Body counts in lowland South American violence

Answer each question based from the readings, need to include examples from the readings for each question. A bibliography need not be submitted with your responses, though should include author name(s) and year of publication when referring to a paper (e.g., “Walker and Bailey (2013) showed…”). Responses should be typed, double-spaced in 12-point, Times New Roman font, and not exceed 1.5 pages per question. You will be assessed on your understanding of the topics addressed as well as how cogently you present the information

Body counts in lowland South American violence
Robert S. Walker a,
⁎, Drew H. Bailey b
a Department of Anthropology, University of Missouri, Columbia, Missouri
b Department of Psychology, Carnegie Mellon University, Pittsburgh, Pennsylvania
article info abstract
Article history:
Initial receipt 3 April 2012
Final revision received 25 August 2012
Keywords:
Violence
Warfare
Intergroup conflict
Revenge killings
Multi-level selection
Amazonia
Violence was likely often a strong selective pressure in many traditional lowland South American societies. A
compilation of 11 anthropological studies reporting cause of death shows that violence led to about 30% of
adult deaths, of which about 70% were males. Here violent deaths are further itemized at the level of
ethnographically-reported death events (particular duels, homicides, and raids) to provide more detailed
insight into the causes and consequences of within- and between-group violence. Data for 238 death events
(totaling 1145 deaths) from 44 lowland South American societies show that attacks are more deadly when
treachery is used, when avenging a previous killing, and on external warfare raids between ethnolinguistic
groups. That revenge raids kill more people on average than the original grievance, at least when conflicts are
between ethnolinguistic groups, indicates a tendency towards increasingly vicious cycles of revenge killings.
Motives of killings as noted in ethnographic sources, in order of importance, reportedly include revenge for
previous killings and other wrong-doings like sorcery, jealousy over women, gain of captive women and
children, fear or deterrence of impending attack, and occasionally the theft of material goods. Results may
have implications for understanding the potential for multi-level selection by delineating the force of
competition at varying scales of analysis within and between lowland South American societies.
© 2013 Elsevier Inc. All rights reserved.
1. Introduction
Amazonian anthropology has long been central to understanding
the nature of warfare in pre-state human societies. The classic raiding
experience in Amazonia (glossed loosely here as tribal warfare) was
for a group of men to attack an enemy village at dawn, kill several
enemies, and quickly retreat into the forest to avoid a counter-attack
(Chagnon, 1968; Larrick, Yost, Kaplan, King, & Mayhall, 1979; Ross,
1988; Conklin, 1989; Verswijver, 1992; Fausto, 2001; Beckerman &
Yost, 2007; Beckerman & Valentine, 2008). The “Great Protein Debate”
crystallized an argument between those that saw tribal warfare
among the Yanomamo and other Amazonian societies as population
control in response to low-density protein availability (Harris, 1974;
Gross, 1975) versus those that argued for plentiful protein availability
(Beckerman, 1979) with warfare as a strategy for status striving and
capturing wives (Chagnon & Hames, 1979). Yet another argument
saw warfare as more novel, driven primarily by competition for
European goods like machetes and shotguns (Ferguson, 1995). The
perspective taken here is that warfare was traditionally a persistent
feature of many tribal societies in lowland South America well before
1500AD and even more paramount in chiefdoms that collapsed
during the early stages of European colonization (Carneiro, 1981;
Hemming, 1978; Redmond, 1994; Balée, 2007).
Comparative ethnographic information on actual death events
may help shed light on the proximate and ultimate causes of violence.
By “death event” we refer to activity that led to violent death(s) as
recorded ethnographically such as a particular duel, homicide, or
warfare raid. Event-level analyses may provide some information for
evaluating two recent evolutionary models of intergroup aggression.
One of these, the “chimpanzee model” (Wrangham, 1999; Wrangham
& Glowacki, 2012), derives from repeated observations of coalitions of
wild chimpanzee males killing members of neighboring communities
when there is a local imbalance of power so that the killing(s) can be
carried out relatively safely for the aggressors. The resultant fitness
benefits may take the form of increased access to more land, food, and
females and even the eventual replacement of the neighboring group
(Wilson, Wallauer, & Pusey, 2004). There are several resemblances
between chimpanzee and tribal warfare, including male coalitions,
group territoriality, low cost but lethal intergroup killings, and similar
fitness benefits, making the chimpanzee model potentially applicable
to warfare in small-scale human societies (Manson & Wrangham,
1991; Wrangham & Peterson, 1996). Wrangham and Glowacki (2012)
have found these chimpanzee–human similarities to hold for human
hunter–gatherers and we extend this analysis here to lowland South
American societies.
Another model, “parochial altruism” (Bowles, 2006, 2009; Choi &
Bowles, 2007), sees widespread cooperation in human societies as the
Evolution and Human Behavior 34 (2013) 29–34
⁎ Corresponding author.
E-mail addresses: walkerro@missouri.edu (R.S. Walker), drewhalbailey@cmu.edu
(D.H. Bailey).
1090-5138/$ – see front matter © 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.evolhumbehav.2012.08.003
Contents lists available at SciVerse ScienceDirect
Evolution and Human Behavior
journal homepage: www.ehbonline.org
result of genetic group (multi-level) selection where within-group
cooperation allows some altruistic groups to better displace or
otherwise out-compete other more self-serving groups of individuals
(Darwin, 1871; Alexander, 1974; Hamilton, 1975; Wilson & Dugatkin,
1997). Parochial altruism resembles the chimpanzee model in that
aggressive male behaviors have been selectively favored through the
success of more dominant groups. However, it further suggests
selection for uniquely human psychological traits adapted for withingroup cooperation and between-group warfare not seen in chimpanzees (Wrangham & Glowacki, 2012), potentially including selfsacrificing behaviors (Bowles & Gintis, 2011), strong reciprocity
(Gintis, 2000), treachery (Wadley, 2003), cultures of honor (Nisbett &
Cohen, 1996), and revenge-seeking (Beckerman & Valentine,
2008; Boehm, 2011). Here we evaluate evidence for these derived
psychological mechanisms in lowland South American warfare. In
general, the most salient levels of selection and the proximate and
ultimate motivations of warfare are underexplored, and that is why
we focus here on quantifying the intensity of competition at multiple
scales of social organization within and between lowland South
American societies.
2. Methods
Ethnographic literature for lowland South America was searched
for event-level information of individual homicides, duels, and raids
by looking through ethnographies for relevant index entries of
“violence”, “war”, “killing”, “raids”, and “homicide”. Of these ethnographic works, 11 include data on the total number of deaths
attributed to violence (1281, not including infanticide but including
violent deaths by non-indigenous perpetrators) as a fraction of total
recorded deaths (4215 deaths, Table 1). The event-level sample
includes 1145 deaths of victims from 44 different ethnolinguistic
groups (see compiled dataset and sources at http://dice.
missouri.edu). Only indigenous-on-indigenous killings were included
as death events, while deaths accorded to strife with European
colonists were excluded. The inclusion criterion for death events is
that the source must mention the body count for a particular event.
Additional information recorded includes whether captive women or
children were taken, number of women taken captive, motivation(s)
for killing (e.g., revenge, gain, fear or deterrence), scale of conflict
(within community, internal between communities in the same
ethnolinguistic group, or external between ethnolinguistic groups),
use of treachery (e.g., feigning friendship such as holding a malicious
feast or inviting enemies to hunt or fish), and sequence of conflict
when an initial killing is recognized as the impetus for a resulting
revenge killing. Motives of raids were recorded as stated in
ethnographic sources and sometimes included multiple motives
from one or more informants.
There are a number of potential sampling issues associated with
the recording of death events. Non-lethal interactions (body count=
0) were not included in this study, and this makes within-village body
counts appear higher per event than they actually are in comparison
to internal and external warfare because many within-village
altercations are not lethal while warfare events usually involve
fatalities. Also, it is possible that informants and ethnographers are
more likely to remember and mention events that are more deadly.
However, the modal body count at all scales of analysis is simply 1
death, and so if there are biases towards larger body counts the effect
does not appear to be serious. In addition, we were forced to assume
that, if captives were not mentioned in an event, then no captives
were taken. This may under-estimate the frequency of captives and
the number of women taken captive in the sample. Likewise, if no use
of treachery was mentioned, we assumed no treachery for that event.
However, this is a conservative choice and only serves to diminish the
real differences in body counts between treacherous and nontreacherous events, although there were likely unsuccessful attempts
at treachery that were unable to entice victims and therefore did not
enter into the sample.
Generalized linear mixed models were run in the lme4 package(Bates & Maechler, 2010) in R (Ihaka & Gentleman, 1996) with
each incident cross-nested in victim’s society and perpetrator’s
society because different societies have variable numbers of death
events (random intercepts model). The dependent variable was
number of victims in an event (body counts). These are count data so a
Poisson link was used. For models using sequence to predict body
counts, incidents were nested within conflict groups (i.e., we used a
grouping variable to indicate which set of conflicts a conflict belonged
to), which were nested in perpetrator’s society (cross-nested models
containing sequence did not converge; however, model results did
not differ whether conflict groups were nested in perpetrator’s group
or victim’s group).
3. Results
3.1. Violent deaths
Cause-of-death data focusing on violent deaths in pre-contact or
more traditional time periods are available for 11 lowland South
American societies (Table 1). Violent deaths include both indigenous
and non-indigenous conflict in the form of duels, homicides, and raids,
Table 1
Sex-specific counts of violent deaths from both indigenous and non-indigenous conflict (including duels, homicides, and raids, but not infanticide).
Society Violent deaths Percent male Total deaths Percent violent deaths Age range of sample Source
males females total
Tsimane 22 8 30 73 525 6 adults (20+) Gurven, Kaplan, &
Zelada Supa, 2007
Xilixana (pre-contact) 8 2 10 80 64 16 all ages Early & Peters, 2000
Ayoreo 276 20 all ages Bugos, 1985
Yanomamo 113 20 133 85 610 22 mostly adults Chagnon, 1974
Wari’ (pre-contact) 110 400 28 all ages Conklin, 1989
Hiwi (pre-contact) 18 10 28 64 86 33 adults (10+) Hill, Hurtado, & Walker, 2007
Arawete 167a 477a 35a all ages Viveiros de Castro, 1992
Kayapo (pre-contact) 62 21 83 75 237 35 mostly adults Werner, 1980
Achuar 73 33 106 69 250 42 adults Ross, 1988
Ache (pre-contact) 44 22 66 67 153 43 adults (15+) Hill & Hurtado, 1996
Waorani (pre-contact) 164 108 272 60 484 56 mostly adults Larrick et al., 1979
SUM 504 224 1281 4215
MEAN 69.2 30.4
An emphasis is made on using earlier pre-contact periods where available (denoted in parentheses) to minimize the impacts of contact and acculturation. Different studies use
different age ranges for analysis, but results are roughly comparable.
a Indicates the inclusion of some captives, mostly women taken by Kayapo, into the death count.
30 R.S. Walker, D.H. Bailey / Evolution and Human Behavior 34 (2013) 29–34
but not infanticide. The percent of deaths arising from violence varies
considerably from the Tsimane at 6%, the only group in the sample
with no active warfare, to the pre-contact Waorani with incessant
revenge raids at 56%. The average percent of violent deaths across all
11 studies is 30% indicating that violence was probably a potent
selective pressure for many traditional societies. The Yanomamo have
been characterized as “The Fierce People” (Chagnon, 1968) but are
below the mean at 22% violent deaths. The lowland societies in this
cause-of-death sample are roughly divided amongst contexts where
internal strife between communities of the same ethnolinguistic
group was the primary threat (Achuar, Waorani, Yanomamo) versus
those where external conflict outside ethnolinguistic boundaries
predominated (Ache, Arawete, Wari’) versus those with considerable
levels of both internal and external conflict (Ayoreo, Hiwi, Kayapo).
Sex-specific counts of violent deaths allow calculation of the
percent of males versus females that died violent deaths. An average
of 69% of all violent deaths were males and 31% females. The
percentage of male versus female violent death varies little across
societies (standard deviation=8%). This result suggests that the
opportunity for a selective pressure of violence is slightly more than
twice as strong on males as it is on females.
3.2. Event-level body counts
Fig. 1 and Table 2 give event-level body counts at 3 scales of
analysis (within village, internal war, and external war). Internal
warfare events within ethnolinguistic boundaries dominate the
database (55% of death events) with the remaining death events
split between within-village homicides and external warfare events.
Median body count is 1 death for within-village homicides and 2
deaths for both internal and external warfare events, with mean body
counts higher for external warfare. One event of an estimated 300
deaths occurred in an external raid by the Mbaya in 1763 that
destroyed another tribe (Muriel, 1918). Attackers were rarely killed in
our sample; only 5 death events (out of 238 total or 2%) involved a
death of an aggressor. Chagnon (1968) notes that few Yanomamo
attackers were killed or injured while raiding, and Beckerman et al.
(2009) reported no reports of deaths or serious injuries in a large
sample of Waorani raids.
To test for differences in the number of individuals killed across
scales of analysis, a model with number of victims as the dependent
variable and scale as the independent variable was estimated (with a
Poisson link). The reference scale was within-village homicides; both
internal and external warfare contrasts were statistically significant
indicating that external warfare had the highest body count, followed
by internal warfare, and then within-village homicides with the
lowest (Fig. 1, Tables 2 and 3). Results were not altered when the
outlying value of 300 victims was removed (237 observations; z=
2.62, p=.009; z=4.39 pb.0001). A comparison between only internal
and external warfare yields a Poisson regression coefficient of 0.7,
which indicates the increase in the log of expected counts of external
over internal warfare deaths per raid.
In order of importance, the tallied motives for killings (including
multiple responses) were revenge for previous killings or other
wrong-doings like adultery or sorcery (n=63 or 70% of responses),
jealousy over women (n=16 or 18% of responses), gain of captive
women and children (n=6 or 7% of responses), fear or deterrence of
an impending attack (n=3 or 3% of responses), and lastly the theft of
material goods (n=2 or 2% of responses). Other reported motives
were rather idiosyncratic (e.g., duels as a result of gossip, insults, stone
throwing, provocations, garden theft, and boastful attitude of leaders)
and were not included in these counts.
Captives of women and some children are more often taken on
external raids (54% of raids) than on internal raids (26%). However, on
those raids where women are taken, about twice as many are
captured on average during internal raids (2.2) than in external raids
(1.1), perhaps because longer distances must be travelled. This
pattern leads to internal and external raids having the same overall
average of women captured per raid (0.59). Unfortunately, ethnographic accounts rarely mention the actual number of men on raiding
Frequency
50
40
30
20
10
0
50
40
30
20
10
0
Body count
1 10 100
50
40
30
20
10
0
Within village Internal warfare External warfare
Fig. 1. Frequency distribution of body counts for different scales of conflict (within community versus internal war between communities in the same ethnolinguistic group versus
external war between ethnolinguistic groups). Note the log scale on x-axis.
R.S. Walker, D.H. Bailey / Evolution and Human Behavior 34 (2013) 29–34 31
parties but a rough estimate is probably around 10–15 but with a
skewed distribution that may occasionally reach into the hundreds.
3.3. Sequence of death events
Body counts of death events linked in sequences were collected
whenever ethnographic information indicated that a revenge killing
was a direct result of a previous killing (n=30 sequences, 7 of which
have body counts for counter-revenge attacks). To test for differences
in the number of individuals killed across sequences of death events, a
model with number of victims as the dependent variable and
sequence as the independent variable was estimated. The main effect
of sequence was positive, but only marginally significant (67 observations; z=1.69, p=.09; the outlying value of victim count was not
included in these analyses because it was not part of a sequence).
Because there are only a small number of observations of sequences
within villages, the observations within villages and from internal
warfare were combined. This scale variable, along with its interaction
with sequence, was added to the previous model. The main effect of
sequence was not statistically significant (z=.67, p=.51), indicating
that for internal warfare, victim count was unrelated to sequence,
while the interaction between sequence and scale was marginally
significant (z=1.91, p=.056), indicating that the relation between
sequence and victim count was more positive in external warfare.
Finally, when the scale was recoded so that external warfare was the
reference group, the main effect of sequence was statistically
significant (z=2.50, p=.01), indicating that, for external warfare,
victim count increased with sequence.
3.4. Treachery
Treachery, such as hosting a malicious feast or feigning friendship
in some manner to entice victims, is used more in internal warfare (19
treacherous death events or 15% of internal warfare attacks versus
only 2 treacherous death events or 4% for external raids). To test for
differences in body counts as a function of the use of treachery by the
perpetrating group, a model with number of victims as the dependent
variable and treachery as the independent variable was estimated.
The effect of treachery was statistically significant (238 observations;
z=3.98, pb.0001), despite the outlying value of victim count
belonging to the no treachery category. When scale of analysis was
added to this model, the effect of treachery increases in size (238
observations; z=4.31, pb.0001). A comparison between treachery
and other body counts yields a Poisson regression coefficient of 0.53,
which indicates the increase in the log of expected counts of
treacherous over non-treacherous death events. Table 3 summarizes
the results from the event-level body count analyses.
4. Discussion
The compilation of 11 anthropological studies reporting cause of
death in traditional lowland South American societies shows that
violence led to about 30% of all deaths of which about 70% were males.
These results are similar to mean values found in global surveys of
farmer-foragers (Keeley, 1996; Bowles, 2009; Pinker, 2011). Cause-ofdeath studies are mixed in terms of the seriousness of internal versus
external warfare. Event-level analyses suggest that external warfare
events have higher body counts per raid, although internal raids are
more than twice as frequent in our sample. Raids are more deadly
when treachery is used, which is primarily an internal warfare
strategy. External revenge raids kill more people on average than the
original grievance, indicating a tendency towards escalation in
violence and increasingly vicious cycles of revenge killings between
ethnolinguistic groups. Although over half of all external raids take
captives in our sample (over twice as frequently as internal raids),
there are similar numbers of women taken on average per raid. In
sum, internal war is more frequent and treacherous, while external
war has more vicious revenge cycles and higher body counts, perhaps
consistent with the parochial altruism model.
The applicability of the chimpanzee model appears to fit
Amazonian warfare up to a point in that violence is often ongoing
between groups, mostly low cost for attackers during any particular
event, and includes some benefits in terms of access to captured
females and goods and potentially to more territory. However, there
appear to be several qualitative differences in tribal warfare. In
particular, strong motivations for humans to seek revenge, which in
Table 2
Summary statistics for event-level body counts.
Level of events N events Total
body count
Median
body count
Mean
body count
N raids
with captives
Fraction raids
that take captives
Total women
captured
Women captured
per raid
N treachery Fraction
treacherous
Within village 51 100 1 2.0 0 0 0
Internal warfare 131 486 2 3.7 34 0.26 76 0.58 19 0.15
External warfare 56 559 2 10.0 30 0.54 33 0.59 2 0.04
SUM 238 1145 64 109 21
Scale of conflict is defined as within community versus internal war between communities in the same ethnolinguistic group versus external war between ethnolinguistic groups.
Only indigenous-on-indigenous killings are included in these results.
Table 3
Summary of results from the event-level body count analyses and their support for 2 models of warfare.
Death event results N z p-value Poisson Regression coefficient Standard error Supports chimpanzee model Supports parochial altruism
ExternalNInternal 238 4.12 b0.0001 0.72 0.18 yes
ExternalNWithin-village 238 4.55 b0.0001 0.89 0.20 yes
InternalNWithin-village 238 1.72 0.09 0.23 0.13 yes
TreacheryNNon-treachery 238 4.31 b0.0001 0.59 0.14 yes
RevengeNOriginal (External warfare) 64 2.50 0.01 0.68 0.27 yes
Revenge primary motivator (70% of raids cite revenge as motive) yes
Low cost for attackers (attackers suffered deaths in 5 of 238 events or 2%) yes
Fitness benefits (women and children captives; some goods) yes
Fitness benefits (disintegration/weakening of enemy groups) yes yes
Male coalitions yes yes
Group territoriality yes yes
32 R.S. Walker, D.H. Bailey / Evolution and Human Behavior 34 (2013) 29–34
the lowland South American case makes attacks increasingly costly,
appears to represent a novel psychological trait in humans designed
to deal with continual warfare by demonstrating resolve of individuals and the group to not be weak and intimidated by enemies.
Not seeking revenge is an open admission to weakness and
emboldens the enemy to further attacks and insults with impunity.
Treachery also appears to represent a derived trait in humans that
requires considerable in-group cooperation that is designed for
between-group competition. The use of treachery, if convincing and
successful, is an extreme example of a low cost attack by aggressors,
consistent with the chimpanzee model, but involves sophisticated
levels of within-group cooperation and enticement of between-group
cooperation only possible in human societies, and of course opens up
perpetrators to revenge raids in the future. Humans are also unique in
demonstrating a multi-tiered social structure where marriage
exchange, trade, and communication (language) serve to tie together
multiple residential communities (Lévi-Strauss, 1949; Rodseth,
Wrangham, Harrigan, & Smuts, 1991; Chapais, 2008; Hill, Walker,
Bozicevic, Eder, et al., 2011). These mechanisms create meta-groups
that ratchet up the scale of both cooperation and competition in
warfare by uniting multiple families, lineages, villages, and even
chiefdoms against other alliances often pitted against one another in
cycles of ongoing violence motivated by revenge (Beckerman &
Valentine, 2008; Boehm, 2011).
Mulitlevel selection is more salient when there is more genetic
variance between groups and when these groups are in direct
competition with one another, allowing the benefits of being in an
altruistic group to outweigh within-group costs of altruistic behavior
(Price, 1970; Frank, 1998; Henrich, 2004). Levels of genetic differentiation among Amazonian societies are some of the highest in the
world with autosomal FST values averaging around 15%, over twice the
world average (Wang, Lewis, Jakobsson, Ramachandran, et al., 2007;
Lewis, 2010). Genetic heterozygosities within Amazonian societies
average around 0.57, compared to world averages of 0.69 (Wang et al.,
2007; Callegari-Jacques, Tarazona-Santos, Gilman, Herrera, et al.,
2011). These patterns follow from the relatively small and isolated
nature of many Amazonian societies promoting genetic drift, along
with assortative fissioning, and in being one of the world’s regions
farthest in migrational distance from the East African homeland of
modern humans. Migration between groups via marriage exchange
erodes genetic differentiation, but if exchange occurs between
friendly alliances then these entities are less likely to raid one another
and more likely to cooperate in raids against common enemies.
Therefore, the boundary where marriage-exchange networks end and
open hostilities begin is the primary candidate for genetic group
selection, because genetic differentiation and the force of betweengroup selection are expected to be the highest at this interface. While
between-group competition may not consistently lead to direct
extermination, some groups often become weakened, forced to
merge with other groups, and suffer fitness depression.
Reproductive leveling also promotes multilevel selection by
disseminating fitness benefits more widely amongst in-group members (Bowles, 2006), although leveling mechanisms are arguably best
considered a consequence and not a cause of prosociality (Boyd,
2006). Reproductive leveling may be a particularly important
component of multilevel selection in Amazonian societies because a
lack of important heritable resources, like land and livestock (Gurven,
Borgerhoff Mulder, Hooper, Kaplan, et al., 2010), leads to low levels of
polygyny. Partible paternity, the conception belief that multiple men
can be co-genitors of a single child, is also common in Amazonia
(Beckerman & Valentine, 2002; Walker, Flinn, & Hill, 2010) and may
represent another form of reproductive leveling. Extensive warfare,
reproductive leveling, and high genetic differentiation are all
potentially important factors for promoting multilevel selection
(Bowles & Gintis, 2011). The current study cannot, however,
determine whether selection has favored similar behaviors in groups
of humans around the world or whether special socio-ecological
conditions have been particularly favorable towards promoting
predispositions for between-group conflict in Amazonia.
Analogous to genetic multilevel selection is cultural group
selection (Boyd & Richerson, 1985) that may prevail if cultural
differentiation among groups is substantial and between-group
competition is intense (Soltis, Boyd, & Richerson, 1995). The taking
of captives, often an important aspect of much tribal and chiefly
warfare, acts to erode genetic differentiation. However, this may not
be true for cultural differentiation if captive wives and children
quickly acculturate into new lifestyles of their captors, as appears to
generally be the case (Chagnon, 1968; Stuart, 1980) via conformistbiased transmission (Henrich & Boyd, 1998). An unanswered question
for Amazonia is the degree of cultural differentiation (“cultural FST
values”, Bell, Richerson, & McElreath, 2009) because data for cultural
traits at multiple scales of social structure (e.g., individuals within
villages within ethnolinguistic groups) are lacking. It does seem
reasonable to assume that ethnolinguistic boundaries show the most
between group cultural variation with boundaries defined by shared
language, norms, and institutions. That said, detailed economic
studies have found considerable variation between communities
within ethnolinguistic boundaries in Amazonia and elsewhere
(Henrich, Boyd, Bowles, Gintis, et al., 2005; Lamba & Mace, 2011;
Apicella, Marlowe, Fowler, & Christakis, 2012). Future work is needed
to examine whether behavioral and psychological adaptations related
to warfare are truly best explained at the cultural or genetic group
level or are more parsimoniously understood at the individual level.
In sum, we have documented what appears to be the potential for
strong competitive selection of warfare across lowland South America
with violent conflict common both within and between ethnolinguistic groups. Warfare dwarfs violence occurring within villages and
therefore points to the importance of between-group competition
that may help drive parochial altruism. Alternative explanations for
widespread cooperation in warfare, such as mutualism, reciprocity,
and direct individual benefits, do not seem attractive because direct
per-capita benefits (e.g., captive women and limited stolen goods)
appear small in comparison to the risk of reprisals. The two studies to
quantify reproductive success of more renowned warriors have
shown mixed results with both higher (Yanomamo: Chagnon, 1988)
and lower (Waorani: Beckerman et al., 2009) fitness outcomes.
Alternative explanations for cooperation also do not seem to fully
explain the vicious cycles of revenge killings or the warfare-related
cultural norms and institutions supporting group solidarity over
individual interests such as widespread in-group cooperation, low
levels of polygyny, costly initiation rites, training in warfare, men’s
houses, diverse rituals and ceremonies, and norm enforcement
via gossip, fines, ostracism, and witchcraft accusations. For these
reasons we favor the parochial altruism model and multilevel
selection acting above the level of the individual to fully explain
cooperation in lowland South American warfare.
Acknowledgments
This paper benefited from conversations with Karthik Panchanathan, Martin Daly, Ray Hames, Mark Flinn, and Kim Hill. Financial
support was provided by Research Board and Arts & Science Alumni
Organization Faculty Incentive Grants (University of Missouri) to RW.
References
Alexander, R. D. (1974). The evolution of social behavior. Annual Review of Ecology and
Systematics, 5, 325–383.
Apicella, C. L., Marlowe, F., Fowler, J. H., & Christakis, N. A. (2012). Social networks and
cooperation in hunter–gatherers. Nature, 481, 497–501.
Balée, W. (2007). Complexity and causality in Tupinambá warfare. In R. J. Chacon, & R.
G. Mendoza (Eds.), Latin American indigenous warfare and ritual violence
(pp. 180–197). Tuscon: University of Arizona Press.
R.S. Walker, D.H. Bailey / Evolution and Human Behavior 34 (2013) 29–34 33
Bates, D., & Maechler, M. (2010). lme4: Linear mixed-effects models using S4 classes. R
package version 0.999375–33. URL. http://CRAN.R-project.org/package=lme4.
Beckerman, S. (1979). The abundance of protein in Amazonia: A reply to Gross.
American Anthropologist, 81, 533–560.
Beckerman, S., & Valentine, P. (Eds.). (2002). Cultures of multiple fathers: The theory and
practice of partible paternity in lowland South America. Gainesvile: University Press
of Florida.
Beckerman, S., & Valentine, P. (Eds.). (2008). Revenge in the cultures of lowland South
America. Gainesville: University Press of Florida.
Beckerman, S., & Yost, J. (2007). Upper Amazonian warfare. In R. J. Chacon, & R. G.
Mendoza (Eds.), Latin American indigenous warfare and ritual violence
(pp. 142–179). Tucson: University of Arizona Press.
Beckerman, S., et al. (2009). Life histories, blood revenge, and reproductive success
among the Waorani of Ecuador. Proceedings of the National Academy of Sciences of
the United State of America, 106, 8134–8139.
Bell, A. V., Richerson, P. J., & McElreath, R. (2009). Culture rather than genes
provides greater scope for the evolution of large-scale human prosociality.
Proceedings of the National Academy of Sciences of the United State of America,
106, 17671–17674.
Boehm, C. (2011). A natural history of retaliatory violence. British Journal of Criminology,
51, 518–534.
Bowles, S. (2006). Group competition, reproductive leveling, and the evolution of
human altruism. Science, 314, 1569–1572.
Bowles, S. (2009). Did warfare among ancestral hunter–gatherers affect the evolution
of human social behaviors? Science, 324, 1293–1298.
Bowles, S., & Gintis, H. (2011). A cooperative species: Human reciprocity and its
evolution. Princeton, NJ: Princeton University Press.
Boyd, R. (2006). The puzzle of human sociality. Science, 314, 1555–1556.
Boyd, R., & Richerson, P. J. (1985). Culture and the evolutionary process. Chicago:
University of Chicago Press.
Bugos, P. E. Jr. (1985). An evolutionary ecological analysis of the social organization of
the Ayoreo of the Northern Gran Chaco. PhD dissertation, Northwestern University,
Evanston, IL.
Callegari-Jacques, S. M., Tarazona-Santos, E. M., Gilman, R. H., Herrera, P., et al.
(2011). Autosome STRs in native South America: Testing models of association
with geography and language. American Journal of Physical Anthropology, 145,
371–381.
Carneiro, R. L. (1981). The chiefdom: Precursor of the state. In G. D. Jones, & R. R. Kautz
(Eds.), The transition to statehood in the New World (pp. 37–79). New York:
Cambridge University Press.
Chagnon, N. (1968). Yanomamö: The fierce people. New York: Holt, Rinehart, and
Winston.
Chagnon, N. (1974). Studying the Yanomamö. New York: Holt, Rinehart and Winston.
Chagnon, N. (1988). Life histories, blood revenge, and warfare in a tribal population.
Science, 239, 985–992.
Chagnon, N., & Hames, R. (1979). Protein deficiency and tribal warfare in Amazonia:
New data. Science, 203, 910–913.
Chapais, B. (2008). Primeval kinship: How pair-bonding gave birth to human society.
Cambridge: Harvard University Press.
Choi, J. -K., & Bowles, S. (2007). The coevolution of parochial altruism and war. Science,
313, 636–640.
Conklin, B. A. (1989). Images of health, illness and death among the Wari’ (Pakaas
Novos) of Rondonia, Brazil. PhD dissertation. University of California, San Francisco.
Darwin, C. (1871). The descent of man, and selection in relation to sex. London: John
Murray.
Early, J. D., & Peters, J. F. (2000). The Xilixana Yanomami of the Amazon: History, social
structure, and population dynamics. Gainesville: University Press of Florida.
Fausto, C. (2001). Inimigos Fiéis: História, Guerra e Xamanismo na Amazônia. São Paulo,
Brazil: Editora da Universidade de São Paulo.
Ferguson, R. B. (1995). Yanomami warfare: A political history. Santa Fe: School of
American Research Press.
Frank, S. A. (1998). Foundations of social evolution. Princeton, NJ: Princeton University
Press.
Gintis, H. (2000). Strong reciprocity and human sociality. Journal of Theoretical Biology,
206, 169–179.
Gross, D. R. (1975). Protein capture and cultural development in the Amazon Basin.
American Anthropologist, 77, 526–549.
Gurven, M., Borgerhoff Mulder, M., Hooper, P., Kaplan, H., et al. (2010). Domestication
alone does not lead to inequality: Intergenerational wealth transmission among
horticulturalists. Current Anthropology, 51, 49–64.
Gurven, M., Kaplan, H., & Zelada Supa, A. (2007). Mortality experience of Tsimane
Amerindians: Regional variation and temporal trends. American Journal of Human
Biology, 19, 376–398.
Hamilton, W. D. (1975). Innate social aptitudes of man: An approach from evolutionary
genetics. In R. Fox (Ed.), Biosocial Anthropology (pp. 133–153). London: Malaby Press.
Harris, M. (1974). Cows, pigs, wars and witches: The riddles of culture. New York: Random
House.
Hemming, J. (1978). Red gold: The conquest of the Brazilian Indians. Cambridge: Harvard
University Press.
Henrich, J. (2004). Cultural group selection, coevolutionary processes and large-scale
cooperation. Journal of Economic Behavior & Organization, 53, 3–35.
Henrich, J., & Boyd, R. (1998). The evolution of conformist transmission and the
emergence of between-group differences. Evolution and Human Behavior, 19,
215–242.
Henrich, J., Boyd, R., Bowles, S., Gintis, H., et al. (2005). ‘Economic Man’ in cross-cultural
perspective: Ethnography and experiments from 15 small-scale societies. The
Behavioral and Brain Sciences, 28, 795–855.
Hill, K. R., & Hurtado, A. M. (1996). Ache life history: The ecology and demography of a
foraging people. Hawthorne, NY: Aldine de Gruyter.
Hill, K. R., Hurtado, A. M., & Walker, R. S. (2007). High adult mortality among Hiwi
hunter–gatherers: Implications for human evolution. Journal of Human Evolution,
52, 443–454.
Hill, K. R., Walker, R. S., Bozicevic, M., Eder, J., et al. (2011). Co-residence patterns in
hunter–gatherer societies show unique human social structure. Science, 331,
1286–1289.
Ihaka, R., & Gentleman, R. (1996). R: A language for data analysis and graphics. Journal
of Computational and Graphical Statistics, 5, 299–314.
Keeley, L. H. (1996). War before civilization: The myth of the peaceful savage. Oxford:
Oxford University Press.
Lamba, S., & Mace, R. (2011). Demography and ecology drive variation in cooperation
across human populations. Proceedings of the National Academy of Sciences of the
United State of America, 108, 14426–14430.
Larrick, J., Yost, J., Kaplan, J., King, G., & Mayhall, J. (1979). Patterns of health and disease
among the Waorani Indians of Eastern Ecuador. Medical Anthropology, 3, 147–189.
Lévi-Strauss, C. (1949). Les Structures Élémentaires de la Parenté. Paris: Presses
Universitaires de France.
Lewis, C. M. (2010). Hierarchical modeling of genome-wide short tandem repeat (STR)
markers infers Native American prehistory. American Journal of Physical Anthropology, 141, 281–289.
Manson, J. H., & Wrangham, R. W. (1991). Intergroup aggression in chimpanzees and
humans. Current Anthropology, 32, 369–390.
Muriel, D. (1918). Historia del Paraguay desde 1747 hasta 1767. Madrid: V. Suárez.
Nisbett, Richard E., & Cohen, Dov (1996). Culture of honor: The psychology of violence in
the south. Boulder: Westview Press.
Pinker, S. (2011). The better angles of our nature: Why violence has declined. London:
Viking.
Price, G. R. (1970). Selection and covariance. Nature, 227, 520–521.
Redmond, E. M. (1994). Tribal and chiefly warfare in South America. Ann Arbor:
University of Michigan Museum of Anthropology Press.
Rodseth, L., Wrangham, R. W., Harrigan, A. M., & Smuts, B. B. (1991). The human
community as a primate society. Current Anthropology, 32, 221–254.
Ross, J. B. (1988). A balance of deaths: Revenge feuding among the Achuarä Jívaro of the
northwest Peruvian Amazon. New York: Columbia University.
Soltis, J., Boyd, R., & Richerson, P. J. (1995). Can group-functional behaviors evolve by
cultural group selection?: An empirical test. Current Anthropology, 36, 473–494.
Stuart, D. E. (1980). Kinship and social organization in Tierra del Fuego: Evolutionary
consequences. In L. Cordell, & S. Beckerman (Eds.), The Versatility of Kinship
(pp. 269–284). New York: Academic Press.
Verswijver, G. (1992). The club-fighters of the Amazon: Warfare among the Kaiapo Indians
of central Brazil. Gent: Rijksuniversiteit Te Gent.
Viveiros de Castro, E. (1992). From the enemy’s point of view. Humanity and divinity in an
Amazonian society. Chicago: University of Chicago Press.
Wadley, R. L. (2003). Lethal treachery and the imbalance of power in warfare and
feuding. Journal of Anthropological Research, 59, 531–554.
Walker, R. S., Flinn, M. V., & Hill, K. R. (2010). Evolutionary history of partible paternity
in lowland South America. Proceedings of the National Academy of Sciences of the
United State of America, 107, 19195–19200.
Wang, S., Lewis, C. M., Jakobsson, M., Ramachandran, S., et al. (2007). Genetic variation
and population structure in Native Americans. PLoS Genetics, 3, e185.
Werner, D. W. (1980). The making of a Mekranoti chief: The psychological and social
determinants of leadership in a native South American society. PhD dissertation.
City University of New York, New York.
Wilson, D. S., & Dugatkin, L. A. (1997). Group selection and assortative interactions.
American Naturalist, 149, 336–351.
Wilson, M. L., Wallauer, W. R., & Pusey, A. E. (2004). New cases of intergroup violence
among chimpanzees in Gombe National Park, Tanzania. International Journal of
Primatology, 25, 523–549.
Wrangham, R. W. (1999). Evolution of coalitionary killing. Yearbook of Physical
Anthropology, 42, 1–39.
Wrangham, R. W., & Glowacki, L. (2012). Intergroup aggression in chimpanzees and
war: Evaluating the chimpanzee model. Human Nature published, online 03 March
2012.
Wrangham, R. W., & Peterson, D. (1996). Demonic males: Apes and the origins of human
violence. Boston: Houghton Mifflin.