ArticlePDF Available

Perceptions of moralizing agents and cooperative behavior in Northeastern Brazil


Abstract and Figures

Evolutionary theories suggest that gods of world religions are associated with moralizing qualities and impartial behavior toward co-religionists, and that secular authorities approximate this effect. However, there is a lack of theorizing and experimental studies regarding the influence of local religions on inter-personal conduct. In the current research, we obtained data on beliefs regarding the moralizing qualities of the Christian god, a local god (Ogum), and police in a sample from Northeastern Brazil (n = 193). We used these beliefs as predictors of behavior in Dictator Games where players distributed endowed money between anonymous individuals belonging to local and distant communities. We used subtle reminders of the Christian god, Ogum, and police to investigate their influence on game behavior. The correlational and priming results are mostly in agreement, revealing that: (a) the Christian god is perceived as most moralizing, but (b) has only limited impact on game behavior, while (c) adherence to Ogum is associated with ingroup favoritism, as is (d) priming with secular authority. These results illustrate the differential effects of belief in moralizing and local deities on extended prosociality but show that in specific contexts, secular authorities may emulate the effects of local rather than moralizing deities.
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
Religion, Brain & Behavior
ISSN: (Print) (Online) Journal homepage:
Perceptions of moralizing agents and cooperative
behavior in Northeastern Brazil
Montserrat Soler, Benjamin Grant Purzycki & Martin Lang
To cite this article: Montserrat Soler, Benjamin Grant Purzycki & Martin Lang (2022) Perceptions
of moralizing agents and cooperative behavior in Northeastern Brazil, Religion, Brain & Behavior,
12:1-2, 132-149, DOI: 10.1080/2153599X.2021.2006285
To link to this article:
View supplementary material
Published online: 06 Apr 2022.
Submit your article to this journal
Article views: 49
View related articles
View Crossmark data
Perceptions of moralizing agents and cooperative behavior in
Northeastern Brazil
Montserrat Soler
, Benjamin Grant Purzycki
, and Martin Lang
Ob/Gyn and Womens Health Institute, Cleveland Clinic, Cleveland, OH, USA;
Department of the Study of Religion,
Aarhus University, Aarhus, Denmark;
LEVYNA: Laboratory for the Experimental Research of Religion, Masaryk
University, Brno, Czech Republic
Evolutionary theories suggest that gods of world religions are associated
with moralizing qualities and impartial behavior toward co-religionists,
and that secular authorities approximate this eect. However, there is a
lack of theorizing and experimental studies regarding the inuence of
local religions on inter-personal conduct. In the current research, we
obtained data on beliefs regarding the moralizing qualities of the
Christian god, a local god (Ogum), and police in a sample from
Northeastern Brazil (n= 193). We used these beliefs as predictors of
behavior in Dictator Games where players distributed endowed money
between anonymous individuals belonging to local and distant
communities. We used subtle reminders of the Christian god, Ogum,
and police to investigate their inuence on game behavior. The
correlational and priming results are mostly in agreement, revealing
that: (a) the Christian god is perceived as most moralizing, but (b) has
only limited impact on game behavior, while (c) adherence to Ogum is
associated with ingroup favoritism, as is (d) priming with secular
authority. These results illustrate the dierential eects of belief in
moralizing and local deities on extended prosociality but show that in
specic contexts, secular authorities may emulate the eects of local
rather than moralizing deities.
Received 6 March 2020
Accepted 25 June 2020
Brazil; Candomblé; dictator
game; economic games;
moralizing gods; prosocial
behavior; secular authority
The explicit connection between belief in supernatural agents and human moral behavior is a cen-
tral feature of the worlds major religions. This observation has sparked considerable theoretical and
empirical work to explain how and why humans have developed beliefs in supernatural beings that
punish and reward followers (Johnson, 2015; Johnson & Bering, 2006; Norenzayan, 2013; Noren-
zayan & Shari,2008; Roes & Raymond, 2003; Shari& Norenzayan, 2011). Such beliefs are pre-
dicted to reduce cheating and promote intra-group cooperation in large groups where monitoring
of others is dicult, thereby facilitating the emergence of complex societies (Johnson, 2005; Nor-
enzayan, 2013; Roes & Raymond, 2003). Empirical work has shown that a greater degree of belief
in a moralizing, punishing god is associated with prosocial attitudes and behaviors (Atkinson &
Bourrat, 2011; Johnson et al., 2013; Shari& Norenzayan, 2011). Priming experiments that utilize
moralizing god concepts have obtained similar results, suggesting a causal connection between such
beliefs and prosociality (Ahmed & Hammarstedt, 2011; Gervais & Norenzayan, 2012a; McKay et al.,
2011; Shari& Norenzayan, 2007).
© 2022 Informa UK Limited, trading as Taylor & Francis Group
CONTACT Montserrat Soler,; Martin Lang
Supplemental data for this article can be accessed at
2022, VOL. 12, NOS. 12, 132149
Gods that interfere in human aairs tend to possess hyper-mentalizingfeatures (e.g., omnipre-
sence, omniscience) that go beyond typical abilities to infer othersmental states. However, ethno-
graphic evidence indicates that not all deities display an equal concern with behavior toward
anonymous others. Supernatural agents in some religious systems either lack these abilities or dis-
play no interest in enforcing moral norms (e.g., Chalupa, 2010; Purzycki, 2013; Purzycki & Sosis,
2011; Whitehouse, 1996). Empirical work in such settings has been scarce, but some studies suggest
that trust and cooperation are preferentially directed toward those who demonstrate adherence to a
common group identity through costly commitment signals (Purzycki & Arakchaa, 2013; Soler,
2012); that is, these systems rely on face-to-face communication of their members rather than
on internalized beliefs in moralizing deities. In line with this argument, McNamara et al. (2016)
have shown that primes that invoke local gods (i.e., gods that are closely associated with circum-
scribed populations and do not explicitly care about moral issues) can also increase parochialism
in experimental economic games. Similarly, cross-cultural studies using such games suggest that
belief in local godsmoralizing abilities does not predict greater generosity toward distant coreligio-
nists, as belief in punitive and omniscience deities does (Lang et al., 2019; Purzycki et al., 2016).
A corollary of this line of reasoning is related to the degree of secularity observed in Western
countries, where the last few decades have witnessed a dramatic decline in religiosity (Zuckerman,
2007). A classic interpretation of this trend is that religious belief and ritual behavior intensify when
there is uncertainty or danger (Barber, 2011; Gmelch, 1974; Malinowski, 1948; Sosis & Handwerker,
2011). Uncertainty may also motivate individuals to turn to civil institutions as a means of re-assert-
ing control over their circumstances (Kay et al., 2008,2010). Thus, the monitoring role of moraliz-
ing gods may be substituted by ecacious judicial and criminal systems that enforce social norms
among strangers (Gervais & Norenzayan, 2012b; Shari& Norenzayan, 2007). Indeed, priming
subjects with secular reminders such as policeand juryhas been shown to have a similar
eect on increasing cooperation on economic game behavior as moralizing god primes (Shari
& Norenzayan, 2007).
Notably, previous research reporting a relationship between beliefs in gods and how they corre-
spond to the breadth of generosity or honesty has been limited to variation of specically moralistic
deities where the primary contrast is whether or not such gods punish and are omniscient, not
whether non-moralistic gods have similar eects. In fact, precedent studies either do not specify
recipientsreligiosity (McNamara et al., 2016) or restrict recipients to the moralistic deity tradition
or religious outgroups (Lang et al., 2019; Purzycki et al., 2016). In other words, participants who
have engaged in these experiments made economic decisions between themselves, local or distant
co-religionists of the moralistic tradition, and members of religious and/or ethnic outgroups. Reci-
pients have not been dened as subscribing to local traditions. Here, our focus supplements this
research by focusing on co-religionists of more locally-bound beliefs.
We investigate the interaction between local-god communities among a population in North-
eastern Brazil that practices Candomblé, a religion of the African Diaspora. While Candomblé
has a strong presence in the area, the majority of the population practices Catholicism or, increas-
ingly, Evangelical and Pentecostal Christianity. The gods of Candomblé are characterized as non-
moralizing, but most adherents see no conict in professing a concurrent belief in the Christian
god. In addition, decades of economic and political turmoil in the country have resulted in mistrust
of civil institutions (Power & Taylor, 2011). Thus, this population is embedded in multiple layers of
religious belief and moral authority, making it an ideal setting to assess the relationship between
religious beliefs, secular authority, and moral behavior. Using ethnographic interviews, economic
games, and object primes, we assess the impact of religiosity and secular authority on prosociality.
The Site: Cachoeira. The Recôncavo Baiano extends from the littoral around the city of Salvador
da Bahia to surrounding regions which were once the center of the sugar cane industry. The town of
Cachoeira (pop. 15, 000), where the research was conducted, sits about 75 miles inland. Up until the
1800s, it held an important strategic position linking wealthy sugar plantations to the coast through
the Paraguassú River. Today it is better known for its charming but decaying Baroque architecture
which makes it a popular day-trip destination for tourists. Thus, residents have modest but steady
sources of wage labor through daily tour buses and seasonal festivals that attract large numbers of
locals from surrounding areas. Other widespread economic activities include small-scale com-
merce, day labor, and domestic work. While the town has a small number of prominent families,
most study participants were low-income and worked in the informal sector. Given its history as a
central hub of the countrys enslaved labor force, Cachoeira is also well-known as a center of Afro-
Brazilian culture. Candomblé houses or terreiros are present in practically every neighborhood.
Thus, even non-devotee residents have some familiarity with the religion.
Fifteen years ago, Cachoeira was a sleepy maze of narrow cobble-stone streets. Since then,
global and local events have ushered in noticeable changes. Although the Northeast is one of
the poorest regions of the country, Brazils economic boom (20102013) brought a period of
prosperity. The commercial area in the city center grew, a large outdoor market was rehabi-
litated, and the volume of cars and motorcycles signicantly increased. The use of cell phones
and social media are now widespread. The opening of a regional branch of the federal univer-
sity has attracted students from all over Northeast Brazil. However, not all changes have been
positive. The inux of students has meant that prices for housing and commodities around the
town have risen, pricing out locals. Whereas anything other than petty crime was practically
unheard of in the recent past, drug-related violence is on the rise (two young men were mur-
dered during Solers most recent two-week stay in 2015). More dramatically, the spectacular
crash of the Brazilian economy in 2014 had dire consequences. By August 2015, when the pre-
sent study was conducted, Brazilian currency had lost a third of its value in the course of eight
months. Many study participants mentioned recent job losses and held a bleak outlook of their
economic future.
The political fallout from the countrys economic crisis was not yet evident at the time of the
research, but stability has never been a hallmark of Brazilian institutions. Corruption is endemic
and distrust of a government traditionally dominated by oligarchs is widespread (Power & Taylor,
2011). For low-income citizens, the most immediate representative of state authority is the police.
There are various law-enforcement bodies in the country but the best-known is the Polícia Militar
or PM. An ancillary of the military, the PM is ostensibly responsible for street patrols and public
safety at the local level. However, the violent exploits of the organization in the slums or favelas
of large cities are well-documented (Ahnen, 2007; Caldeira, 2002). Raids, retaliatory violence,
and extra-judicial executions, particularly against poor Afro-Brazilians, are routine practices.
Thus, as a representation of secular authority, the PM is unlikely to evoke positive attitudes or sub-
stitute the eects of a moralizing god.
Brazils Religious Landscape. Brazil is home to a complex mix of Christian faiths, indigenous
traditions, and various religions of African origin. Among the latter, the best-known is Yoruban-
inuenced Candomblé, which shares beliefs and rituals with other religions of the African Diaspora.
Candomblé centers on the cultivation of axé, the life-force of the universe. Although there is a dis-
tant creator god (Olorun or Olódùmarè), worship is primarily directed at a group of deities called
orixás (also santos or saints) and various other supernatural entities unique to Brazil. Each orixá is
associated with a specic realm of life, such as contagious disease (Obaluâe) or sacred herbs and
plants (Oxossí). Adherents must nd out which orixá is their dono da cabeça (owner or master
of their head), which makes them the symbolic son or daughter of that deity. Sons and daughters
share the personality traits of their dono da cabeça, rely on that orixá for supernatural intervention,
and must abide by the specic taboos and rituals associated with that entity (e.g., sexual abstinence
on certain days, food restrictions, and specicoerings). Thus, an orixá that is particularly impor-
tant for one person, may be signicantly less so for another.
Candomblé worshippers are organized around autonomous groups called terreiros (other terms
are axé, roça, or casa). The terreiro represents a physical space for worship as well as a community of
followers led by a charismatic priest or priestess (pai or mãe-de-santo). Although each leader may
set his or her own norms, there are sucient commonalities across terreiros to recognize adherents
as belonging to the povo-de-santo (literally, people of the saintwhich refers to all followers of
orixá-based religions).
Candomblé beliefs and rituals are concerned with attaining happiness, health, and prosperity in
this life rather than accumulating good deeds or securing rewards in the afterlife. This pragmatism
has led to descriptions of the religion as amoral(Amaral, 2002; Chesnut, 2003; Prandi, 1991).
More recently, there has been resistance to this characterization and a recognition of an inherent
ethical system centered on the sacred responsibilities (obrigações) of each adherent, which are con-
tingent on their role within the hierarchy of the terreiro and their relationship with the orixá that
ownstheir head (Berkenbrock, 2017; Rabelo, 2016). In addition, Candomblé adherents do not live
in an isolated bubble; they are fully integrated in broader Brazilian society. Thus, the degree to
which ethical or moral dictates emerge from Candomblé or from the surrounding socio-cultural
environment is dicult to ascertain.
For much of its history, Candomblé existed in the shadow of Roman Catholicism. Perhaps as a
cover for their beliefs, Candomblé adherents at some point adopted some aspects of Catholicism.
Orixás are associated with corresponding Catholic saints (e.g., Iansá with Santa Barbara, Iemanjá
with the Virgin Mary) and some feasts correspond with important dates in the Catholic calendar.
There are even Catholic festivals where Candomblé followers play a central role. The feast of Our
Lady of the Good Death in Cachoeira, for example, involves processions and a Catholic mass, but is
organized by a confraternity of women who are Candomblé adherents. Thus, there is little conict
between being a Candomblé follower and believing in the Christian god or other deities. A compli-
cating factor is the growing presence of Evangelical and Pentecostal churches. Evangélicos, as fol-
lowers of these religions are called in Brazil, tend to subscribe to a conservative version of Protestant
Christianity. There is active hostility from some of these churches towards Afro-Brazilian religions,
which are often portrayed as superstitious, dangerous, or evil. Despite this, most Evangélicos are
familiar with Candomblé, may have family members who are Candomblé adherents, or even con-
verted from Candomblé into their current religion. Thus, a notable feature of this site is the reli-
gious heterogeneity of the population and the familiarity of any given person with various
General Predictions. If the gods of world religions are indeed more moralistic, punitive, and
knowledgeable of human behavior, and reminders of such deities can harness their general
eects of extending prosociality beyond the connes of family and ones local community, then
exposure of the Christian god should be more associated with prosociality beyond local commu-
nities (i.e., less parochial game behavior). While we generally expect people to favor anyone closer
to themselves (i.e., individual > local religious community > distant co-religious community > dis-
tant religious outgroup), when individuals entertain the belief in a punitive, omniscient deity that
cares about morality, they should be more generous toward geographically distant co-religionists
and religious outgroup members than those who do not endorse such beliefs. In other words,
such deities should bring more investment to others beyond participantsimmediate lives. How-
ever, as Candomblé deities are locally important spirits, they should be associated with increased
parochialism in game behavior; when reminded of such deities, people should be more willing to
favor their local communities at the expense of co-religionists who live elsewhere and religious out-
group members. Moreover, while secular judicial systems and institutions might generally function
in a similar way to the moralistic gods of the world religions, given the aforementioned climate of
police corruption in Brazil, we posit that the police will be associated with the lowest expression of
prosociality towards others in this context. We tested these predictions using a dictator game and
four priming conditions.
Materials and methods
The study took place in January 2015 and August 2015. Data collection included a preliminary
phase to collect data for ethnographic interviews. Study participants played four dictator games
in four randomly assigned treatment conditions that ostensibly should have allowed us to calculate
the eect of religious and secular primes on prosocial behavior. Players responded to ethnographic
interviews before and after the games.
Preliminary Data Collection. Preliminary questionnaires were conducted to determine which
deities were widely known and which moral traits were viewed as important at the study site.
These data were collected at a public plaza with 20 locals that did not take part in the nal
study. Findings were used to construct ethnographic interviews for the study. Each of the intervie-
wees was asked to list and rank ve spirits or deities that you think people here believe in. By far
the most commonly cited moralizing god was the Christian God (Deus in Portuguese). Choosing a
local god presented more diculties. Several orixás received an equally high ranking and number of
mentions. In addition, some orixás are connected to notions of fairness or the afterlife (e.g., Xangô,
Nanã), some are believed to lead people to bad outcomes, including death (i.e., Exú), and some are
widely syncretized with Catholic gures (e.g., Iemanjá). Taking these issues into consideration,
Ogum, a male orixá syncretized as St. George, was chosen to represent the local god. Among the
at least 12 orixás recognized in Candomblé, Ogum is one the best-known and most widely wor-
shipped. Ogums origin story identies him as a fearsome warrior, and as such he is associated
with ghts, struggles, and conquering adversity. Qualities linked with Ogum are courage, arrogance,
anger, and determination, but there is nothing inherently moralizing about him (i.e., these qualities
may be channeled for whatever purpose suits a devotee). Interviews also included a free-list of ve
qualities that make someone a bad or immoral person. We used these data for validation checks
(see Results section).
Recruitment. Participants were 56 males and 128 females (n= 193; 9 participants have missing
data for sex). All participants spoke Portuguese and the vast majority reported living in the town
where the study was conducted (99.5%). Descriptive variables of the sample are shown in Table
1. The goal was to recruit members of the povo-de-santo, that is, devoted Candomblé followers,
in order to explore extended prosociality among followers of a local god. However, responses to
questions about engagement with Deus and Ogum suggest lower levels of adherence to Candomblé
than expected. These questions were How often do you think about Deus/Ogum?and How often
do you perform activities or practices to talk to or appease Deus/Ogum?and were answered on a
ve-point frequency scale; all interview questions are available online at All
participants except two reported thinking daily about Deus and 85% of the sample reported daily
performance of activities to talk to or appease Deus. In contrast, 61% reported thinking about
Ogum at least a few times a year (24% did so daily), while 55% reported performing activities to
talk to or appease Ogum a few times a year (16% did so daily). That is, despite recruitment through
personal visits at local terreiros to maximize participation of devoted adherents, we sampled partici-
pants with varying commitment to Ogum, from fully initiated members to many who simply enjoy
attending feasts and ceremonies. To limit collusion, potential subjects were invited to attend a
Table 1. Sample descriptive variables.
Mean SD Range
Age 34.8 12.3 18 to 75
Yrs of formal education 8.6 4.1 0 to 20
Number of children 1.9 1.9 0 to 10
Household size 4.1 2.0 1 to 17
Yrs lived in a large urban area
3.7 7.4 0 to 36
Co-religionist closeness 1.5 1.0 1 to 5
Outgroup closeness 1.6 1.1 1 to 5
Police evaluation 0.3 1.1 2to2
Freq. thinking about Deus 4.0 0.3 0 to 4
Freq. thinking about Ogum 1.7 1.7 0 to 4
Freq. ritual Deus 3.8 0.7 0 to 4
Freq. ritual Ogum 1.4 1.5 0 to 4
Pop. >100,000.
morning or afternoon session at a centrally located community center over the course of eight days.
However, Cachoeira is a small town and word travels fast. As people found out from friends and
neighbors that study participants earned cash, it is likely that not all attendees considered them-
selves as belonging to the povo-de-santo. One woman, for example, did not complete the study
because she identied as evangélica and felt uncomfortable giving money in the games to a Can-
domblé adherent. To control for the possibly varying attachment to Candomlé (co-religionists)
and Evangelical (outgroup) communities, we included in our statistical models questions about
emotional closeness to co-religionists and outgroups (see Table 1).
Ethnographic Interviews. There were four sets of interviews used in the study: a demographic
questionnaire, a material uncertainty scale, an income and household wealth questionnaire, and
a religiosity/moral values interview which included questions about the Christian moralizing god
(Deus), the local god (Ogum), and the secular authority (the police). Only data from the demo-
graphic and religiosity/moral values interviews are presented here. Questions from the religios-
ity/moral values interview were answerable as either yesor no(coded as dummy variables)
or through a 5-point frequency scale (0 = never, 1 = once a year, 2 = once a month, 3 = once a
week, 4 = every day or several times a day). The basic predictor variables pertaining to a gods pun-
ishment and omniscience were constructed from four binary questions asking whether a god (1)
can punish people for their behavior; (2) inuence what happens after death; (3) know what people
think and feel; and (4) see what people are doing in São Paulo. Answers to these questions were
averaged, creating a Punishment-Monitoring scores for Deus and Ogum.
Economic Games. To measure prosocial behavior, we employed modied versions of the dicta-
tor game (DG). In its simplest form, the DG consists of a player who receives an economic allot-
ment to split between herself and an anonymous recipient in any manner she chooses. The
amount allocated by the player to the recipient is a baseline measure of prosociality. Here, each sub-
ject played four games in random order: local coreligionist vs. distant coreligionist, themselves vs.
distant coreligionist, distant coreligionist vs. outgroup member, and themeselves vs. outgroup
member. Recipient categories were described to subjects as an individual from the povo-de-
santo in Cachoeira(local-coreligionist member), an individual from the povo-de-santo in Salva-
dor(distant co-religionist), and an Evangelical person in Salvador(outgroup member). In other
words, recipients were identied as either local or distant Candomblé followers, and the outgroup
members were geographically distant Evangelicals. We used the nearby city of Salvador as a refer-
ence because it has many Candomblé terreiros and it is large enough to hold many anonymous co-
religionists. Subjects were provided with $10 Brazilian Reais (BR) (approximately $2.85 USD at the
time) in coins of $1BR for each round of the DG. Players were instructed to divide a $10BR during
each game in whatever manner they chose using labeled pairs of covered cups with a slit to place
coins. It was explained that coins in the selfcup would be taken home by the player at the end of
the game and that coins in the other cups would be given to individuals belonging to the recipient
categories. Prior to the beginning of play, the game was described in detail to individual participants
and test questions were used to ensure full understanding of the procedure. Upon nishing the data
collection, coins allocated to local and distant co-religionists and to outgroups were donated to the
respective communities.
Object Primes. Subjects were randomly assigned to one of four possible conditions: a control,
moralizing god, local god, and secular authority. In the control condition, subjects played the econ-
omic game on an empty table. In the other three conditions, object primes were placed on one side
of the table where they were clearly visible. The moralizing god prime was a Christian Bible, the
local god prime was a blue bead necklace easily recognizable as representing Ogum, and the secular
prime consisted of a cap used by the military police with a prominent seal representing the
General Procedure. All parts of the experiment were carried out at a local community center
which had several rooms with doors that aorded necessary privacy. Participants waited in a class-
room as they arrived. Groups of four or ve were then led into another room where the lead author
explained that the task involved an economic game and a series of questionnaires. Subjects were
assured that all their choices in the game would be condential and that they were free to stop
the activity at any moment they chose without forfeiting a show-up fee of $10 BR. After providing
consent, each subject lled out an identication card with basic contact information that randomly
assigned each person to one of the four possible prime treatments and a card with 4 digits that rep-
resented the order in which the games would be played (also distributed randomly). Subjects were
then individually led by a eld assistant to one of two rooms where they could play the games in
complete privacy. Each of these rooms had a table and chair facing a blank wall. Object primes
were placed on one side of the table. All pictures and posters were either removed from the
walls or covered with white sheets of paper. A eld assistant provided the subject with the rst
set of cups and coins, explained the game in detail, ran sample runs of the game, and asked testing
questions to ensure a full understanding of the procedure. Assistants then left the room, closed the
door, and waited outside. After each game, subjects alerted the assistant who then provided the next
pair of cups while the lead author picked up the cups with the allocated coins inside. This was
repeated until the subject completed the four games. Following the games, subjects moved to an
outside area where two eld assistants conducted the demographic and religiosity interviews. In
the meantime, the lead author recorded each game oer and organized pay-outs behind closed
doors in another room. After the rst interviews were over, each subject entered this room
where the lead author conducted a household wealth interview and a market integration interview.
Following this, subjects were thanked and provided cash pay-outs in an envelope. All methods, pro-
cedures, and data collection forms were reviewed and approved by the Institutional Review Board
of Montclair State University.
Predictions.The cup dyads follow most of our general predictions (see above). We expect a gen-
eral pattern of allocations benetting players, followed by their local communities, followed by dis-
tant co-religionist communities, with distant outgroup members receiving the least (i.e., S>L>D
>O). However, getting primed with a Bible or expressing higher ratings of Gods omniscience and
punishment should increase allocations to distant co-religionists in the local vs. distant co-religio-
nist game and in the self vs. distant co-religionist game, as well as increase allocations to the reli-
gious outgroups in the distant co-religionist vs. outgroup game, and in the self vs. outgroup game
(Prediction #1). This prediction reects the fact that while co-religionist cups were designated to
Ogum followers, these followers are mostly members of the Catholic church; hence, belief in the
Catholic god (Deus) should positively aect allocations to Evangelicals (outgroups) who believe
in the same god. In contrast, getting primed with the Ogum prime or expressing higher ratings
of Ogums omniscience and punishment should boost parochialism such that participants will pre-
fer the self and local co-religionist cups (Prediction #2). In other words, belief in and priming with
Ogum should aect cup preferences in the opposite direction than belief in and priming with Deus.
We should see a similar pattern in the police condition; if the police are generally held to be a hostile
and ineective secular institution, we should see people withholding resources from relatively more
socially distant players (Prediction #3).
Analysis. We performed all analyses in R, version 3.5.3 (R Core Team, 2019). For each of the
four games, we predicted the number of coins in the less parochial cups: a distant co-religionist
cup in the games that pitted self and local co-religionist against distant co-religionist; and an out-
group cup in the games that pitted self and distant co-religionist against outgroup. Since the Dic-
tator Game data were censored at 0, we used the Tobit regression (McDonald & Mott, 1980),
which proved to t the data better than a negative binomial regression that could also account
for the data-censoring at 0. Data were analyzed using the tobit function from the AER package
(Kleiber & Zeileis, 2008) and plots were created using the package ggplot2 (Wickham, 2016).
We divide the main analyses into two partsmodeling the relationship between belief in omnis-
cient and punitive aspects of Deus and Ogum, and allocations in the Dictator Game; and testing the
eects of priming with Deus, Ogum, police, and control primes on Dictator Game allocations. That
is, models in the rst part tested Predictions 13 using self-reported beliefs in Deus, Ogum, and
opinions about police, and models in the second part tested Predictions 13 using the priming
materials. In the rst set of models, we rst included the predictor variables mapping the monitor-
ing and punitive characteristic of Deus and Ogum. In the second step, we adjusted these models for
demographic variables pertaining to age, sex, years of formal education, the number of children,
household size, years lived in a city, and for emotional closeness to co-religionists and outgroups.
In the priming analyses, we rst compared the dierences between priming with Deus and the other
primes, and in the second step adjusted these eects for the same demographic variables as in the
correlational analyses.
To investigate the dierential eects of belief in Deus and Ogum on behavior in the Dictator Game,
we rst assessed whether participants believed that Deus was indeed more moralizing compared to
Ogum and the police. Subsequently, we tested whether this purported dierence in moralization
predicted game behavior, and, nally, tested whether priming with dierent god concepts aected
game behavior.
Preliminary Analyses. By way of a validation check, we rst assessed participantsattributions to
the Christian god, a Candomblé deity, and the police in order to demonstrate that the Christian god
is the most morally concerned and knowledgeable of the three agents. As expected, responses
demonstrate that Deus is ascribed with both moralizing concerns and punishment-monitoring abil-
ities to a higher degree than Ogum (Figure 1, top row). While the dierences between Deus and
Ogum were relatively subtle for punishing, the belief in Deusmonitoring abilities were twice as
high compared to Ogum. In fact, these beliefs reached ceiling levels. Furthermore, there was no per-
son in the whole sample who answered Noto all of the four questions related to Deus. Figure 1,
middle row, also revealed stark dierence between Deus and Ogum in rewarding abilities: the for-
mer again reaches ceiling levels while the latter hovers around the mean of the reward question.
Recall that in the preliminary interviews, we asked participants about what constitutes a bad or
immoral person. The most frequently mentioned items were deceitfulness, envy, wickedness, alco-
hol and drug use, and stealing. These items were translated based on Solers own ethnographic
experience (see Soler, 2012,2016) and include synonymous terms. Thus, frequently mentioned
variations on the word falso (fake or phony) included falsedade (deceitfulness or phoniness) or
even mentiroso (liar) or mentira (lie), but only falsedade was included in the nal ethnographic
interviews. Alcohol use is frequent in this population, but alcohol and drug use (rather than
abuse) were ranked high as moral failings. Envidia and colloquial expressions such as olho grande
(literally, large eyewhich refers to coveting what others have) were translated as envy. Maldoso
refers to someone who is wicked, mean, or cold-hearted. Finally, ladrão and roubar were subsumed
under stealing. The term ladrão translates literally as thief,but is used more broadly to refer to
young men involved in crime, particularly drug-tracking. The usage is similar to thugor hood-
lumin the United States and can reect socio-economic and racial prejudice. Drawing from these
data, we asked participants how much Deus, Ogum, and police cared about these four immoral
behaviors (last row in Figure 1). Deus was perceived as most concerned with all moral items.
Ogum was believed to be more concerned than police with deceitfulness, envy, and wickedness,
but less concerned about the use of alcohol and drugs. Together, these summary statistics show
that Deus was indeed perceived as a relatively more moralizing deity compared to Ogum and police.
Supernatural Beliefs and Game Behavior. We analyzed four permutations of the Dictator Game
where participants divided money between themselves, local co-religionists, distant co-religionists,
and outgroups. In the games that included cups for players themselves, the modal contributions to
the distant co-religionists or outgroup players was 0. In other words, participants strongly preferred
to allocate all money to themselves. However, when allocating between two groups based on their
geographical and social proximity, the modal allocation was 5; participants often opted for impartial
divisions. Since the games containing the Self cuprevealed more selsh behaviors, belief in Deus
had more space to drive allocations closer to the impartial division. Indeed, for both games where
participants were pitted against distant co-religionists and outgroups, we observed a substantial
positive eect of belief in Deuspunitive and monitoring abilities on allocations to the less parochial
cup, albeit the 95% CI crossed zero (supporting Prediction #1). Believing in all four aspects of Deus
monitoring and punishing increased the estimated positive allocations to distant co-religionists on
average by 1.8 coins in the self vs. distant game, and by 2.1 coins in the self vs. outgroup game (see
Table 2 for untransformed coecients and 95% CIs and Figure 2).
On the contrary, belief in Ogum
had no stable eect on allocations in the games comprising the Self cup, and Prediction #2 was not
Interestingly, this trend was reversed for games where local co-religionists were pitted against
distant co-religionists and where distant co-religionists were pitted against outgroups. While belief
in Deusmonitoring and punitive abilities had variable eects, belief in Ogums abilities predicted
Figure 1. Moralizing concerns and punishment-monitoring abilities of Deus, Ogum, and the police (means with SE). The rst row
displays the percentage of yes answers to binary questions pertaining to gods punishment and monitoring abilities. The second
row displays gods interest in punishing various misconduct and rewarding good behaviors. The third row displays perceived
concerns with various behaviors for Deus, Ogum, and the police (see R code for plot regarding stealing).
Table 2. Beta estimates with 95% CI from Tobit models for deity punishment-monitoring abilities predicting coin allocations to most geographically/socially distant recipient.
Local vs. Distant Self vs. Distant Distant vs. Outgroup Self vs. Outgroup
Deus 0.59 0.78 2.07 2.06 0.57 1.75 3.65 2.51
(2.63, 1.44) (1.50, 3.07) (0.28, 4.42) (0.45, 4.56) (1.58, 2.73) (0.76, 4.27) (1.04, 6.03) (0.36, 5.39)
Ogum 0.81 0.84 0.32 0.24 2.22 1.91 0.04 0.53
(1.59, 0.04) (1.73, 0.05) (0.57, 1.20) (0.73, 1.21) (3.04, 1.40) (2.89, 0.93) (0.98, 0.90) (1.63, 0.58)
Age 0.05 0.09 0.03 0.31
(0.61, 0.51) (0.52, 0.71) (0.64, 0.58) (1.01, 0.40)
Sex (1 = male) 0.78 0.13 0.85 0.27
(1.63, 0.06) (1.05, 0.79) (1.79, 0.09) (1.33, 0.79)
Education 0.02 0.14 0.13 0.004
(0.37, 0.41) (0.57, 0.30) (0.56, 0.30) (0.49, 0.49)
Children 0.003 0.03 0.06 0.03
(0.26, 0.27) (0.33, 0.26) (0.24, 0.35) (0.37, 0.31)
Household 0.30 0.19 0.09 0.17
(0.58, 0.01) (0.50, 0.13) (0.39, 0.22) (0.52, 0.17)
Material 0.17 0.21 0.34 0.39
(0.65, 0.32) (0.74, 0.32) (0.88, 0.20) (0.22, 1.01)
Years in city 0.06 0.11 0.03 0.04
(0.01, 0.11) (0.06, 0.16) (0.02, 0.08) (0.01, 0.10)
Police eval 0.29 0.01 0.28 0.04
(0.68, 0.10) (0.44, 0.42) (0.71, 0.15) (0.45, 0.53)
Corel close 0.01 0.14 0.34 0.1
(0.42, 0.40) (0.59, 0.30) (0.80, 0.11) (0.62, 0.42)
Outgroup close 0.35 0.29 0.21 0.31
(0.003, 0.71) (0.10, 0.69) (0.18, 0.61) (0.14, 0.75)
Intercept 3.95 3.59 0.60 1.02 5.27 4.70 0.34 1.08
(2.31, 5.58) (1.27, 5.91) (1.30, 2.49) (1.52, 3.57) (3.53, 7.00) (2.17, 7.23) (2.35, 1.67) (1.79, 3.95)
N177 128 180 130 180 130 178 129
Note: Coecients refer to less parochial cup allocations. The Deus and Ogum variables represent punishment-monitoring scores for the respective deity. Age, education, and material insecurity are
centered at their mean. Police eval is evaluation of police. Corel and outgroup emo is emotional closeness to co-religionists and outgroup. The dierent number of participants between games
reect mistakes in the game behavior of individual participants. Note that intercept is the intercept of latent uncensored variable.
increased parochiality in both games (see Table 2 for untransformed coecients and 95% CIs and
Figure 2). Specically, reporting that Ogum is both punitive and monitoring was associated with 0.8
fewer coins contributed to the Distant co-religionist cupin the local vs. distant game and with 1.9
fewer coins in the Outgroup cupin the distant vs. outgroup game (supporting Prediction #2).
Positive evaluation of police had similar trends as belief in Ogum, that is, toward more parochialism
in the local vs. distant and distant vs. outgroup games, but these eects were generally small and
imprecise, suggesting ambiguous support for Prediction #3.
Regarding the demographic variables used to adjust the main coecients of interest, there were
no substantial eects that would be precisely estimated across all four games. For instance, while
males were on average more parochial than females in all games, this eect was precisely estimated
only in the distant vs. outgroup game. Similarly, household size predicted more parochialism, but
these eects were weak and imprecise. The only precisely estimated eects were for years lived in a
city, which positively predicted contributions to less parochial cups in all four games (see Table 2).
Priming Analyses. Participants were randomly assigned to one of four treatments: control (N=
45), moralizing god prime (Christian Bible) (N= 41), local god prime (Ogum beads) (N= 60), and
secular prime (police hat) (N= 47). We run the same Tobit model as in the analyses of supernatural
beliefs above, only substituting variables related to belief in Deusand Ogums monitoring-punitive
abilities with the treatment variable.
Figure 2. Regression estimates from Tobit models. Note that our sample did not contain any participant who would answer No
to all four questions about Deuspunitive and monitoring abilities, hence the regression lines for Deus belief start at 0.25.
As stated in Prediction #1, the Deus prime was associated with largest contributions to the less
parochial cups across all four permutations of the Dictator Game (see Figure 3). However, the
dierence from the control condition was on average quite low, ranging from 0.4 to 0.1 coins
across all model specications and imprecisely estimated (see Table 3 for untransformed coe-
cients and 95% CIs). More stable dierences were revealed in comparison between the Deus
and Ogum primes with dierences ranging from 0.2 to 0.9 coins across all models (providing lim-
ited support to Prediction #2). The largest dierences were observed for the comparison between
Deus and police primes: the estimated dierences ranged from 0.1 to 1.9 coins across the four
Dictator Games (supporting Prediction #3). These results suggest that while the Deus prime pro-
duced limited and small positive eects (compared to the control prime), the Ogum and police
primes exerted larger negative eects on allocations to less parochial cups (except for the distant
vs. outgroup games, where the dierences between Deus, Ogum, and police prime were the
Figure 3. Raw means with SE for the eects of primes on Dictator Game allocations. Priming with Deus produced largest allo-
cations in all four games.
Table 3. Beta estimates with 95% CI from Tobit models for priming eects on coin allocations to most geographically/socially distant recipient.
Local vs. Distant Self vs. Distant Distant vs. Outgroup Self vs. Outgroup
Control prime 0.63 0.43 0.29 0.11 0.12 0.01 0.62 0.03
(1.59, 0.33) (1.46, 0.61) (1.43, 0.85) (1.05, 1.28) (1.24, 1.00) (1.25, 1.23) (1.82, 0.57) (1.29, 1.36)
Ogum prime 1.17 0.96 0.69 0.75 1.04 0.19 1.18 1.14
(2.08, 0.27) (1.87, 0.04) (1.76, 0.38) (1.78, 0.28) (2.10, 0.02) (1.29, 0.90) (2.29, 0.07) (2.32, 0.03)
Police prime 2.22 2.04 1.40 1.50 0.99 0.15 2.14 1.99
(3.18, 1.26) (3.09, 0.99) (2.54, 0.26) (2.69, 0.32) (2.11, 0.12) (1.39, 1.10) (3.34, 0.95) (3.36, 0.61)
Age 0.01 0.2 0.22 0.3
(0.50, 0.52) (0.37, 0.78) (0.82, 0.38) (0.96, 0.37)
Sex (1 = male) 0.21 0.24 0.62 0.24
(0.99, 0.57) (0.63, 1.10) (1.54, 0.30) (0.76, 1.24)
Education 0.01 0.15 0.28 0.07
(0.37, 0.35) (0.56, 0.26) (0.71, 0.15) (0.54, 0.40)
Children 0.01 0.03 0.05 0.08
(0.24, 0.26) (0.25, 0.32) (0.25, 0.34) (0.25, 0.41)
Household 0.23 0.2 0.02 0.21
(0.48, 0.01) (0.48, 0.08) (0.31, 0.26) (0.52, 0.11)
Material 0.25 0.15 0.43 0.35
(0.65, 0.16) (0.60, 0.30) (0.91, 0.05) (0.17, 0.88)
Years in city 0.06 0.08 0.03 0.02
(0.02, 0.10) (0.03, 0.13) (0.02, 0.09) (0.03, 0.08)
Corel close 0.14 0.01 0.66 0.18
(0.50, 0.21) (0.39, 0.41) (1.09, 0.24) (0.65, 0.29)
Outgroup close 0.23 0.12 0.32 0.2
(0.10, 0.56) (0.25, 0.49) (0.08, 0.72) (0.23, 0.62)
Intercept 4.20 4.60 2.98 3.24 5.34 5.41 3.39 3.74
(3.51, 4.89) (3.15, 6.04) (2.16, 3.80) (1.60, 4.88) (4.53, 6.15) (3.68, 7.14) (2.54, 4.24) (1.86, 5.61)
N190 142 193 144 193 144 191 143
Note: Intercept reects allocation to Deus prime condition. Deus and Ogum are punishment-monitoring scores for the respective deity. Age, education, and material insecurity are centered at their
mean. Sex is females vs. males. The dierent number of participants between games reect mistakes in the game behavior of individual participants.
The descriptive results were consistent with the hypotheses that moralizing gods are believed to
possess higher mentalizing abilities and are more concerned with morality. Deus was universally
believed to possess omniscient abilities, while these were ascribed to Ogum by less than half of
the sample. Deus was also believed to possess greater ability to interfere in human aairs than
Ogum, such as punishing and inuencing the afterlife. These dierences may reect broader dier-
ences between religions associated with texts that preserve the uniformity of belief, like Christianity,
and those that rely on oral transmission, like Candomblé, where theological orthodoxy is more
likely to vary according to individual levels of adherence. Dierences were also found in the per-
ceived concern of deities and secular authorities for moral behaviors. Deus was believed to be
more concerned about moral failings than both Ogum and the police. Item-by-item comparisons
between Ogum and the police, however, reveal suggestive dierences between moral sentiments
and moral behaviors. While Ogum cares more about the former (e.g., deceitfulness), the police
were believed to be more concerned with the latter (e.g., alcohol and drug abuse and stealing).
Most participants had no diculty understanding the games after the rst explanation and one
or two sample rounds. As explained in the Methods section, eld assistants demonstrated additional
sample rounds to those who still seemed unclear and gave them an opportunity to ask their question
until it was clear that they had full comprehension of the rules. During delivery of nal pay-outs and
in conversation with the lead author, several individuals mentioned that they found the interviews
interesting and described their understanding of the games as an experiment related to generosity
and religion. Game behavior was within the typical range observed in other studies (e.g., Engel,
2011), with generally more selsh and parochial choices across games. The predicted eects of belief
in Deuspunitive and omniscience abilities on positive allocations to less parochial cups (Prediction
#1) were detected only in the games that included the Self cup. Exploring Figure 2 revealed that
while these eects are quite substantial (2 coins more in the less parochial cups), they likely reect
aversion toward total selshness (putting 0 coins in the less parochial cups) observed for partici-
pants that expressed the lowest rating of Deuspunitive and omniscient abilities. In other words,
the moralizing aspects of Deus were associated with allocating at least some amount to distant
co-religionists and outgroups, rather than impartially dividing the endowed money. On the
other hand, in the local vs. distant and distant vs. outgroup games where allocations were generally
close to the impartial split, the ratings of Deusabilities did not rise the allocations beyond the fair
split, as should be expected. Self-reported devotion to Ogum, on the other hand, acted on that varia-
bility of allocations in the local vs. distant and distant vs. outgroup games (Prediction #2) exactly
because there was room for making the oers more parochial (while the allocations in the self-
games were already at 0).
The priming results do not support the hypothesis that moralizing gods decrease selshness and
increase generosity to strangers. While players exposed to the moralizing god prime behaved less
selshly than those in the control condition, these dierences were only small and imprecise. How-
ever, we found substantial dierences between the Deus prime and Ogum primes, which had oppo-
site eects on allocations, mostly driven by the Ogum prime. As in the correlational analysis of
belief in omniscience and punishment, the priming results suggested that priming participants
with Ogum reminders increased parochiality of their allocations in all games except for the distant
vs. outgroup game (Prediction #2). This result complements McNamara et al.s(2016)nding that
local god primes in Fiji increase ingroup favoritism under some conditions and is consistent with
previous similar work in Brazil (Cohen et al., 2018).
Even stronger results in this regard were observed for the comparison of Deus and police primes
(Prediction #3). Exposure to the secular prime, a police hat, increased parochial behavior in all
games except the distant vs. outgroup game. While the perception of the police was marginally posi-
tive in our sample (see Table 1), this positive view was associated with more parochial allocations
(see Table 2). Likewise, the police prime was associated with more parochial oers. These ndings
challenge the idea that secular authority functions similarly to moralizing deities (e.g., Kay et al.,
2010; Shari& Norenzayan, 2007). This may be the case in wealthy Western countries where secu-
lar institutions are functional and respected, but the situation in Brazil is more complex. In exten-
sive ethnographic interviews with low-income residents of Salvador, Paes-Machado and Noronha
(2002) describe attitudes toward the police as ambivalentand contradictory. They point out that
the urban poor in Brazil are equally familiar with drug-related crime and the brutal tactics of the
police. In smaller towns like Cachoeira, residents may be familiar with narratives surrounding
police violence, but their actual experience of police authority may be quite dierent. Until recently,
Cachoeira experienced very little serious crime, and the population is small enough that members of
the local police are personally known to many residents. These factors may account for the para-
doxical ndings of positive perceptions of secular authority and less prosocial game behavior in
the presence of a police prime.
Most demographic variables had no eect on game behavior. Women were overrepresented in the
sample, but this gender imbalance is reective of a similar trend in Candomblé terreiros where
females make up the bulk of membership. More years living in a large city were associated with
increased prosociality and a decrease in ingroup favoritism. Salvador was most frequently men-
tioned as the large city where participants had lived and this was also used as the location of the
distant coreligionist and outgroup member in the games. It is possible that this triggered a sense
of familiarity in some players that, in turn, led to more generous oers. An alternative interpretation
is that the experiences with anonymous strangers that are common in large urban centers decrease
parochialism, as other studies have reported (Henrich, Ensminger et al., 2010).
Much of the literature surrounding the cognitive and evolutionary origins of religious belief has
been developed around Abrahamic traditions (see Johnson, 2015; Norenzayan, 2013). As is true of
most experimental psychological work, socioeconomically diverse and non-Western populations
have been underrepresented (Henrich, Heine et al., 2010). The present study centers on a non-mor-
alizing religion in a setting that shares many characteristics with the Western world. Brazil is one of
the largest democracies in the world and the research site is a small city that is fully integrated in the
global market economy. Despite these similarities many subjects reported a simultaneous adher-
ence to a moralizing god and a local god that forms a distinct religious tradition, and have opposite
eects on extended prosociality. Thus, the postulate that belief in moralizing gods is instrumental in
promoting intra-group cooperation may be complicated by the presence of a parallel belief system
with dierent network structure and universalizing tendencies. For future work, it would be impor-
tant to disentangle moralizing abilities of local gods from their broader role as signiers of group
identity. Here, it is dicult to discern if Ogum promotes ingroup favoritism because of the mor-
alizing qualities associated with this deity or because any orixá is bound to trigger favorable behav-
ior toward other Candomblé followers. Furthermore, future work should also focus on better
sampling of Candomblé followers, which proved dicult in the current project. Indeed, close to
50% of participants did not express adherence to the local god that was featured in the interviews
and games. This low number is likely due in part to the large Candomblé pantheon. Since each
devotee is considered the son or daughter of a specic orixá, those who belongto other orixás
are not expected to consider Ogum as equally important. In addition, some orixás are closely linked
to ideas of justice (e.g., Xangô) while others are perceived as rather sinister (e.g., Exú). While our
religious landscape interview suggested that Ogum is one of the most salient local gods, using
another Candomblé deity would be a good test of generalizability of the current results.
Finally, the role of secular institutions in regulating moral behavior merits further investigation.
In some ways, the police elicit similar reactions to the local god (i.e., both increase ingroup favorit-
ism). It is unclear what, if anything, links the police and local god in a way that predicts similar
behaviors in the Dictator Game. Perceptions of the police in this sample are paradoxical and appear
to encompass the fear and longingfor a strong control system described in other studies (Cal-
deira, 2002; Paes-Machado & Noronha, 2002). Relationships between Candomblé followers and
deities are characterized by a similar combination of service provision, a sense of obligation and
appreciation, and respect tinged with fear. A pai-de-santo in Salvador explained this in the following
We have a day, every month, so adepts and initiates of the house come to take care of what is theirs. How?
Doing their oerings and every year paying their obligation to the orixá. And if its an urgent case, they
make some oerings to obtain success in what they want in life. Because orixá has a positive side, but it
also has a negative side. I do know what is the positive side of the axé, [but] of the negative side of the axé,
to this day I dont know and I dont want to know.
Although this is a speculative explanation, it is worth considering parallels between secular insti-
tutions and local religious beliefs in non-Western settings as a novel avenue of research.
1. To ease the interpretation, we present marginal eects estimated by the Tobit model for non-zero values. The
original Tobit estimates for the uncensored latent variable (presented in Tables 2 and 3) were multiplied by an
adjustment factor, which takes into account the scale parameter computed across the mean values of predictor
variables in the model (see Wooldridge, 2016 and also the Supplementary R code for more details).
Disclosure statement
No potential conict of interest was reported by the author(s).
We acknowledge funding from a research grant, The Emergence of Prosocial Religionsfrom the John Templeton
Foundation, and the Cultural Evolution of Religion Research Consortium (CERC), funded by a generous partnership
grant (895-2011-1009) from the Social Sciences and Humanities Research Council of Canada.
Data availability statement
The data set and associated R code used to produce the reported results are freely available at
Montserrat Soler
Benjamin Grant Purzycki
Martin Lang
Ahmed, A., & Hammarstedt, M. (2011). The eect of subtle religious representations on cooperation. International
Journal of Social Economics,38(11), 900910.
Ahnen, R. E. (2007). The politics of police violence in democratic Brazil. Latin American Politics and Society,49(1),
Amaral, R. (2002). Xirê!: o Modo de Crer e de Viver no Candomblé. Pallas/EDUC.
Atkinson, Q. D., & Bourrat, P. (2011). Beliefs about God, the afterlife and morality support the role of supernatural
policing in human cooperation. Evolution and Human Behavior,32(1), 4149.
Barber, N. (2011). A cross-national test of the uncertainty hypothesis of religious belief. Cross-Cultural Research,45
(3), 318333.
Berkenbrock, V. J. (2017). O conceito de ética no candomblé. HORIZONTE-Revista de Estudos de Teologia e Ciências
da Religião, 905928.
Caldeira, T. P. (2002). The paradox of police violence in democratic Brazil. Ethnography,3(3), 235263. https://doi.
Chalupa, A. (2010). Why did Greeks and Romans prayed aloud? Anthropomorphism, dumb gods and human cogni-
tion. Chasing Down Religion: In the Sights of History and the Cognitive Sciences,15,8195.
Chesnut, R. A. (2003). Competitive spirits: Latin Americas new religious economy. Oxford University Press.
Cohen, E., Baimel, A., & Purzycki, B. G. (2018). Religiosity and resource allocation in Marajó, Brazil. Religion, Brain
and Behavior,8(2), 168184.
Engel, C. (2011). Dictator games: A meta study. Experimental Economics,14(4), 583610.
Gervais, W. M., & Norenzayan, A. (2012a). Like a camera in the sky? Thinking about God increases public self-aware-
ness and socially desirable responding. Journal of Experimental Social Psychology,48(1), 298302.
Gervais, W. M., & Norenzayan, A. (2012b). Reminders of secular authority reduce believersdistrust of atheists.
Psychological Science,
Gmelch, G. J. (1971). Baseball magic. Trans-action,8, 3941.
Henrich, J., Ensminger, J., McElreath, R., Barr, A., Barrett, C., Bolyanatz, A., Cardenas, J. C., Gurven, M., Gwako, E.,
Henrich, N., Lesorogol, C., Marlowe, F., Tracer, D., & Ziker, J. (2010). Markets, religion, community size, and the
evolution of fairness and punishment. Science,327(5972), 14801484.
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences,33
(2-3), 6183.
Johnson, D. (2015). God is watching you: How the fear of God makes us human. Oxford University Press.
Johnson, D., & Bering, J. (2006). Hand of God, mind of man: Punishment and cognition in the evolution of
cooperation. Evolutionary Psychology,4(1), 219233.
Johnson, D. D. P. (2005). Gods punishment and public goods. Human Nature,16(4), 410446.
Johnson, K. A., Li, Y. J., Cohen, A. B., & Okun, M. A. (2013). Friends in high places: The inuence of authoritarian
and benevolent god-concepts on social attitudes and behaviors. Psychology of Religion and Spirituality,5(1), 1522.
Kay, A. C., Gaucher, D., Napier, J. L., Callan, M. J., & Laurin, K. (2008). God and the government: Testing a com-
pensatory control mechanism for the support of external systems. Journal of Personality and Social Psychology,95
(1), 1835.
Kay, A. C., Shepherd, S., Blatz, C. W., Chua, S. N., & Galinsky, A. D. (2010). For God (or) country: The hydraulic
relation between government instability and belief in religious sources of control. Journal of Personality and
Social Psychology,99(5), 725739.
Kleiber, C., & Zeileis, A. (2008). Applied econometrics with R. Springer-Verlag.
Lang, M., Purzycki, B. G., Apicella, C. L., Atkinson, Q., Bolyanatz, A., Cohen, E., Klocová, E. K., Handley, C.,
Lesorogol, C., Mathew, S., McNamara, R., Moya, C., Placek, C., Soler, M., Vardy, T., Weigel, J., Willard, A.,
Xygalatas, D., Norenzayan, A., & Henrich, J. (2019). Moralizing gods, impartiality and religious parochialism
across 15 societies. Proceedings of the Royal Society B: Biological Sciences,286(1898), 20190202.
Malinowski, B. (1948). Magic, science and religion. Doubleday.
McDonald, J. F., & Mott, R. A. (1980). The uses of Tobit analysis. The Review of Econoics and Statistics,62(2), 318
McKay, R., Eerson, C., Whitehouse, H., & Fehr, E. (2011). Wrath of God: Religious primes and punishment.
Proceedings of the Royal Society of London B: Biological Sciences,278(1713), 18581863.
McNamara, R. A., Norenzayan, A., & Henrich, J. (January 2016). Supernatural punishment, in-group biases, and
material insecurity: Experiments and ethnography from Yasawa, Fiji. Religion, Brain & Behavior,6(1), 3455.
Norenzayan, A. (2013). Big gods: How religion transformed cooperation and conict. Princeton University Press.
Norenzayan, A., & Shari,A.F.(2008). The origin and evolution of religious prosociality. Science,322(5898), 5862.
Paes-Machado, E., & Noronha, C. V. (2002). Policing the Brazilian poor: Resistance to and acceptance of police bru-
tality in urban popular classes (Salvador, Brazil). International Criminal Justice Review,12(1), 5376. https://doi.
Power, T. J., & Taylor, M. M. (Eds.). (2011). Corruption and democracy in Brazil: The struggle for accountability.
University of Notre Dame Press.
Prandi, R. (1991). Segredos guardados: Orixás na alma brasileira. Companhia Das Letras.
Purzycki, B. G. (2013). The minds of gods: A comparative study of supernatural agency. Cognition,129(1), 163179.
Purzycki, B. G., Apicella, C., Atkinson, Q. D., Cohen, E., McNamara, R. A., Willard, A. K., Xygalatas, D., Norenzayan,
A., & Henrich, J. (2016). Moralistic gods, supernatural punishment and the expansion of human sociality. Nature,
530(7590), 327330.
Purzycki, B. G., & Arakchaa, T. (2013). Ritual behavior and trust in the Tyva Republic. Current Anthropology,54(3),
Purzycki, B. G., & Sosis, R. (2011). Our gods: Variation in supernatural minds. In U. J. Frey (Ed.), Essential building
blocks of human nature (pp. 7793). Springer-Verlag.
Rabelo, M. (2016). Considerações sobre a ética no candomblé. Revista de Antropologia,59(2), 109130. https://doi.
R Core Team. (2019). R: A language and environment for statistical computing. R Foundation for Statistical
Roes, F. L., & Raymond, M. (2003). Belief in moralizing gods. Evolution and Human Behavior,24(2), 126135.
Shari, A. F., & Norenzayan, A. (2007). God is watching you priming God concepts increases prosocial behavior in an
anonymous economic game. Psychological Science,18(9), 803809.
Shari, A. F., & Norenzayan, A. (2011). Mean gods make good people: Dierent views of God predict cheating behav-
ior. International Journal for the Psychology of Religion,21(2), 8596.
Soler, M. (2012). Costly signaling, ritual and cooperation: Evidence from candomblé, an Afro-Brazilian religion.
Evolution and Human Behavior,33(4), 346356.
Soler, M. (2016). The faith of sacrice: Leadership trade-osin an Afro-Brazilian religion. Human Nature,27(4),
Sosis, R., & Handwerker, W. P. (2011). Psalms and coping with uncertainty: Religious Israeli womens responses to
the 2006 Lebanon War. American Anthropologist,113(1), 4055.
Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. New York: Springer-Verlag.
Whitehouse, H. (1996). Apparitions, orations, and rings: Experience of spirits in Dadul. In J. Major & A. Howard
(Eds.), Spirits in culture, history and mind (pp. 173194). Routledge.
Wooldridge, J. M. (2016). Introductory econometrics: A modern approach. Thomson South-Western (6th ed.).
Zuckerman, P. (2007). Atheism: Contemporary numbers and patterns. In M. Martin (Ed.), The Cambridge compa-
nion to atheism (pp. 4765). Cambridge University Press.
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
The emergence of large-scale cooperation during the Holocene remains a central problem in the evolutionary literature. One hypothesis points to culturally evolved beliefs in punishing, interventionist gods that facilitate the extension of cooperative behaviour toward geographically distant co-religionists. Furthermore, another hypothesis points to such mechanisms being constrained to the religious ingroup, possibly at the expense of religious outgroups. To test these hypotheses, we administered two behavioural experiments and a set of interviews to a sample of 2228 participants from 15 diverse populations. These populations included foragers, pastoralists, horticulturalists, and wage labourers, practicing Buddhism, Christianity, and Hinduism, but also forms of animism and ancestor worship. Using the Random Allocation Game (RAG) and the Dictator Game (DG) in which individuals allocated money between themselves, local and geographically distant co-religionists, and religious outgroups, we found that higher ratings of gods as monitoring and punishing predicted decreased local favouritism (RAGs) and increased resource-sharing with distant co-religionists (DGs). The effects of punishing and monitoring gods on outgroup allocations revealed between-site variability, suggesting that in the absence of intergroup hostility, moralizing gods may be implicated in cooperative behaviour toward outgroups. These results provide support for the hypothesis that beliefs in monitoring and punitive gods help expand the circle of sustainable social interaction, and open questions about the treatment of religious outgroups.
Full-text available
A partir da descrição de eventos desenrolados em terreiros de Salvador o texto visa contribuir para uma reflexão acerca do modo próprio do candomblé colocar e resolver questões éticas, formar e oferecer condições para o exercício de uma sensibilidade ética no dia a dia do terreiro. Ao abordar casos relativos à iniciação e formação gradativa de adeptos e ao desenrolar de relações entre pessoas humanas e orixás, pretende-se discutir tanto alguns dos elementos que definem o agir ético no candomblé, quanto caracterizar a montagem ética que sustenta esse agir.
Full-text available
Despite secular trends in some countries, prestige-based authority in the form of religious leadership remains hugely influential in the everyday lives of millions of people around the world. Here, the costs and benefits of religious leadership are explored in an urban setting in northeastern Brazil. An economic game, within-group cooperation questionnaires, and social network analyses were carried out among adherents of an Afro-Brazilian religion. Results reveal that leaders display high levels of religious commitment and disproportionally provide cooperative services to group members. On the other hand, initiates cooperate less than leaders but do not differ in levels of received cooperation or social cohesion measures. This may indicate some level of exploitation or free-riding. Demographic and group variables also appear to play an important role in the degree of social cohesion a group achieves. These findings are discussed in the context of non-Western urban settings where religious leadership may represent both an alternative to social advancement and a crucial source of material aid, social support, and a strong sense of community.
Full-text available
Since the origins of agriculture, the scale of human cooperation and societal complexity has dramatically expanded. This fact challenges standard evolutionary explanations of prosociality because well-studied mechanisms of cooperation based on genetic relatedness, reciprocity and partner choice falter as people increasingly engage in fleeting transactions with genetically unrelated strangers in large anonymous groups. To explain this rapid expansion of prosociality, researchers have proposed several mechanisms. Here we focus on one key hypothesis: cognitive representations of gods as increasingly knowledgeable and punitive, and who sanction violators of interpersonal social norms, foster and sustain the expansion of cooperation, trust and fairness towards co-religionist strangers. We tested this hypothesis using extensive ethnographic interviews and two behavioural games designed to measure impartial rule-following among people (n = 591, observations = 35,400) from eight diverse communities from around the world: (1) inland Tanna, Vanuatu; (2) coastal Tanna, Vanuatu; (3) Yasawa, Fiji; (4) Lovu, Fiji; (5) Pesqueiro, Brazil; (6) Pointe aux Piments, Mauritius; (7) the Tyva Republic (Siberia), Russia; and (8) Hadzaland, Tanzania. Participants reported adherence to a wide array of world religious traditions including Christianity, Hinduism and Buddhism, as well as notably diverse local traditions, including animism and ancestor worship. Holding a range of relevant variables constant, the higher participants rated their moralistic gods as punitive and knowledgeable about human thoughts and actions, the more coins they allocated to geographically distant co-religionist strangers relative to both themselves and local co-religionists. Our results support the hypothesis that beliefs in moralistic, punitive and knowing gods increase impartial behaviour towards distant co-religionists, and therefore can contribute to the expansion of prosociality.
Este texto irá apresentar os pressupostos para se pensar o conceito de ética na religião do Candomblé e de sua visão de mundo. Na visão de mundo do Candomblé, a existência é dividida em dois níveis: o Orum e o Aiyê. A realidade finita, individualizada e histórica é a existência Aiyê. Mas esta é uma existência que tem sua origem no Orum, a realidade permanente. No Orum tudo existe sempre. Quem administra as existências individualizadas são as forças chamadas de Orixás. Cada ser humano é filho ou filha de um Orixá. A ética no Candomblé se constrói a partir da relação entre o adepto e o seu Orixá. Não há assim no Candomblé uma ética de valores ou de princípios. A ética no Candomblé é uma ética relacional. Ela depende da relação de cada pessoa com o seu Orixá e da manifestação deste. Depende também do estágio de iniciação do fiel, pois cada estágio gera diferentes obrigações. A relação do adepto com o seu Orixá tem por objetivo alcançar uma existência em equilíbrio, numa busca que é constante e dinâmica.
This new edition to the classic book by ggplot2 creator Hadley Wickham highlights compatibility with knitr and RStudio. ggplot2 is a data visualization package for R that helps users create data graphics, including those that are multi-layered, with ease. With ggplot2, it's easy to: • produce handsome, publication-quality plots with automatic legends created from the plot specification • superimpose multiple layers (points, lines, maps, tiles, box plots) from different data sources with automatically adjusted common scales • add customizable smoothers that use powerful modeling capabilities of R, such as loess, linear models, generalized additive models, and robust regression • save any ggplot2 plot (or part thereof) for later modification or reuse • create custom themes that capture in-house or journal style requirements and that can easily be applied to multiple plots • approach a graph from a visual perspective, thinking about how each component of the data is represented on the final plot This book will be useful to everyone who has struggled with displaying data in an informative and attractive way. Some basic knowledge of R is necessary (e.g., importing data into R). ggplot2 is a mini-language specifically tailored for producing graphics, and you'll learn everything you need in the book. After reading this book you'll be able to produce graphics customized precisely for your problems, and you'll find it easy to get graphics out of your head and on to the screen or page. New to this edition:< • Brings the book up-to-date with ggplot2 1.0, including major updates to the theme system • New scales, stats and geoms added throughout • Additional practice exercises • A revised introduction that focuses on ggplot() instead of qplot() • Updated chapters on data and modeling using tidyr, dplyr and broom
Current research suggests that certain features of religion can harness our sociality in important ways, curbing selfish behavior and/or boosting prosocial behavior. If this is the case, embodied symbols of religious devotion should induce these effects. To test the claim that religious symbolism has an effect on sociality, we conducted the Random Allocation Game with a symbolic prime in Pesqueiro, on the island of Marajó, Brazil, among Christians. Our prime – a Bible and a crucifix pendant – appears to have influenced the allocations made toward distant co-religionists; people who played the game in the prime condition allocated more coins to the distant co-religionist. Additionally, self-reported beliefs about God’s knowledge and punishment had strong effects on fair gameplay across games.