Download INTRODUCTION

To be published in Cross-Cultural Research, November 2008
PURPOSIVE SOCIAL SELECTION
AND THE EVOLUTION OF HUMAN ALTRUISM
Christopher Boehm
Departments of Anthropology and Biological Sciences
University of Southern California
Normally scientific evolutionary approaches eschew any element of teleology in
theorizing about how natural selection processes work, but social decisions pose a
problem for this position. I examine both positive and negative social sanctioning
by human groups, to show that purposive social selection at the level of phenotype
can have parallel effects at the level of genotype and that social control has shaped
human genetic nature profoundly. A small crosscultural sample of Pleistoceneappropriate foragers is used to suggest that human groups have been favoring generous individuals for at least 45,000 years, in ways that affect their fitness. The impact on gene pools could have been substantial, since runaway selection is likely to
have been involved. This type of explanation, which takes human purposes into account as part of evolutionary process, offers a new solution for the genetic paradox
of altruistic behavior.
INTRODUCTION
In the broader evolutionary debates about the ultimate causal forces which drive natural selection process, positions taken have been both predictable and highly disparate. Theorists who
believe the “design” somehow to be “intelligent” include not only religious fundamentalists but
many others, far less doctrinaire, who argue that some higher entity must at least have created natural selection process and turned it loose to do its remarkable mechanical thing. On the scientific
side, of course, are an array of “mechanists” whose world views are based on the totally blind
“variation-and-selective-retention” evolutionary paradigm characterized by Campbell (1965).
1
Behind this controversy lurks the cosmological question of “teleology” (see van Woudenberg 2004). Is the natural world basically self-organizing—or has it been shaped by some kind of
purpose? In this context Ernst Mayr (1983) created a special term to describe how a system of natural selection can appear to be solving problems actively when, in fact, it is self-organizing. The
process is not purpose-driven but teleonomic, meaning that individual genetic competition supports gene pools which automatically adjust their members’ phenotypes to environmental change,
and do so unthinkingly—even though there may be a striking appearance of ingenuity.
Mayr’s position is accepted by the great majority of evolutionary scientists today, who defensively remain highly suspicious of any attempt to introduce teleological elements into evolutionary analyses. Keeping this polarization of opinion in mind, I shall examine some important
patterns of purposive human decisionmaking, to see whether they might have had a special “guidance” effect on certain aspects of natural selection process as humans evolved socially. This question is of particular interest because a major focus in the pages that follow will be on altruistic
generosity and how it could have evolved in humans and, because when social decisions become
involved with natural selection processes, not only is there introduced an element of intentionality
that provides some guidance, but in theory runaway selection could also be taking place (e.g.,
Fisher 1930).
Of course, these same scientists will readily acknowledge that modern humans, with their
scientific understanding of heredity, can purposefully guide evolutionary process in some very
special and insightful ways such as genetic engineering, or, deplorably, in trying to implement policies of eugenics. But these were not part of our deep evolutionary past. It was only at the end of
the Pleistocene that the intensive domestication of animals and plants first brought humans some
real potency to shape the genotypes of other species under their control, and until Darwin their insights into the overall effects of these practices surely remained very limited.
2
Darwin grew up in proximity to livestock breeding, and there is little doubt that his precocious insights into natural selection process sprang in part from this experience as well from what
he absorbed as a descriptive naturalist. In The Origin of Species, when Darwin (1859) discussed
the breeding activities of farmers, he came to the conclusion that they were practicing unconscious
selection. By this he meant that they were unaware of the larger effects of their selective efforts—
even though their immediate, deliberate decisions about which useful variants to preserve were
having an impact on the course of natural selection.
Were humans the only species to influence natural selection by making focused, purposeful decisions? Subsequently, Darwin (1871) took up the question of sexual selection in other animals. Even though they obviously are not equipped to understand either the immediate or the ultimate genetic effects of their decisions, in some species the females do instinctively choose males
whose displays please the most, and the “runaway” power of such decisions as an agency of selection is remarkable (Fisher 1930). Indeed, secondary sexual characteristics like the tails of male
peacocks can evolve so decisively that they seem exaggerated—and in noteworthy ways costly to
fitness. It is the highly-patterned and surely coevolved social decisions of the females, that hold
such male traits in place.
“Social selection” has been discussed in a wide variety of contexts (e.g., Frank 2005;
Hodgson and Knudsen 2006; Simon 1990; Stark 1961; Tanaka 1996; Wolf et al. 1999; Wrangham
2002), and sexual selection is an important subset of social selection. A number of scholars (e.g.,
Breden and Wade 1991; Nesse 2007) believe that this model can be applied not just to selection in
favor of secondary sexual characteristics, but to other aspects of mate choice—including preferences which favor altruistic behavioral traits in a mate. The argument is that because such selection can rise to a runaway rate, substantial altruistic costs may be defrayed as both wiser choosers
and the more altruistic individuals chosen profit from the interaction.
3
Here we will be examining several types of choice regularly made by human foraging
bands, to build the case that innate human altruism could have been supported by social choices
which rewarded generous behavior at the level of phenotype. The argument will be made that a
high degree of “selective focus” (i.e., individuals or groups deciding to reproductively reward or
punish certain specific phenotypes) can transform long-running, culturally-supported decision biases into a relatively powerful mechanism of natural selection. However, in contradistinction to
peacock hens, with humans such decisions are far from being heavily “instinctive” or based on
thoughtless imitation. In fact, our adaptively-significant, cognitively-sophisticated decisions may
become both highly patterned culturally (see Boehm 1978, 1982a) and also highly collectivized
(Boehm 1991, 1996). Insofar as these decisions in fact are purposive, some “lower-level teleology” (Boehm 1991) would seem to be creeping into the process of gene selection.
The main descriptive thrust, here, is to ethnographically identify certain prosocial decision
biases among Pleistocene-type foragers, biases which could have resulted in strong selection in
favor of altruistic generosity. However, as we examine in some detail the social biases of their
small, moralistic groups, I will undertake one other discussion on the side. This has to do with
group punishment’s impact upon human genetic nature, which also involved “guided” social selection.
Techniques of cross-cultural research will be used, to explore certain highly purposeful
patterns of human social preference—ones backed by moralistic sanctioning—which seem likely
to have been sufficiently consistent, over enough thousands of generations, to have affected human gene pools profoundly. But first, it will be instructive to look at one very suggestive matingselection case, where human preferences appear to have shaped a gene pool in precisely the same
direction as the preferences being expressed.
4
EXPLAINING UNUSUALLY TALL SERBS
In On the Origin of Species, Darwin (1859) states that humans sometimes apply unconscious selection to themselves, mentioning the ancient Greeks’ choosing their wives with an eye to
having healthy progeny and the Spartans’ practice of culling their own children. He also mentions
Prussian grenadiers who were selected by height, noting that these exceptionally tall men had exceptionally tall wives. In addition he surmises that nonliterate people may practice some culling,
as when Eskimos select dogs of a certain color for their teams.
When I was doing field work with quasi-literate Serbian mountain pastoralists in Montenegro in the 1960s, I was intrigued by the fact that these “tribesmen,” along with a few contiguous
neighbors in Northern Albania and Southern Hercegovina, were the tallest “caucasian” types in the
world. Furthermore, they contrasted culturally with people in surrounding areas, who lacked the
very prominent historical warrior ethos of these tribes (e.g., Boehm 1983, 1986) and at the same
time were significantly shorter in stature (Ehrich 1946). After Ehrich measured these Serbian
tribesmen anthropometrically, he had no good explanation for their outstanding height aside from
a guess about superior nutrition. However, in the course of 3 years’ cultural fieldwork in Montenegro, I inadvertently obtained two small clues.
As a result of having a warrior society, tribal Montenegrins before 1850 were among the
world’s most patrifocal people (Boehm 1983). Accordingly, having sons to perpetuate the male
family line was a major cultural preoccupation which often appeared obsessive to outsiders (e.g.,
Wilkinson 1848). The physical characteristics of progeny were also of interest, and one informant
told me that sometimes parents would marry a precious son—one who was unusually short—to a
very tall woman, in order to have taller progeny. In addition, I had noted that being stasita (tall)
was considered a desirable characteristic in brides in general.
5
Knowing that these pastoralists bred livestock in a purposeful way, at the time I was not
surprised to hear about parents seeking tall wives for their sons. But after I left the field I began to
wonder if, over time, such intentions might have affected the regional gene pool in ways unrealized by the people involved. To adequately test this unconscious-selection hypothesis, it would
have been necessary to show not only that such marriage preferences existed and were acted on,
but also that they were likely to have affected the fecundity of tall females compared to other
women. To this end, I asked a female informant from the tribe I had lived in, Milenka Vuksanović,
to make a list of men and women she knew of who were both unmarried and very unlikely to marry (see Boehm 1991). While there were no certain male examples, there were almost a dozen unmarried and essentially unmarriageable females she could identify within her own fairly extensive
social networks, which included perhaps a third of a tribe of 1800.
First, I asked Milenka to guess about why these women never married, and subsequently I
asked her to rank them as tall, medium, or short in stature. My socially-knowledgeable informant
gave two reasons for a woman’s failing to marry in a culture in which marriage was very much the
norm. Half were “old maids” (posidjelice) who, even though they were poštena (sexually virtuous), failed to marry because they were “ugly” (grdna). The other half were “loose” women
(kurve) whose premarital affairs had become known in a patriarchal society which valued female
virginity and had no mercy in “talking” about women and giving them bad reputations. When the
ratings for height were made, the virtuous but “ugly” females invariably turned out to be “short,”
while the women with problematic reputations appeared to follow a normal distribution for height
(see Boehm 1991). The numbers were small and the measurements intuitive, but the patterns were
certainly suggestive. None of the short-but-virtuous women had given birth, and in effect they
composed a sterile caste.
6
Written historical information about the warlike Montenegrin Serbs suggests that the tribal-warrior syndrome which even today distinguishes them from their neighbors goes back at least
to the times of the ancient Greeks and Romans (see Wilkinson 1848), so this shorter time frame
would have allowed minimally 80 generations for this pattern of unconscious selection to have
some effect. Since radical changes in the social genotype of domesticated foxes have been accomplished in only 40 generations (Hare et al. 2005), 80 generations might have sufficed for the Montenegrins to significantly raise their average height, particularly if runaway selection (e.g., Fisher
1930) was operating. Unfortunately, there are no historical data for still earlier times in Montenegro, to tell us whether the pattern of being extremely warlike and patrifocal has greater historical
depth, but the cultural pattern of seeking taller progeny could in fact be much older.
Indeed, Balkan people appear to have been domesticating plants and animals for at least
7,400 years (Diamond 1997), so potentially these Montenegrins could have been applying limited
insights about the hereditary transmission of traits to patrifocal marriage choices for at least 296
generations. If animal breeders can change major traits in a species within a few dozen generations
when totally-systematic selective breeding is being implemented, several hundred generations
might well have provided sufficient time for long-term, culturally-patterned biases in favor of
gaining taller children to have affected the gene pool—even though only one sex was targeted for
culling.
This Montenegrin hypothesis definitely involves Darwinian unconscious selection, and it is
useful for thinking more generally about the influence of purposeful decision patterns on natural
selection process. In this case there were perceptually-obvious variations in the human phenotype,
and these people already understood how to select variable traits in their animals and plants. Their
conscious intention was to have taller sons, but rather than “culling” male progeny, they simply
married all their sons to the tallest (sexually-virtuous) females they could find. The outcome
7
would have affected height in both males and females, of course, and over the evolutionary long
run this pattern of choices could have affected an entire regional breeding population.
WIDENING THE SCOPE OF “UNCONSCIOUS SELECTION”
If one is interested in the human-evolutionary role of purposeful decisions, the obvious
place to look is not at tribesmen like these, who maximally have been around for only 12,000
years (less than 500 generations), but at hunter-gatherers of the culturally-modern type who have
been making their livings in bands for nearly 50,000 years or almost 2,000 generations. Such people evolved in the Late Pleistocene (Klein 1999), and their direct descendants have been studied
by ethnographers everywhere in the world.
These people did not practice any intensive domestication of plants, and we can’t be at all
sure whether it was deliberate culling that led to the domestication of dogs at least 15,000 years
ago (see Vila et al. 1997), so we will not be looking for instances of “unconscious selection” precisely as Darwin defined it, nor will the treatment be limited to mate choice. Our potential subject
matter will be any case of purposeful social decisionmaking which affects gene pools in ways that
seem consistent with the intentions of the decision-makers—even if the latter are not consciously
trying to manipulate hereditary outcomes at all. The type of decisions to be treated will be those
involved with social control, since I believe such behavior can be confidently projected into our
human past for at least 1800 generations (see Boehm 1997, 2000).
Natural selection proceeds at variable rates (see Eldridge 1971), but Wilson (1978) has estimated that “normally” a given selection pressure must operate on a species for at least 1000 generations to introduce a new trait. Thus, hunter-gatherers should have had sufficient time to evolve
new or highly modified traits after they became culturally modern like us. However, in the case of
traits selected in a highly focused way, through decisions, the time frame needed might have been
less—and if runaway selection process was also in force the time frame could have become short8
er still.
“PLEISTOCENE-APPROPRIATE” FORAGERS
The goal, here, is to assess certain patterns of individual and group decisionmaking that
could have impacted on human gene pools over the past 45,000 years, so the preliminary crosscultural analysis will be based on extant mobile hunter-gatherers, whose socioecological patterns
most closely approximate those of culturally-modern foragers who lived in the Late Pleistocene.
For brevity I will be calling these contemporary hunter-gatherers “Pleistocene-appropriate,” and
this cross-cultural sample will include only hunting-and-gathering bands which meet the following
criteria:
1. They are mobile. Excluded are foragers who live in permanent, year-round settlements
made possible by intensive food storage.
2. Basically, they are economically independent. (This excludes fur-trade foragers, symbiotic adaptations with agriculturists like those of the Pygmies or the Agta, people living part-time at missions like the Aché, etc.)
3. They are “pure” hunter-gatherers. Excluded are foragers who practice domestication,
such as mounted hunters in North and South America, or even those who practice a little horticulture. One exception will be the use of dogs, because dogs have been domesticated since at least about 15,000 years ago (Vila et al. 1997), and probably earlier.
Recently (Boehm 2002), I have informally surveyed several dozen of these “Pleistoceneappropriate” foragers (see also Knauft 1991; Marlowe 2005), and, in spite of their being adapted
to an extremely wide diversity of environments (Kelly 1995), a number of important behavioral
patterns appear to hold constant (see also Boehm 1997). For instance, these people invariably
combine hunting and gathering but show a preference for large game. Furthermore, there is a highly-predictable ethos which makes for a strong consensus about holding down dominance behaviors that affect the entire band, and as a result they are predictably and assertively egalitarian in
their political life (Boehm 1993; see also Erdal and Whiten 1994; Wiessner 1996).
9
There is also band-level cooperation which is centered on sharing large game, along with
strong (but not absolute) tendencies to pair-bonded monogamy (Marlowe 2005), a tendency to exhibit substantial homicide rates due to male competition over females (Knauft 1991), a strong concern for conflict resolution within the multi-family band (see Boehm 2000; Fry 2000), and, when a
common problem is faced, a capacity of entire bands to arrive at a decision consensus in spite of
having very weak leadership (Boehm 1997). As will be seen, their social preferences include not
only a dislike of interpersonal domination and self-interested deception, but an active appreciation
of those with generous traits.
By examining in some detail a crosscultural corpus of 10 of these forager societies, I will
try to pinpoint some widespread social preferences which could have had a major impact on the
human behavioral genome. But first it will be necessary to describe briefly the ecological niche
occupied by Late-Pleistocene foragers, to see how this constrained and in effect standardized
much of their sociopolitical life.
PREHISTORIC GROUP LIFE
In reconstructing Late-Pleistocene social life, I shall be operating on the assumption that if
universals or strong central tendencies can be identified in the social behavior of “Pleistoceneappropriate” foragers today, these same central tendencies may be projected backwards in time, to
apply to prehistoric humans (see Boehm 2002)—as long as they were culturally modern, invested
substantial energy in hunting large game, and for logistical reasons needed to equalize their meat
distribution.
I will be basing this reconstruction on the notion of a cultural core (e.g., Steward 1955),
which holds that environment and subsistence determine social organization (see Bettinger 1991).
However, I have expanded Steward’s concept to include also certain social and ideological factors
which contribute causally to subsistence success, and at the same time help to shape decision bias10
es and influence the overall social organization. Elsewhere (Boehm 1997), I have described this
ideologically-supported core pattern in detail, calling it an “egalitarian syndrome.” I did so because one of its central features is for people to deliberately hold down alpha behavior—both so
that meat distributions can be efficient, and, more generally, because dominance is predictably resented by egalitarian humans (Boehm 1999).
Large game hunting has figured prominently in human social evolution (see Stanford
1999). Today, with respect to vegetable foods and small game, in their bands mobile foragers basically subsist as independent families, without much interfamilial sharing (Kelly 1995). However,
to cope with the nutritionally-potent large-game carcasses they strongly prefer as food, they are
obliged to deal with the fact that meat takes effort to preserve, and in any event is hard to carry
when camps frequently move. This means that if single families were to consume all the meat
their own hunters provided, there would be the very rare feast and basically a meat-famine the rest
of the time.
To reduce this extreme variance in meat intake, in theory hunter-gatherers are obliged to
live in multifamilial bands with at least 5 or 6 hunters, who must transfer major portions of the
meat they kill to the other families (see Winterhalder 2001; Winterhalder and Smith 1981). This
theory is born out on the ground (see Kelly 1995; Marlowe 2005), for it is the “averaging” effect
of having multiple hunters share nutritious meat that makes large game such a viable subsistence
item. This helps to explain a quarter of a million years of archaeological evidence for continuous
large-game hunting by both archaic and modern Homo sapiens, our direct ancestors, on a number
of continents (see Stanford 1999; Stiner 2002). It is in the interest of keeping the analysis conservative that only culturally-modern humans are being considered here.
Whereas sharing food within the family is very strongly supported by innate nepotistic
tendencies, sharing among unrelated families is not (Boehm 2004b). And in fact, in the wider con11
text of band-level cooperation people sometimes seem to share quite ambivalently (see Peterson
1993)—even though actual conflicts over meat are rare. This overt psychological ambivalence fits
with the fact that band members must “police” those who would seriously take advantage of the
system, and this includes not only people who might be stingy, but also bullies, laggards, cheaters,
and thieves (Boehm 1997). A prime challenge is to keep alpha-male bullying propensities under
control, so that this very important sharing system can both work equitably (Boehm 2004c) and
not be impaired by serious social conflict (Boehm 1982b).
In quickly surveying almost a third of the approximately 150 foraging societies I have now
identified as “Pleistocene-appropriate” (see Boehm 2002), the earmarks of a consistent “egalitarian syndrome” having these features seemed to be pervasive. And by the time that prehistoric Culturally-Modern Humans arrived, we may be reasonably certain that an egalitarian syndrome very
similar to what is found today had developed. I make this assertion because by 45,000 years ago
not only were human skeletal remains the same as today, but by then archaeological signs of cultural inventiveness and flexibility had become dramatic and “contemporary” (see Klein 1999); and
reliance on large game was comparable to today (Stiner 2002).
SOME SPECIAL EFFECTS OF PUNISHMENT
Today’s Pleistocene-appropriate people tend to move camp about 9 times a year (see Marlowe 2005) and they always favor large game because it is fattier (see Cordain et al. 2000, 2001,
2002)—even though vegetable foods, along with small game, so often provide the bulk of their
calories (Kelly 1995). Decisions about where to go next sometimes are made at the family level
(e.g., Palmer et al. 1997), but entire bands often make collective decisions about where to relocate,
and they can make group decisions in response to social challenges, as well (Boehm 1997).
One type of collective decision stands out in such egalitarian bands (Boehm 2002). When
an individual departs too far from the egalitarian ethos, the entire group either turns against him
12
and cuts him down to size, or else it simply does away with him by means of ejection from the
group or through capital punishment (see Boehm 1997). These outcomes have reproductive consequences that ultimately can impact our gene pools, for even being ostracized can tell on reproductive success—while an especially aggressive male’s being put to death, even at the midpoint of
his reproductive career, will seriously damage his inclusive fitness.
If unusual dominant aggressivity has any basis in the genes carried by these political upstarts, and this seems extremely likely, then the egalitarian practice of punishing such males
should be affecting gene pools—an effect that has not gone unnoticed. Trivers (1971), in discussing the “moralistic aggression” of small human groups, took note of how genetic behavior traits
making for deviance could be suppressed by severe punishment, and Otterbein (1988) has applied
this insight specifically to capital punishment. Wrangham (2002) has referred to the effects of such
punishment as “autodomestication,” and he suggests that an originally more aggressive ancestral
political nature (see Wrangham and Peterson 1996) could have become “docilized” as a result.
I have agreed that aggressive traits would have been reduced due to punishments meted out
by groups, and I also have proposed that the human capacity for “self-monitoring” (see Campbell
1983) and self control could have been augmented and transformed as a result of these special,
social-selection effects (see Boehm 1999). The specific hypothesis is that individuals better disposed to internalize group rules, evaluate their own behavior, and hold back from deviant acts,
would have had a net gain in fitness by avoiding punishment. We call the special type of selfappraisal that has evolved in conjunction with our brain’s large prefrontal cortex (see Damasio et
al. 1990) a conscience (see Darwin 1871), and socially-sensitive consciences serve as internal testing grounds where individuals can think about mistakes they are likely to make, and better desist
from making them.
13
These theories about genetic effects of social punishment could help to explain the evolution of moral behavior, and the conscience-evolution hypothesis is currently under development.
In this paper, I will be concentrating on a different hypothesis, which focuses on how altruism
evolved, but again the agency of selection will be a prehistoric band expressing its predictable social preferences.
THE “SAMPLE”
My goal is to assess the likely social biases of hunter-gatherers over the past 45,000 years
and, as a way of generating hypotheses for future testing, to explore the possibility that human
dispositions to altruism could have been produced by totally-unconscious “sanctioning selection”
operating through social control at the group and individual levels. A crosscultural sample of 10
extant “Pleistocene-appropriate,” hunter-gatherer societies will be used to do this.
This preliminary sample has not been fully randomized. At present I am constructing a
large database on Pleistocene-appropriate hunter-gatherers, which when complete will permit
scholarly access to an enormous amount of published information which is relevant to social behavior, morals, cooperation, conflict resolution, and politics, for the scores of mobile, band-level
societies of this Late-Pleistocene type which have been studied ethnographically. The coding protocol is 6 pages long just for these types of behavior, and to date approximately 50 societies have
been completely or partially coded, with the assistance of the John Templeton Foundation and the
Goodall Research Center at USC.
This raw coded information has not yet been incorporated into a searchable database, but a
useful portion of it is now able to be analyzed by hand. The societies are from all over the world,
and the standard procedure is to code every ethnography in the English language for each culture
that is fully processed. The eventual aim is to create an electronic multimedia database for use by
14
scholars interested in forager behavior and human evolution, and an appropriate search-engine architecture is now beginning development.
Several dozen cultures have been fully processed now, and to conduct this pilot survey I
informally selected ten societies, partly on the basis of wanting to have a representative worldwide
distribution, but also on the basis of which of the societies were fully coded and had particularly
rich or abundant ethnography available. Included are Murngin and Tiwi from Australia; !Ko
Bushmen from Africa; Plateau Yumans from North America; Waorani and Yahgan from South
America; and four Inuit groups selected from different parts of the Arctic. Asian foragers were
omitted for logistical reasons alone; the Andaman Islanders were not yet fully processed. Once I
chose this list, it was final.
All of these groups qualify as Pleistocene-appropriate, even though the Inuit societies relied unusually heavily on large game and their dogs could serve as an emergency food source. Because they were obviously oversampled I will note any possible pattern anomalies in their social
preferences or antipathies, but these anomalies were few and they will not affect the analysis.
In Table I it should be kept in mind that different societies were described by varying
numbers of ethnographers, of varying personal and theoretical perspectives, and that there was
varying descriptive quality. The table does indicate, however, how widespread a given preference
or sanction or prescription was in this sample of 10 societies, and, very roughly, how emphasized
a given behavior was, ethnographically speaking. Obviously, given the nature of ethnography, if
an attitude or a behavior is not being mentioned for a given society, this cannot be taken as evidence that in fact it was absent.
If negative evidence must be looked at with some suspicion, for likely “universals” the
positive evidence is all but indisputable. For instance, when 10 societies sampled from most world
continents all report murder (see Table I, item 3.00) as an indigenous category of moral deviance,
15
we may reasonably assume that were the sample size increased, murder would continue to appear
on at least a very widespread basis as a specific category of deviance. Knauft’s (1991) research on
“simple societies” bears this out (see also Brown 1991).
[TABLE I HERE]
16
TABLE I – SOCIAL PREFERENCES AND SANCTIONS
C O D I N G
CATEGORY
Area
# of Sources
MORAL
DEVIANCE
3.00 Murder
3.01 Beating
someone
3.020 Bullying
3.021 Trying to
grab power or
position
3.022 Leader's
trying to boss
others around
3.03 Incest
3.04 Adultery
3.05 Man dishonors a woman
3.06 Cheating
3.060 Cheating
an individual
3.061 Cheating
the group [as in
meat sharing]
3.062 Failure to
share
3.07 Stealing
3.08 Lying
3.09 Taboo violation/other supernatural transgression
3.10 Failure to
cooperate
3.2 Commits
sorcery or witchcraft
MORAL
SANCTIONING
4.0 Public Opinion
4.00 Gossip [as
private expression of public
opinion]
4.1 Direct Criticism by group or
spokesman for
group
4.2 Group shaming
4.20 Ridicule
4.21 Other shaming
4.3 Group Ostracism [social
distancing]
4.30 Social
aloofness [cool
relations, reduced
speaking]
4.31 Tendency to
avoid culprit
4.32 Total Shun-
% of
Societies
Total
Cites
Murngin
Tiwi
!Ko
Greenland
Inuit
Netsilik
Inuit
Polar
Inuit
Plateau
Yumans
Waorani
Yahgan
AFR
8
N.
Alaskan
Inuit
ARC
3
AUS
9
AUS
11
ARC
3
ARC
8
ARC
4
NA
9
SA
1
SA
3
25
16
5
3
11
11
2
2
7
77
8
25
15
11
10
34
2
38
12
4
3
1
1
1
1
11
11
4
5
2
1
6
14
7
3
12
4
1
1
7
1
2
9
7
1
3
4
6
1
14
23
26
15
106
13
19
34
6
4
26
1
1
8
19
3
17
3
1
4
2
100%
90%
235
79
70%
0%
12
0
10%
1
90%
100%
70%
65
102
15
17
28
50%
30%
24
9
3
2
30%
9
80%
33
3
100%
60%
100%
95
45
378
13
4
59
21
40%
10
100%
118
25
2
2
1
10
36
10
7
21
4
100%
174
6
11
27
19
1
37
31
11
7
24
100%
61
5
1
4
9
1
12
10
6
2
11
90%
33
2
3
5
10
1
3
3
1
5
60%
38
2
5
3
15
12
1
90%
50%
70
9
4
1
4
1
4
28
3
21
3
4
70%
48
8
8
3
18
1
4
6
70%
16
5
1
3
1
2
3
50%
11
3
1
2
2
3
50%
6
1
1
1
1
1
2
5
3
15
18
4
1
1
1
1
1
1
4
1
1
1
6
3
70
17
1
2
3
1
6
19
1
3
2
ning [no speaking; total avoidance]
4.33 Spatial
distancing [move
or re-orient domicile or camp]
4.34 Temporary
expulsion from
group
4.35 Permanent
expulsion from
group
4.4 Physical
Punishment
4.40 Administer
blows
4.41 Group
member selected
to assassinate
culprit
4.42 Conspire
with enemies to
have culprit
killed
4.43 Entire group
kills culprit
4.5 Supernatural
sanctions
4.50 Belief that
supernatural
powers will
punish transgressions
4.51 Belief that
individual deviance harms entire
group
4.52 People use
supernatural to
manipulate deviants
4.7 Social control
SOCIAL
IDEOLOGY
5.02 Cooperation
favored
5.020 Generosity
or altruism favored
5.021 Sharing
favored
5.022 Aid to
relatives favored
5.023 Aid to
nonrelatives
favored
5.024 Demand
Sharing
25
5
6
1
2
3
40%
15
7
2
5
90%
86
11
3
50%
22
4
1
50%
21
0%
0
60%
13
1
90%
157
8
3
9
90%
155
10
2
9
60%
39
2
50%
18
5
80%
66
5
8
5
100%
225
6
5
100%
219
21
100%
296
100%
100%
98
30%
2
2
2
6
16
6
4
10
1
21
21
13
7
9
1
1
10
3
5
3
6
1
1
9
12
70
31
1
14
9
12
67
31
1
14
3
3
24
6
1
9
2
7
3
27
4
14
49
27
63
31
8
11
11
4
24
12
13
59
45
8
10
23
19
3
23
32
30
92
57
8
9
23
59
5
1
1
6
4
15
21
4
1
1
100%
64
3
1
11
1
5
26
14
1
1
1
80%
48
3
1
1
1
29
6
2
5
2
1
1
6
22
2
1
7
As Table I shows, certain prescriptions and proscriptions are so ethnographically salient
that right across the board they are reported regularly, and often with high frequency. Under “Social Ideology,” invariably cooperation, generosity, and sharing are preferred indigenously, as de18
7
sirable behaviors. Under “Moral Deviance,” all 10 societies agreed unanimously that murder,
theft, adultery, taboo violation, and sorcery were undesirable behaviors, while the majority of societies had mentions of bullying behavior or beating someone, incest, dishonoring a female, failure
to share or cooperate, and either cheating or lying, as deviant behaviors. It is likely that with less
imperfect ethnographic descriptions, some of the latter would have been unanimously reported, as
well.
In general, oversampling the Inuit appears to have made no significant difference in the
patterns that emerged. It is, in fact, exclusively the four Inuit societies which account for the 4
condemnatory mentions of “failure to cooperate.” However, “failure to share,” which may be seen
as a heavily-overlapping but more specific category, was mentioned 8 times, and this included 5
non-Inuit societies. And if these two categories are collapsed, the overall distribution rises to 9 out
of 10.
The purpose of this preliminary survey was simply to look for potentially universal or very
widespread moral beliefs and practices in band life, and consider them as central tendencies which
might be projected into the past. With respect to practices, it is no surprise that social-control techniques also are mentioned ubiquitously: for instance, public opinion and gossiping are always reported, as is some kind of social distancing—the types running from informal social aloofness, to
outright shunning or ejection from the group, to physical elimination by capital punishment. Furthermore, all but one society had shaming or ridicule mentioned, as was the case with direct criticism. In addition, all but one society mentioned either “physical punishment” or “capital punishment,” as acts of social control, while 7 specifically mentioned executions—either by the entire
group or by someone acting for the group.
We now have an outline of potentially universal or very widespread moral concerns and
social control practices of 10 mobile hunter-gatherer societies. It seems likely that decision biases
19
mentioned unanimously within this small sample are likely to apply, in at least a very widespread
fashion, to extant Pleistocene-appropriate societies in general—including the very few that remained in Asia. It also seems likely that empirically any pattern mentioned by, say, 70% or more
of this reasonably representative sample should have quite a substantial distribution worldwide.
For that reason I am including all of the abovementioned behaviors in the central tendencies that I
think can be confidently projected backwards in time—as long as the culturally-modern humans in
question were fulfilling the socioecological criteria specified previously.
To summarize, in terms of punitive social control this limited “sampling” provides some
very strong central tendencies in values, rules, and manipulative practices which mobilize public
opinion and result in criticism, ridicule, and varying degrees of ostracism or physical punishment.
These practices control not only recidivist murder and bullying, but theft, cheating or failure to
share, and also incest and adultery. If we think about the probable long term genetic-selection effects of these group behaviors, obviously they were likely to have suppressed the frequencies of
traits that prepared dominantly-aggressive violence or other ways of selfishly taking away the resources of others, along with certain types of sexual misconduct.
Although bullying behavior was condemned specifically for only 7 out of 10 societies,
more broadly-based surveys that focused on finding high-quality ethnographies (Boehm 1993,
1999) suggest that this is a widespread and probably universal concern of these mobile foragers—
even though severe punitive sanctioning of serious political upstarts by outraged bands takes place
rather rarely. The fact that executions of political bullies have been reported on a number of continents helps to make the case not only that humans are innately disposed to dominate other people—but that small bands of potential victims are likely to strongly oppose such behavior (Boehm
1999).
20
THE EVOLUTIONARY EFFECTS OF PUNISHMENT
The argument for prehistoric negative sanctioning selection’s impacting gene pools is easiest to make where capital punishment dramatically curtails a younger male’s reproductive career,
which precludes creating more progeny, and also cuts off further support of existing progeny,
breeding partners, progeny’s offspring, and other close relatives. Ejection from the group can also
be very dangerous to the ejectee’s inclusive fitness, unless a safe haven is found. And within the
band social distancing up to and including shunning will at least disrupt an unreformed deviant’s
cooperative networks (e.g., Briggs 1970), and reduce the chances of making an advantageous marriage.
Thus, the fitness costs would have been significant for individuals unduly disposed to
dominate others through physical aggression or sorcery, or for those disposed to engage in theft or
cheating, or commit incest. In the case of adultery, however, it is my ethnographic intuition that
such behavior is widely condemned but not actively punished in many hunting bands, and sometimes is even winked at. Thus I would suggest that in spite of its being morally condemned, this
particular deviant behavior could be on average improving the individual fitness of adulterers, and
probably the males in particular.
Let us return briefly to the conscience. In other hierarchical animals, individuals keep
themselves out of trouble with their superiors through simple fear of punishment, which translates
into individually-adaptive self-control and submission. The difference, with humans, is that we
actually internalize our groups’ rules (Darwin 1871; see also Boyd and Richerson 1982; Campbell
1983, 1991; Gintis 2003, 2004; Simon 1990; Waddington 1960), and we have a sense of shame
(Piers and Singer 1971). This complex, cognitively-sophisticated emotional reaction helps us to
fully assess in advance societal reactions to our own potentially deviant behaviors—and also to
our prosocial behaviors—before we express them. Thus, over the past 45,000 years not only have
21
humans become less innately-prone to commit certain types of crime, but at the same time we
have become much better at following rules in general. This was because punitive social selection
was favoring the evolution of rule-sensitive consciences which slowed down the expression of our
deviant tendencies.
In referring to the group-punishment that brought all this sanctioning selection about, I
emphasize that such social processes were highly “focused,” insofar as just a handful of antisocial
behaviors were being regularly singled out and regularly punished over dozens of millennia. It was
because of this enduring and widespread cultural preoccupation with controlling certain serious
deviances, that people were unwittingly guiding genetic selection process in complementary directions. This can be considered a very special case of gene-culture coevolution (e.g., Durham 1991),
special because it is possible to directly link decision biases with genetic outcomes (see Boehm
1978).
If we give a high probability to the hypothesis that this type of social selection has been
operating for at least several thousand generations, the arguments of Trivers, and then Otterbein,
Boehm, and Wrangham, which suggest that our genetic nature has been modified accordingly,
make excellent sense. This provides one major instance of sanctioning selection’s having had an
important impact on gene pools, and, because decisions were involved it raises the interesting
question of whether some type of runaway-type selection pressures (see Andersson 1994; Fisher
1930) may have been operating—even though sexual selection was in no way involved.
POSITIVE “SANCTIONING SELECTION” IS LESS OBVIOUS
We turn now to the stubborn evolutionary problem of trying to explain innate altruism. Sociologists (e.g., Turner 2000) define social control as having both a very obvious negative (or punitive) aspect, and also far more subtle positive manifestations. On the negative side, both entire
groups and individuals acting on behalf of the group serve as manipulators or enforcers, and in
22
terms of political dynamics often enough it is most or all of the group that is involved—for instance, when people are obliged to gang up in order to cope with truly dangerous deviants. In this
context the type of group moralistic aggression Trivers (1971) talked about would appear to be
universal among these foragers (Boehm 2000).
In contrast to punitive group outrage, positive sanctioning takes place more at the individual level. Although once in a while a foraging band as a whole may reward good behavior, as in
consensually conferring an informal or formal leadership role on a wise, unassuming, and generous individual, the majority of social rewards would seem to come from individualized or familylevel reactions to someone’s having a good reputation (see Alexander 1987). This takes place
when a socially attractive person figures in a desirable marriage alliance—or is otherwise preferred as a close partner in cooperation. The latter may come in collaborative hunting (Kelly
1995), in setting up trading partnerships that also provide important safe-havens in case of localized scarcity (e.g., Burch 2005; Wiessner 1982), and when entire bands have to make difficult decisions about supporting band members who are incapacitated (e.g., Alexander 1987).
However, even though the actual payoffs to individuals mainly come from other individuals or families that show them favor, basically it is the band as a whole, as a tightly-interconnected
gossiping network (e.g., Haviland 1977), that actually arrives at these assessments of people’s personal reputations (Alexander 1987; see also Wiessner 2005). Thus, both negative and positive social sanctioning may be seen as results of “public opinion” being galvanized into group or individual action. Such opinion is based on careful mutual monitoring (e.g., Campbell 1983), and it is
guided by decision biases which can be highly consistent over space and time—wherever human
foraging groups remain very small, mobile, and independent.
In the context of positive sanctioning selection, the “Social Ideology” section at the end of
Table I shows that the behaviors which stimulate such approval include cooperation, generosity,
23
and sharing in general, and in all 10 societies giving aid both to kin and to nonrelatives were ethnographically mentioned at least once as being desirable. However, such social approval does not
imply that the actual sharing between unrelated families is devoid of ambivalence. Indeed, the
aforementioned “demand sharing,” which involves badgering those who divide the meat to make
them behave more generously (see Peterson 1993), is mentioned for 8 societies. Curiously, the
Netsilik account for 29 out of a total of 48 mentions for demand sharing, yet we know that to
make their winter seal-hunting specialization work they must, in fact, share meat very effectively
(see Balikci 1970). More generally, interfamilial meat sharing, although often somewhat agonistic
in its style—seems basically to be quite effective in equalizing meat consumption (see Kelly
1995).
GENETIC EFFECTS OF POSITIVE SANCTIONING
Compared to punishment positive social sanctioning was much less obvious, ethnographically speaking, and conceivably it was less drastic in its genetic effects—even though mate selection, as one important component, may present some special “runaway” possibilities (e.g., Breden
and Wade 1991). The decision biases were predictable and consistent, and positive sanctioning
very much favored tendencies to share and cooperate, with a specific cultural focus on generosity—both within and between families.
As a matter of common sense we know that the latter, broader type of generosity does
seem to come into play in human life, but basically contemporary evolutionary theory holds this to
be genetically “paradoxical” (e.g., Alexander 1987). In essence, sociobiologists (e.g., Trivers
1971, 1972) have held that genetically humans are capable of nepotism, which fits well with intrafamilial generosity—but does not fit with generosity between unrelated families. This theoretical
stance also holds for evolutionary scholars more generally, including most evolutionary anthropologists and many others. Indeed, the evolutionary paradox of “altruism” that was so lucidly de24
fined by Wilson (1975) has given rise to an enormous interdisciplinary academic industry, whose
work continues strongly today precisely because the altruism-paradox challenge continues to be
challenging. This seems to be the case even though multilevel selection paradigms, which include
group-selection possibilities, are becoming more widely accepted (see Wilson and Wilson 2007).
If we consider the genetic consequences of these strong hunter-gatherer decision biases
that favor generosity, it is clear in everyday life that individuals who vary genetically in the direction of being unusually generous in their responses to the needs of others—including unrelated
others—will be socially favored and their fitness will benefit. For that reason, up to a point their
altruistically-generous genes can be favorably selected in spite of the individual fitness costs incurred. If positive moralistic sanctioning selection becomes sufficiently robust in producing such
effects, then for humans at least there should be no evolutionary paradox.
Because for decades genetic group-selection explanations of altruism have been in doubt
(e.g., Alexander 1974, 1987; Trivers 1985; Williams 1966; Wilson 1975), one alternative explanation has been as follows. In a species that is innately limited to nepotism, altruistic donations to
others might be supported solely at the level of phenotype. Simon (1990) suggests that being unusually culturally “docile” is extremely useful to individual reproductive success because better culture-learners gain higher fitness, and he holds that the large gains from carrying this trait can in
fact be reduced somewhat because such individuals will be particularly susceptible to group messages that call for altruism. The idea is that there should still be a net fitness gain from carrying
this “docility gene” because it is so very useful in other contexts. This is, in effect, a pleiotropic or
“piggybacking” theory of altruism (see also Boehm 1981), which applies only to a highly-cultural
species which moralizes in order to get others to behave altruistically (Campbell 1983; see also
Alexander 1987).
25
Another way of explaining how altruistic cooperation can take place among selfishly nepotistic individuals is advocated by Boyd and Richerson (1992). The title states the position clearly:
“Punishment Allows the Evolution of Cooperation (or Anything Else) in Sizable Groups.” It follows that innate altruism is not necessary for altruistic cooperation to take place; people will cooperate simply out of immediate fear of group social pressure and reprisals (see also Bowles and
Gintis 2004). This last explanation may not look to genes, but it is based on the same negative social forces I discussed earlier.
These everyday punitive sanctioning forces were clearly identified at the social level by
Durkheim (1933), and, before him, by Maine (1861), and the likely genetic effects of negative
sanctioning-selection are, in fact, well-identified by Trivers, Otterbein, myself, Wrangham, and,
more by implication, by Bowles and Gintis (2004). Thus, punishment and the threat of punishment
can immediately stimulate altruistic behavior at the level of phenotype, while over many generations, at the level of genotype, this same punishment can at least reduce genetic tendencies which
favor selfish deviance. What punishment cannot do, however, is to directly select for altruism at
the level of genotype.
There are at least some ethnographic “hints” that positive sanctioning in favor of wider
generosity itself might have some genetic basis—even though in terms of inclusive fitness it
makes sense that humans should be evolved to exhort others to be more generous than oneself (see
Alexander 1987). In surveying some hunter-gatherer prosocial activities in an attempt to find clues
about whether genetically-altruistic traits might underlie them, my ethnographic impression was
that because people seem so susceptible to messages calling for altruism, their willingness to be
generous might, in fact, have some genetic basis aside from merely being culturally docile. Likewise, the fact that small human groups so very predictably come up with prosocial messages favoring altruistic generosity could offer further indirect evidence that people actually carry some
26
innate traits favoring altruism (Boehm 2004a). It is at least logical that it should be much easier to
persuade others to be altruistic if both parties harbored some genetic altruistic tendencies.
Of course, in theory group selection could explain such tendencies—if it was strong
enough. However, if we turn to positive sanctioning selection, as a parsimonious and potentially
powerful model for the genetic selection of altruistically-generous traits, we have an alternative
that can easily work in conjunction with whatever group-selection forces may exist. The main
concerns with group selection have been its apparent lack of robustness (Wilson 1975)—in conjunction with free-rider problems (Williams 1966). However, I have shown elsewhere that today’s
Pleistocene-appropriate bands are quite good at discouraging free-riders, and when they fail to do
so they usually can punish or eliminate them (Boehm 1997).
The various types of explanation discussed above are far from being incompatible, and
what we are talking about here is a complex process of gene-culture coevolution that has kept cooperation around meat-sharing in place for at least 45,000 years. Thus, ultimately positive sanctioning selection and group selection could have been working in tandem, at the level of genotype,
to make humans better able to cooperate around meat. At the same time, at the level of immediate
cultural dynamics, both positive and negative social sanctioning were directly supporting such cooperation on an everyday basis.
IS RUNAWAY SANCTIONING SELECTION LIKELY?
According to Andersson (1994:25), it is the escalating interactions of preferences and preferred traits that can help natural selection to reach a “runaway” status:
Fisher (1930) suggested that “a sexual preference of a particular kind may
confer a selective advantage, and therefore become established in the species.” He envisaged a two-step process. Suppose there arises genetic variation
in a male trait such as tail length, and that males with, say, a longer than average tail have a slight survival advantage, for example owing to improved
agility. Assume also that females choose mates, and that there is genetic variation among females in their tendency to mate with males of different tail
lengths. Females preferring males with long tails tend to bear sons with high
27
survival. Hence alleles that code for longer tails in males will spread, and so
will alleles that make females prefer long-tailed males, as the two types of alleles become associated in their offspring.
When this process continues, a new effect grows in importance: males with
long tails are favored not only by better survival, but also by higher mating
success as the preference for long tails spreads among females. The higher
mating success of long-tailed males helps carry the associated alleles for the
long tail and the female preference to yet higher frequency, and a feedback
“runaway process” develops at accelerating pace. Although Fisher did not
explain fully how this would happen, he suggested that females will prefer
tails of ever-increasing length, until it finally becomes so long that higher
mortality balances the mating advantage of long-tailed males. This brings the
process to rest.
In defining this runaway process, (Fisher 1930) concentrated on mate selection,
which favored secondary sexual traits like male peacocks’ tails. It was suggested that
these sexual-selection pressures, generated by female choice, were so powerful that they
might “carry” some noteworthy deleterious effects associated with the preferred traits.
With his model, this was true up to the point that the traits became so “exaggerated”
that costs began to mount up too much, as with higher vulnerabilities of the larger-tailed
peacocks to predators. At that point the intensive escalating genetic interaction between
the choosers and the carriers of chosen traits could go no further, and an equilibrium
was reached.
What is distinctive about this runaway process is that the selection is not only
social, but interactive and—up to a point—escalating. More recent runaway scenarios
have been developed, which focus on choosing socially-attractive altruistic traits in a
potential breeding partner (Breden and Wade 1991; Nesse 2007), as opposed to responding to secondary sexual characteristics. The processes are similar, for if the pattern of social choice is both consistent and potent as a means of gene selection, and if
both the decisionmakers and the individuals carrying favored traits are both profiting
reproductively, then the selection process might, likewise, “run away”—up to a point.
28
Of course, in humans culture is heavily involved. And in a cultural context, it is not difficult to conceive how this type of selection in favor of more and more altruism might have started.
All that was needed was for purposeful common sense to guide the choices of people in small
bands, in which personal characteristics of group members were well known because of gossiping.
A wise chooser would try to ally with people who were exceptionally reliable or industrious or
generous, not only in marriage but in other important types of association in which cooperation
was involved. In this context a stable cultural perspective would have informed a regular pattern of
decisionmaking, and altruistic traits would have been selected positively—as long as, over time,
the stronger altruists, as choice-magnets, were gaining a net advantage over others. However, if an
individual was predisposed to behave so altruistically that the losses became greater than the
gains, genes promoting that degree of altruism would not have increased in frequency and the runaway-selection process would have escalated as far as it could.
What one sees on the ground is in accordance with what this genetic model predicts. Empirically, the generosity of hunter-gatherers to genetic strangers in the same group is, in fact, obvious—but limited. Sharing within the family, which in its genetic basics is readily explained just by
nepotism, is sufficiently unambivalent that it can take place strongly and without a great deal of
social reinforcement. Quite predictably, extrafamilial generosity involves much more ambivalence. That is why “demand-sharing” patterns appear so widely, and why cheaters sometimes have
to be punished. However, it would be a mistake to assume that routine bickering and occasional
cheating attempts mean altruistic tendencies must be entirely absent.
With the help of moralistic messages and sanctions, this extrafamilial generosity tends to
be expressed actively whenever large game are being taken sporadically on a routine basis. However, when famine arrives this system of cooperation among unrelated families is prone to collapse
as selfish nepotism wins out, and does so long before the intrafamilial sharing system collapses. In
29
the rare cases where famine reactions have been well documented for other types of humans, even
intrafamilial sharing patterns can diminish or disappear if hunger becomes dire (see Laughlin and
Brady 1978), and among Netsilik Inuit-speakers people may even resort to cannibalism within the
family (Balikci 1970), as an indigenously abhorrent alternative to sharing.
Overall, what Pleistocene-appropriate hunter-gatherers do in their groups fits quite well
with a positive sanctioning-selection model that sees gene selection as being “guided” by decision
biases that favor generosity. Individuals in their bands are innately generous enough toward nonkin to permit a culturally-based system of indirect reciprocity to flourish (see Alexander 1987),
and this does involve making donations to nonkin when exact reciprocation (Trivers 1971) is extremely unlikely. Indeed, in such “insurance systems” (e.g., Wiessner 1982) positive “payback”
very often comes from a different party than the one being helped (Alexander 1987). At the same
time, though, this remarkable generosity among nonkin is far more limited than is sharing within
families. These patterns fit with a process of purposive social selection which is strong enough to
support genetically altruistic traits to the point that the costs begin to get out of hand. At that point,
the process can go no further.
Could All Altruistic Behavior Be Dissembled?
Table I suggests a great deal about the likely fitness effects of punitive sanctioning, especially capital punishment, but relatively little about positive sanctioning’s direct effects. This is
because benefits accruing to altruists as leaders or as economic or breeding partners are not nearly
as obvious ethnographically. There is also another potential problem, that of trying to differentiate
helpful behavior that is “genuinely” altruistic from helpful behavior that is merely expedient because it is aimed solely at reputation enhancement. The latter has been discussed at length by Alexander (1987), and he sees such dissembling as an evolved trait that is useful to inclusive fitness
30
and requires no genetic altruism. In fact, he believes that all generosity toward nonkin may well be
of this dissembling type.
I would agree that we are likely to have been naturally-selected to dissemble generosity;
for instance, the self-serving presentation of self in everyday life (see Goffman 1959) is widely
noted by ethnographers. However, in Alexander’s (1987) opinion the only (slim) hope for innately-genuine (as opposed to dissembled) altruistic generosity’s having evolved would have been
strong group selection taking place in the Pleistocene through band-level warfare and group extinctions. This may be conceivable, but in that epoch frequent outright band-level extinctions due
to warfare seem less than likely (see Kelly 2000).
More recent versions of group-selection theory do allow for significant trait-group selection in the absence of outright group extinctions (e.g., Sober and Wilson 1998), so in fact it is
quite likely that genetically speaking group selection has been contributing some, and possibly
significantly to our “altruistic quotient.”. But it remains a challenge to estimate exactly how
strongly such effects have operated (see Bowles 2006), and whether they alone could account for
the rather striking (if limited) degrees of altruism we see in nonliterate foragers’ everyday behavior, and in our own lives as well.
What about dissembling, then? In my view it was obvious to group members that being
seen as exceptionally generous would make others think better of them, so they put to work not
only their genuinely-generous impulses in this direction—but also a sly capacity to “showcase”
such generous actions publicly and sometimes even to fake them. Both real and feigned generosity
facilitated cooperation, for in either case the behavior was helpful to others, and to a large-game
subsistence. Thus, social-selection processes would have favored innate, non-dissembled generosity at the same time that they were favoring a convincing capacity to showcase one’s generous acts
in the interest of reputation enhancement. I might add that to the extent that the choosers could
31
make this discrimination, this social-selection process would have been biased generally toward
non-dissembled generosity.
THE QUESTION OF PURPOSE
With respect to the larger picture, this exploratory treatment suggests that the phrase “intelligent design” should be applied to modern genetic engineering, and to it alone. However, this
does not mean that all evolutionary process is teleonomic from bottom to top. When females of an
animal species react favorably to the displays of certain males, they are likely to be either choosing instinctively (Darwin 1871) or mechanically imitating the choices of other females (e.g., Dugatkin 1992). However, when Montenegrin Serbs arrange for their sons to marry the tallest women
available, the immediate choices involve some limited insight into biological-evolutionary process. This makes the Montenegrins’ choices an instance of unconscious selection as Darwin defined it.
In contrast, with both positive and negative sanctioning selection the decision biases are
simply aimed at improving group or individual social life—with no thought at all being given to
manipulating heredity. While this is not quite “unconscious selection” as Darwin defined it, these
purposeful decision biases are, in fact, affecting gene pools in directions that are consistent with
the purposes involved. Thus, the mechanisms I have been discussing might be called “totally unconscious selection,” in that decision biases are guiding selection process in the absence of any
understanding of hereditary issues.
While such behavior may not be limited to humans, I believe that our capacity to develop
an ethos at the group level, and to deliberately sanction group members both negatively and positively, is distinctive both in the degree to which it is cultural and also in its being elaborated symbolically. The cultural symbols involved carry not only meaning but intentionality, for people in
bands care about their groups and discuss how well they are functioning (see Sober and Wilson
32
1998), are quick to judge their fellows in this light, and feel empowered to actively manipulate
group life.
Our main focus has been on positive sanctioning selection, which might also be called
“positive social selection.” This involves group attitudes about what constitutes good behavior,
manipulative moralizing messages in favor of behavior that is generous or otherwise desirable,
gossiping to evaluate people in this light, and, finally, the choices made by individuals, families,
and entire bands which reward such behaviors. I have tried to show that a longstanding and purposeful cultural emphasis on generous behavior has impacted both phenotype and genotype in
very much the same direction.
This wholly unconscious selection process should have had a consistent and powerful impact on our gene pools because presumably it was so deliberate, focused, and consistent. If runaway genetic processes have been involved, the impact would have been greater still. Surely one
result of all this deliberate positive sanctioning, with its possible runaway effects, has been a species that can individually showcase its altruistically-generous behaviors in a self-interested fashion. But another result has been the generous behavior, itself. It appears that such purposive sanctioning selection has had substantial effects on human gene pools, and in this light it seems expedient to take a fresh look at the altruism paradox—as well as at the types of evolutionary “causality” we are willing to talk about.
Indeed, scientific evolutionists may wish to admit some shades of grey into their analyses,
with respect to “cause” and “purpose” as these are inherent in social decision processes. Exactly
how to incorporate these limited purposes into the usual teleonomic explanatory mode remains a
question. Could we treat decisions as an “intelligent” type of variation (e.g., Campbell 1983), but
leave the rest of the standard paradigm unchanged? Or for humans is some more radical paradigmatic adjustment needed? A full exploration of these questions is beyond the scope of the present
33
treatment. However, the complex interactions of genes and cultures we have been discussing may
provide a useful case study for further exploration.
This model involves not only gene-culture evolution (e.g., Boyd and Richerson 1985;
Durham 1991), but a purposeful and socially-sophisticated mind which with help from symbolic
communication creates potent and highly “directional” selection forces. With respect to this coevolutionary dimension, Baldwin (1896; see also Ananth 2005; Griffiths 2003; Weber and Depew
2003) made suggestions about the involvement of culturally-involved developmental factors with
natural selection process, which in modern terms may be considered under the rubric of “epigenetic” processes (see Lumsden and Wilson 1980) or, perhaps, “canalization” (Waddington 1942). In
addition, niche-construction theory (Odling-Smee 1988; see also Laland et al 2001; Sterelny 2001)
has looked to the impact of resident species on environments as well as vice-versa, and, in this
context the decisions we have discussed help to create social-environmental effects which in turn
help to determine the course of natural selection.
This case history involving positive sanctioning may be useful for testing any of these theories, and in doing so it may be beneficial to raise the question of exactly how purposive minds
may be articulating with evolutionary process, overall.
CONCLUSIONS
For decades, cultural anthropology has been preoccupied with issues of “cultural diversity,” with relatively little attention going to universals—many of which are so obvious that we
simply take them for granted. This is unfortunate because universals (e.g., Brown 1991) provide a
special arena in which the complex relations between cultures and gene pools can be explored. In
this paper I have given certain hunter-gatherer universals some rather intensive scrutiny, in the interest of seeing whether human social intentions could, in fact, have been guiding natural-selection
process in ways that were significant.
34
In the context of both positive and negative sanctioning selection, it would appear that such
“guidance” can, in effect, select in favor of genetic outcomes which might not have come about
otherwise. As a product of decision biases at both group and individual levels, this limited but significant evolutionary “directionality” has provided genetic selection process with an influence
which did not stem just from teleonomic selection pressures originating in natural environments—
which, of course, are still the main past determinants of the content of our gene pools.
When species are highly social, this means they are constantly making choices about
whom to associate with, and this results in potentially powerful social selection effects which have
a strong and consistent “direction.” Because humans have moral communities and individual moral reputations, it would appear that some truly unusual possibilities for social selection have arisen
in our species—and that the power of such selection may be sufficient to keep certain altruistic
traits in place.
This would particularly be the case when selection becomes “runaway,” and I have suggested that positive sanctioning selection in particular might fit quite well with Fisher’s process
because mate selection is involved, along with economic-partner selection. Punitive sanctioning
selection also may be a candidate for applying Fisher’s model, even though it is aversions rather
than attractions that are driving decision processes, and even though many of the decisions are collective, which potentially introduces another level of selection. In combination, these two types of
sanctioning selection can explain a great deal about the evolution of our human capacity to cooperate.
Both methodologically and theoretically this paper has been exploratory. Yet I would suggest that human social intentions, expressed in the form of group and individual decisions, can impinge significantly on the process of gene selection. They can do so to a degree not expectable in
any animal that lacks purposeful, morally-catalyzed social control. The reduction of antisocial
35
traits and the evolution of a conscience are two very likely results, while the positive socialselection hypothesis we have concentrated on can explain something very important about altruism’s evolutionary underpinnings.
This hypothesis requires some further development, with more extensive and more rigorous ethnographic coverage combined with appropriate mathematical modeling and simulationtesting. However, such exploration should be worthwhile simply because human altruism and altruistic cooperation remain provocative and sometimes contentious evolutionary issues. It would
also be useful because socially rewarding altruists provides such a good example of how natural
selection might be driven by longstanding decision biases, largely shaped by group mores, which
were also modifying behavioral dispositions in similar directions—even though directed at phenotypic behavior were also modifying behavioral dispositions in similar directions. If this type of
lower-level teleology is, in fact, making a significant difference for selection outcomes, I believe
this to be more than just a question for the philosophers.
36
ACKNOWLEDGEMENTS
I thank David Krakauer, and this journal’s three blind reviewers and its editor, for useful
comments and guidance. I also am grateful to the John Templeton Foundation for two grants
which supported coding ethnographic materials, and to the Goodall Center at USC and USC’s Department of Computer Sciences, which also have supported the development of the huntergatherer database.
REFERENCES CITED
Alexander, Richard D.
1974 The Evolution of Social Behavior. Annual Review of Ecology and Systematics
5:325-384.
1987 The Biology of Moral Systems. New York: Aldine de Gruyter.
NC Ananth, Mahesh
2005 Psychological Altruism vs. Biological Altruism: Narrowing the Gap with the Baldwin Effect. Acta Biotheoretica 53:217-239.
Andersson, Malte
1994 Sexual Selection. Princeton: Princeton University Press.
NC Baldwin, James M.
1896 A New Factor in Evolution. American Naturalist 30:441-451.
Balikci, Asen
1970 The Netsilik Eskimo. Prospect Heights: Waveland Press.
Bettinger, Robert L.
1991 Hunter-Gatherers: Archaeological and Evolutionary Theory. New York: Plenum
Press.
Boehm, Christopher
1978 Rational Pre-Selection from Hamadryas to Homo sapiens: The Place of Decisions in
Adaptive Process. American Anthropologist 80:265-296.
1981 Parasitic Selection and Group Selection: A Study of Conflict Interference in Rhesus
and Japanese Macaque Monkeys. In, Primate Behavior and Sociobiology: Proceedings of
the International Congress of Primatology. A. B. Chiarelli and R. S. Corruccini, eds., pp.
161-182. Heidelberg: Springer.
1982a A Fresh Outlook on Cultural Selection. American Anthropologist 84:105-124.
1982b The Evolutionary Development of Morality as an Effect of Dominance Behavior
and Conflict Interference. Journal of Social and Biological Sciences 5:413-422.
1983 Montenegrin Social Organization and Values. New York: AMS Press.
1986 Blood Revenge: The Enactment and Management of Conflict in Montenegro and
Other Tribal Societies. Philadelphia: University of Pennsylvania Press.
1991 Lower-Level Teleology in Biological Evolution: Decision Behavior and Reproduc37
tive Success in Two Species. Cultural Dynamics 4:115-134.
1993 Egalitarian Society and Reverse Dominance Hierarchy. Current Anthropology
34:227-254.
1996 Emergency Decisions, Cultural Selection Mechanics, and Group Selection Current
Anthropology 37:763-793.
1997 Impact of the Human Egalitarian Syndrome on Darwinian Selection Mechanics.
American Naturalist 150:100-121.
1999 Hierarchy in the Forest: The Evolution of Egalitarian Behavior. Cambridge: Harvard
University Press.
2000 Conflict and the Evolution of Social Control. In, Journal of Consciousness Studies
7:79-183. Special issue on Evolutionary Origins of morality. Leonard Katz, editor.
2002 Variance Reduction and the Evolution of Social Control. Paper presented at Santa Fe
Institute, 5th Annual Workshop on the Co-Evolution of Behaviors and Institutions, November 2002, Santa Fe. (Posted at Santa Fe Institute Web site).
2004a Explaining the Prosocial Side of Moral Communities. In, Evolution and Ethics:
Human Morality in Biological and Religious Perspective. P. Clayton and J. Schloss, eds.,
pp. 78-100. New York: Eerdmans.
2004b Large-Game Hunting and the Evolution of Human Sociality. In, Origins and Nature
of Sociality among Nonhuman and Human Primates. R. W. Sussman and A. R. Chapman,
eds., pp. 270-287. New York: Aldine.
2004c What Makes Humans Economically Distinctive? A Three-Species Evolutionary
Comparison and Historical Analysis. Journal of Bioeconomics 2:109-135.
Submitted Conscience Origins, Sanctioning Selection, and the Evolution of Altruism in
Homo sapiens. Article Submitted to Behavioral and Brain Sciences.
Bowles, Samuel, and Herbert Gintis
2004 The Evolution of Strong Reciprocity: Cooperation in Heterogeneous Populations.
Theoretical Population Biology 65:17-28.
Bowles, Samuel
2006 Group Competition, Reproductive Leveling, and the Evolution of Human Altruism.
Science 314:1569-1572.
Boyd, Robert, and Peter J. Richerson
1982 Cultural Transmission and the Evolution of Cooperative Behavior. Human Ecology
10:325-351.
1985 Culture and the Evolutionary Process. Chicago: University of Chicago Press.
1992 Punishment Allows the Evolution of Cooperation (or Anything Else) in Sizable
Groups. Ethology and Sociobiology 13:171-195.
Breden, F. and M. J. Wade
1991 Runaway Social Evolution: Reinforcing Selection for Inbreeding and Altruism.
Journal of Theoretical Biology 153:323-337.
Briggs, Jean L.
1970 Never in Anger: Portrait of an Eskimo Family. Cambridge: Harvard University
Press.
Brown, Donald
1991 Human Universals. New York: McGraw-Hill.
Burch, Ernest S., Jr.
2005 Alliance and Conflict: The World System of the Iñupiaq Eskimos. Lincoln: University of Nebraska Press.
38
Campbell, Donald T.
1965 Variation and Selective Retention in Socio-Cultural Evolution In, Social Change in
Developing Areas. H. R. Barringer, B. I. Blanksten, and R. W. Mack, eds., pp. 19-49.
Cambridge, MA: Schenkman.
1983 The Two Distinct Routes Beyond Kin Selection to Ultrasociality: Implications for
the Humanities and Social Sciences. In, The Nature of Prosocial Development: Interdisciplinary Theories and Strategies. D. Bridgeman, ed., pp. 11-41. New York: Academic Press.
1991 A Naturalistic Theory of Archaic Moral Orders. Zygon 26:91-114.
Cordain, Loren, Janette Brand Miller, S. Boyd Eaton, Neil J. Mann, Susanne H. A. Holt, and John
D. Speth
2000 Plant-Animal Subsistence Ratios and Macronutrient Energy Estimations in Worldwide Hunter-Gatherer Diets. American Clinical Nutrition 71:682-692.
Cordain, Loren, Bruce A Watkins, and Neil J. Mann
2001 Fatty Acid Composition and Energy Density of Foods Available to African Hominids: Evolutionary Implications for Human Brain Development. World Review of Nutrition and Diet 90:144-161.
Cordain, L., B. A. Watkins, G. L. Florant, M. Kelher, L. Rogers, and Y. Li
2002 Fatty Acid Analysis of Wild Ruminant Tissues: Evolutionary Implications for Reducing Diet-Related Chronic Disease. European Journal of Clinical Nutrition 56:1-11.
Damasio, Antonio R., Daniel Tranel and Hanna Damasio
1990 Individuals with Sociopathic Behavior Caused by Frontal Damage Fail to Respond
Automatically to Social Stimuli. Behavioral Brain Research 41:81-94.
Darwin, Charles
1859 On the Origin of Species. London: John Murray.
1871 The Descent of Man and Selection in Relation to Sex. London: John Murray.
Diamond, Jared M.
1997 Guns, Germs and Steel: The Fates of Human Societies. New York: W. W. Norton
and Company.
Dugatkin Lee Alan
1992 Sexual Selection and Imitation: Females Copy the Mate Choice of Others.
American Naturalist 139:1384-1389.
Durham, William H.
1991 Coevolution: Genes, Culture, and Human Diversity. Stanford: Stanford University
Press.
Durkheim, Émile
1933 The Division of Labor in Society. New York: Free Press.
Ehrich, Robert
1946 A Racial Analysis of Montenegro. Doctoral dissertation, Harvard University.
Eldridge, Niles
1971 The Allopatric Model and Phylogeny in Paleozoic Invertebrates. Evolution 25:156167.
Fisher, Ronald A.
1930 The Genetical Theory of Natural Selection. New York: Dover.
Frank, Stephen A.
2005 Social Selection. In, Evolutionary Genetics: Concepts and Case Studies. C. W. Fox
and J. B. Wolf, eds., pp. 350-363. Oxford: Oxford University Press.
Fry, Douglas P.
2000 Conflict Management in Cross-Cultural Perspective. In, Natural Conflict Resolution.
39
F. Aureli and F. B. M. de Waal, eds., pp. 334-351. Berkeley: University of California
Press.
Gintis, Herbert
2003 The Hitchhiker's Guide to Altruism: Gene-Culture Coevolution and the Internalization of Norms. Journal of Theoretical Biology 220:407-418.
2004 The Genetic Side of Gene-culture Coevolution: Internalization of Norms and Prosocial Emotions. Journal of Economic Behavior and Organization 53:57-67.
Goffman, Erving
1959 The Presentation of Self in Everyday Life. New York: Anchor Books.
NC Griffiths, Paul E.
2003 Beyond the Baldwin Effect: James Mark Baldwin’s “Social Heredity,” Epigenetic
Inheritance and Niche-Construction. In, Learning, Meaning, and Emergence: Possible
Baldwinian Mechanisms in the Co-evolution of Mind and Language. B. Weber and D. Depew, eds., pp. 193-215. Cambridge: MIT Press.
Hare, Brian, Irene Plyusnina, Natalie Ignacio, Olesya Schepina, Anna Stepika, Richard Wrangham, and Lyudmila Trut
2005 Social Cognitive Evolution in Captive Foxes is a Correlated By-Product of Experimental Domestication. Current Biology 15:226-230.
Haviland, John B.
1977 Gossip, Reputation, and Knowledge in Zinacantan. Chicago: University of Chicago
Press.
Hodgson, Geoffrey, and Thorbjørn Knudsen
2006 The Nature and Units of Social Selection. Journal of Evolutionary Economics
16:477-489.
Kelly, Robert L.
1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington D.C.:
Smithsonian.
Kelly, Raymond C.
2000 Warless Societies and the Evolution of War. Ann Arbor: University of Michigan
Press.
Klein, Richard G.
1999 The Human Career: Human Biological and Cultural Origins. Chicago: University of
Chicago Press.
Knauft, Bruce M.
1991 Violence and Sociality in Human Evolution. Current Anthropology 32:391-428.
NC Laland, K. N., F. J. Odling-Smee, and M. W. Feldman
2001 Niche Construction, Ecological Inheritance, and Cycles of Contingency in Evolution. In, Cycles of Contingency: Developmental Systems and Evolution. S. Oyama, P. E.
Griffiths and R. D. Gray, eds., pp. 117-126. Cambridge: MIT Press.
Laughlin, Charles D., and Ivan A. Brady, eds.
1978 Extinction and Survival in Human Populations. New York: Columbia University
Press.
Lumsden, Charles J., and Edward O. Wilson
1980 Translation of Epigenetic Rules of Individual Behavior into Ethnographic Patterns.
Proceedings of the National Academy of Sciences 77:4382-4386.
Maine, Sir Henry
1861 Ancient Law: Its Connection with the Early History of Society and its Relation to
Modern Ideas. London: John Murray.
40
Marlowe, Frank W.
2005 Hunter-Gatherers and Human Evolution. Evolutionary Anthropology 14:54-67.
Mayr, Ernst
1983 How to Carry Out the Adaptationist Program? American Naturalist 121:324-334.
Nesse, Randolph M.
2007 Runaway Social Selection for Displays of Partner Value and Altruism. Biological
Theory 2:143-155.
Odling-Smee, F. J.
1988 Niche-Constructing Phenotypes. In, The Role of Behavior in Evolution. H. C. Plotkin, ed., pp. 73-132. Cambridge: MIT Press.
Otterbein, Keith
1988 Capital Punishment: A Selection Mechanism. Comment on Robert K. Dentan, On
Semai Homicide. Current Anthropology 29:633-636.
Palmer, Craig T., B. Eric Fredrickson, and Christopher F. Tilley
1997 Categories and Gatherings: Group Selection and the Mythology of Cultural Anthropology. Ethology and Sociobiology 18: 291-308.
Peterson, Nicolas
1993 Demand Sharing: Reciprocity and the Pressure for Generosity among Foragers.
American Anthropologist 95:860-874.
Piers, Gerhart and Milton B. Singer
1971 Shame and Guilt: A Psychoanalytic and a Cultural Study. New York: Norton.
Simon, Herbert
1990 A Mechanism for Social Selection and Successful Altruism. Science 250:1665-1668.
Sober, Elliott, and David S. Wilson
1998 Unto Others: The Evolution and Psychology of Unselfish Behavior. Cambridge:
Harvard University Press.
Stanford, Craig B.
1999 The Hunting Apes: Meat Eating and the Origins of Human Behavior. Princeton:
Princeton University Press.
Stark, W.
1961 Natural and Social Selection. In, Darwinism and the Study of Society: A Centenary
Symposium. M. Banton, ed., pp. 49-61. London: Tavistock Publications.
NC Sterelny, Kim
2001 Niche Construction, Developmental Systems, and the Extended Replicator. In, Cycles of Contingency: Developmental Systems and Evolution. S. Oyama, P. E. Griffiths and
R. D. Gray, eds., pp. 333-350. Cambridge: MIT Press.
Steward, Julian H.
1955 Theory of Culture Change. Urbana: University of Illinois Press.
Stiner, Mary C.
2002 Carnivory, Coevolution, and the Geographic Spread of the Genus Homo. Journal of
Archeological Research 10:1-63.
Tanaka, Yoshinari
1996 Social Selection and the Evolution of Animal Signals. Evolution 50:512-523.
Trivers, Robert L.
1971 The Evolution of Reciprocal Altruism. Quarterly Review of Biology 46:35-57.
1972 Parental Investment and Sexual Selection. In Sexual Selection and the Descent of
Man. B. Campbell, ed. Pp. 136–179. Chicago: Aldine-Atherton.
1985 Social Evolution. Menlo Park: Benjamin/Cummings.
41
Turner, Jonathan H.
2000 On the Origins of Human Emotions: A Sociological Inquiry into the Evolution of
Human Affect. Stanford: Stanford University Press.
van Woudenberg, Rene
2004 Is There an Evolutionary Foundation for Human Morality? In, Evolution and Ethics:
Human Morality in Biological and Religious Perspective. P. Clayton and J. Schloss, eds.,
pp. 187-203. New York: Eerdmans.
Vila, Caries, Peter Savolainen, Jesus E. Maldonado, Isabel R. Amorim, John E. Rice, Rodney L.
Honeycutt, Keith A. Crandall, Joakim Lundeberg, Robert K. Wayne
1997 Multiple and Ancient Origins of the Domestic Dog. Science 276:1687-1689.
Waddington, Charles H.
1942 Canalization of Development and the Inheritance of Acquired Characters. Nature
150:563-565.
1960 The Ethical Animal. Chicago: University of Chicago Press.
NC Weber, Bruce H., and David J. Depew, eds.
2003 Evolution and Learning: The Baldwin Effect Reconsidered. Cambridge: MIT Press.
Wiessner, Polly
1982 Risk, Reciprocity and Social Influences on !Kung San Economics. In, Politics and
History in Band Societies. E. Leacock and R. B. Lee, eds., Pp. 61-84. Cambridge: Cambridge University Press.
1996 Leveling the Hunter. In, Food and the Status Quest: An Interdisciplinary Perspective. P. Wiessner and W. Schiefenhövel, eds., pp. 171-191. Oxford: Berghahn Books.
2005 Norm Enforcement among the Ju/’hoansi Bushmen: A Case of Strong Reciprocity?
Human Nature 16:115-145.
Wilkinson, Sir Gardner
1848 Dalmatia and Montenegro, Volume 1. London: John Murray.
Williams, George C.
1966 Adaptation and Natural Selection: A Critique of some Current Evolutionary
Thought. Princeton: Princeton University Press.
Wilson, David S. and Edward O. Wilson
2007 Rethinking the Theoretical Foundation of Sociobiology. Quarterly Review of Biology 82:327-348.
Wilson, Edward O.
1975 Sociobiology: The New Synthesis. Cambridge: Harvard University Press.
1978 On Human Nature. Cambridge: Harvard University Press.
Winterhalder, Bruce
2001 Intragroup Resource Transfers: Comparative Evidence, Models, and Implications for
Human Evolution. In, Meat-Eating and Human Evolution. C. B. Stanford and H. T. Bunn,
eds., pp. 279-301. New York: Oxford University Press.
Winterhalder, Bruce, and Eric A. Smith, eds.
1981 Hunter-Gatherer Foraging Strategies: Ethnographic and Archeological Analyses.
Chicago: University of Chicago Press.
Wolf, Jason B., Edmund D. Brodie III, and Allen J. Moore
1999 Interacting Phenotypes and the Evolutionary Process, II: Selection Resulting from
Social Interactions. American Naturalist 153:254-266.
Wrangham, Richard
42
2002 The Evolution of Cooking: A Talk with Richard Wrangham. Edge, February 29,
2002. http://www.edge.org/documents/archive/edge98.html.
Wrangham, Richard, and Dale Peterson
1996 Demonic Males: Apes and the Origins of Human Violence. New York: Houghton
Mifflin.
43