's ability to achieve these benefits from nongroup members” (Sosis & Bressler, 2003, p. 219). Only those who are truly committed to the group and its shared values will be willing to forgo otherwise pleasant or productive

Methodology

A preliminary database of 55 terreiros was constructed using a combination of random sampling from an existing catalogue (Mott & Cerqueira, 1998) and snowball sampling. This preliminary database was used to understand the variability present in the population of terreiros and to begin to identify potential sites for further research. From this general database, a subsample of 14 terreiros was created for more detailed investigation. To ensure that participating terreiros were active, only those

Results

Two-hundred and fifty-three subjects (138 female, 115 males) from 14 terreiros participated. Additional information on participants is provided in Table 1. The 13 terreiros that provided group-level information ranged from 4 to 37 in years since foundation (mean±S.D.=19.33±10.8). Number of participants at each terreiro ranged from 11 to 30 (mean±S.D.=20.17±6.33). Household income for each participant was measured as seven categories related to monthly minimum wage (R$350 or US$165 at the time).

Discussion

Consistent with a costly signaling interpretation of religious ritual, results show that hard-to-fake expressions of religiosity are related to measures of intragroup cooperation among Candomblé adherents. Controlling for various other variables, higher scores on a religiosity scale specifically designed for Candomblé, the CRSS, significantly predict game offer in an economic game and self-reported acts of cooperation toward other group members. In the case of the game, the amount of variation

Supplementary Materials

The following is the Supplementary data to this article.

scale-soler 2011.

soler 2007.

Acknowledgments

I am especially grateful to Lee Cronk for his long-time support and insightful comments and to Rich Sosis for his extensive help with various versions of this article. I also thank six anonymous reviewers for comments on this and a previous version of the manuscript. Many thanks to members of the Rutgers Evolution, Psychology and Culture seminar for helpful comments and discussions. My special gratitude to the people of Candomblé in Salvador da Bahia for their generosity, patience and

References (85)

  • E.A. Smith et al.

    Turtle hunting and tombstone opening: public generosity as costly signaling

    Evolution and Human Behavior

    (2000)
  • R. Sosis et al.

    Scars for war: evaluating alternative signaling explanations for cross-cultural variance in ritual costs

    Evolution and Human Behavior

    (2007)
  • A. Zahavi

    Mate selection — a selection for a handicap

    Journal of Theoretical Biology

    (1975)
  • C. Alcorta et al.

    Ritual, emotion and sacred symbols: the evolution of religion as an adaptive complex

    Human Nature

    (2005)
  • R. Amaral

    Xirê! O modo de crer e de viver no Candomblé

    (2002)
  • S. Atran

    In Gods We Trust

    (2002)
  • J.L. Barrett

    Why Would Anyone Believe in God?

    (2004)
  • R. Bastide

    [1958] O Candomblé da Bahia

    (2001)
  • J.M. Bering

    Intuitive conceptions of dead agents' minds: the natural foundations of afterlife beliefs as phenomenological boundary

    Journal of Cognition and Culture

    (2002)
  • J.M. Bering et al.

    The natural emergence of reasoning about the afterlife as a developmental regularity

    Developmental Psychology

    (2004)
  • J.M. Bering et al.

    Recursiveness in the cognitive evolution of supernatural agency

    Journal of Cognition and Culture

    (2005)
  • R. Bliege-Bird et al.

    Signaling theory, strategic interaction and symbolic capital

    Current Anthropology

    (2005)
  • P. Boyer

    The Naturalness of Religious Ideas: A Cognitive Theory of Religion

    (1994)
  • P. Boyer

    Religion Explained: The Evolutionary Origins of Religious Development

    (2001)
  • J. Bulbulia

    The cognitive and evolutionary psychology of religion

    Biology and Philosophy

    (2004)
  • J. Bulbulia

    Religious costs as adaptations that signal altruistic intent

    Evolution and Cognition

    (2004)
  • S. Capone

    A Busca da Africa no Candomblé: Tradição e Poder no Brasil

    (2004)
  • ChestnutR.A.

    Competitive spirits: Latin America's new religious economy

    (2003)
  • E. Cohen

    The Mind Possessed: The Cognition of Spirit Possession in an Afro-Brazilian Religious Tradition

    (2007)
  • L. Cronk

    Evolutionary theories of morality and the manipulative use of signals

    Zygon

    (1994)
  • M.S. Dawkins

    Are there general principles of signal design?

    Philosophical Transactions of the Royal Society Series B

    (1993)
  • R. Dawkins

    The God Delusion

    (2006)
  • D.C. Dennett

    Breaking the Spell: Religion as a Natural Phenomenon

    (2007)
  • R.G. Downey et al.

    Missing data in Likert ratings: a comparison of replacement methods

    Journal of General Psychology

    (1998)
  • P. Fry

    Male homosexuality and spirit possession in Brazil

    Journal of Homosexuality

    (1986)
  • V. Griskevicius et al.

    Blatant benevolence and conspicuous consumption: when romantic motives elicit strategic costly signals

    Journal of Personality and Social Psychology

    (2007)
  • M. Gurven

    Does market exposure affect economic behavior? The ultimatum game and public goods game among the Tsimané of Bolivia

  • S.E. Guthrie

    Faces in the Clouds: A New Theory of Religion

    (1993)
  • R.E. Harding

    A Refuge in Thunder: Candomblé and Alternative Spaces of Blackness

    (2000)
  • L.R. Iannaccone

    Sacrifice and stigma: reducing free-riding in cults, communes, and other collectives

    Journal of Political Economy

    (1992)
  • Cited by (0)

    View full text