ArticlePDF AvailableLiterature Review

Evolutionary Psychology: New Perspectives on Cognition and Motivation



Evolutionary psychology is the second wave of the cognitive revolution. The first wave focused on computational processes that generate knowledge about the world: perception, attention, categorization, reasoning, learning, and memory. The second wave views the brain as composed of evolved computational systems, engineered by natural selection to use information to adaptively regulate physiology and behavior. This shift in focus-from knowledge acquisition to the adaptive regulation of behavior-provides new ways of thinking about every topic in psychology. It suggests a mind populated by a large number of adaptive specializations, each equipped with content-rich representations, concepts, inference systems, and regulatory variables, which are functionally organized to solve the complex problems of survival and reproduction encountered by the ancestral hunter-gatherers from whom we are descended. We present recent empirical examples that illustrate how this approach has been used to discover new features of attention, categorization, reasoning, learning, emotion, and motivation.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Evolutionary Psychology:
New Perspectives on
Cognition and Motivation
Leda Cosmides1and John Tooby2
1Department of Psychological & Brain Sciences and Center for Evolutionary Psychology
and 2Department of Anthropology and Center for Evolutionary Psychology, University of
California, Santa Barbara, California 93106; email:,
Annu. Rev. Psychol. 2013. 64:201–29
The Annual Review of Psychology is online at
This article’s doi:
Copyright c
2013 by Annual Reviews.
All rights reserved
motivation, domain-specificity, evolutionary game theory, visual
attention, concepts, reasoning
Evolutionary psychology is the second wave of the cognitive revolu-
tion. The first wave focused on computational processes that gener-
ate knowledge about the world: perception, attention, categorization,
reasoning, learning, and memory. The second wave views the brain
as composed of evolved computational systems, engineered by natu-
ral selection to use information to adaptively regulate physiology and
behavior. This shift in focus—from knowledge acquisition to the adap-
tive regulation of behavior—provides new ways of thinking about every
topic in psychology. It suggests a mind populated by a large number of
adaptive specializations, each equipped with content-rich representa-
tions, concepts, inference systems, and regulatory variables, which are
functionally organized to solve the complex problems of survival and re-
production encountered by the ancestral hunter-gatherers from whom
we are descended. We present recent empirical examples that illustrate
how this approach has been used to discover new features of attention,
categorization, reasoning, learning, emotion, and motivation.
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
Click here for quick links to
Annual Reviews content online,
Other articles in this volume
Top cited articles
Top downloaded articles
• Our comprehensive search
Fur ther
PS64CH08-Cosmides ARI 11 November 2012 8:40
INTRODUCTION.................. 202
VISUALATTENTION.............. 205
Animal Monitoring: An Appendix
inVisual Attention?............. 205
Automatic Regulation of Attention
by High-Level Social Cues . . . . . . 207
NAVIGATION................... 208
Spatial Specializations
forForaging.................... 209
SOCIALBEHAVIOR ............. 210
The Evolution of Two-Party
Cooperation: Constraints from
GameTheory .................. 211
CollectiveAction .................. 212
CATEGORIZATION............. 213
Concepts for Collective Action:
Free Riders Versus
Cooperators . . . . . . . . . . . . . . . . . . . . 213
REASONING........................ 215
Conditional Reasoning and
Social Exchange . . . . . . . . . . . . . . . . 215
Investigations with the Wason
SelectionTask.................. 216
VARIABLES...................... 218
Internal Regulatory Variables . . . . . . . 218
Genetic Relatedness and
Motivation: Siblings, Incest,
andAltruism.................... 219
AKin DetectionSystem............ 219
WelfareTrade-Offs................ 223
Anger and the Recalibration of
Welfare Trade-Off Ratios . . . . . . . 223
CONCLUSION..................... 224
Both before and after Darwin, a common
view among philosophers and scientists has
been that the human mind resembles a blank
slate, virtually free of content until written
on by the hand of experience. Over the years,
the technological metaphor used to describe
the structure of the human mind has been
consistently updated, from blank slate to
switchboard to general-purpose computer, but
the deeper assumption remained. The impli-
cations are wide ranging. According to this
view, the mechanisms that produce learning
operate in the same way, whether they are
acquiring the grammar of a language, a fear of
snakes, or an aversion to sex with siblings. The
mechanisms that produce reasoning deploy
the same procedures, whether they are making
inferences about the trajectory of a billiard
ball, the beliefs and desires of another person,
or what counts as cheating in social exchange.
The same goes for attention, categorization,
memory, motivation, and decision making.
This perspective grants that evolution may
have equipped the mind with a few primary re-
inforcers that have hedonic value (food, water,
pain avoidance, sex). But it assumes that the
neurocomputational systems that collect and
process experiences are largely content free and
domain general, designed to operate uniformly
on information drawn from any stimulus class
(cf. Herrnstein 1977, Gallistel 1995).
A very different picture of the human mind is
emerging from evolutionary psychology, an ap-
proach to the cognitive sciences that integrates
evolutionary biology, psychology, information
theory, anthropology, cognitive neuroscience,
and allied fields (for reviews, see Barkow et al.
1992, Buss 2005). In this view, human nature—
the species-typical information-processing
architecture of the human brain—is packed
with content-rich adaptive problem-solving
systems. Like expert systems (in artificial intel-
ligence), each is designed to deploy different
concepts, principles, inference procedures,
regulatory variables, and decision rules when
activated by cues of its proper domain. Why?
202 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
From this perspective, the cognitive and
evolutionary sciences are connected as follows:
1. Each organ in the body evolved to serve
a function: the intestines digest, the heart
pumps blood, the liver detoxifies poisons.
The brain is also an organ, and its evolved
function is to extract information from
the environment and use that information
to generate behavior and regulate phys-
iology. From this perspective, the brain
is a computer, that is, a physical system
that was designed to process information.
Its programs were designed not by an en-
gineer, but by natural selection, a causal
process that retains and discards design
features on the basis of how well they
solved problems that affect reproduction
(Williams 1966, Dawkins 1982).
The fact that the brain processes in-
formation is not an accidental side ef-
fect of some metabolic process: The brain
was designed by natural selection to be a
computer. Therefore, if you want to de-
scribe its operation in a way that cap-
tures its evolved function, you need to
think of it as composed of programs
that process information. The question
then becomes, what programs are to be
found in the human brain? What are the
reliably developing, species-typical pro-
grams that, taken together, constitute the
human mind?
2. These programs were sculpted over
evolutionary time by the ancestral
environments and selection pressures ex-
perienced by the hunter-gatherers from
whom we are descended. Each evolved
program exists because it produced
behavior that promoted the survival and
reproduction of our ancestors better
than alternative programs that arose
during human evolutionary history.
Evolutionary psychologists emphasize
hunter-gatherer life because it takes a
long time for natural selection to build
a computational adaptation of any com-
plexity. Simple, quantitative traits can
change faster, but it takes thousands of
adaptations: evolved
systems designed (by
natural selection) to
monitor information
and use it to
functionally regulate
behavior or physiology
Environment of
Adaptedness: the
series of ancestral en-
pressures that sculpted
the design of an
Replicator dynamics:
how genes change in
frequency in a
years (i.e., many human generations) for
natural selection to assemble a complex
program composed of many different,
functionally integrated parts (Tooby &
Cosmides 1990a).
3. Although the behavior our evolved pro-
grams generate would, on average, have
been adaptive (i.e., reproduction promot-
ing) in the ancestral environments that se-
lected for their design (their environment
of evolutionary adaptedness), there is no
guarantee that it will be so now (Tooby
& Cosmides 1990b, Symons 1992). Mod-
ern environments differ importantly from
ancestral ones, particularly when it comes
to social behavior. We no longer live in
small, face-to-face societies, in semino-
madic bands of 25–200 men, women, and
children, many of whom were close rel-
atives. Yet our cognitive programs were
designed for that social world.
4. Perhaps most importantly, the brain must
be composed of many different programs,
each specialized for solving a different
adaptive problem our ancestors faced.
Our hunter-gatherer ancestors were,
in effect, on a camping trip that lasted
a lifetime, and they had to solve many
different kinds of problems well to
survive and reproduce under those
conditions: hunting, evaluating plant
resources, cooperating with others,
avoiding predators, dividing resources
among kin, selecting fertile mates, de-
terring sexual rivals, avoiding infectious
diseases, detecting alliances, avoiding
incest, learning grammar, negotiating
dominance hierarchies, and managing
aggression, for example. When natural
selection was reconceptualized as repli-
cator dynamics and combined with game
theory (Williams 1966, Dawkins 1982,
Maynard Smith 1982), it became possible
to derive powerful (and nonintuitive)
inferences about what counts as adaptive
behavior in these domains.
Results from evolutionary game theory
and data about ancestral environments Evolutionary Psychology 203
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
can be used to identify and dissect adap-
tive information-processing problems, to
see what properties programs capable of
solving them would need. This exercise
often reveals that what counts as a so-
lution differs radically and incommen-
surably for different adaptive problems.
Consider, for example, food choice versus
mate choice. The computational struc-
ture of programs that are well engineered
for choosing nutritious foods will fail to
produce adaptive behavior unless they
generate different preferences and trade-
offs than programs designed for choos-
ing fertile sexual partners. Similarly, ma-
chinery that reliably and efficiently learns
which local organisms are predators and
the best way to respond to each (freeze?
run? climb a tree?) lacks properties that
will cause the reliable and efficient acqui-
sition of grammar (and vice versa).
Evolutionary psychologists therefore
expect (and find) that the human mind
contains a large number of information-
processing devices that are functionally
specialized and therefore domain spe-
cific, with different devices activated by
different kinds of content (snakes versus
smiles, food versus mates, cues of social
exchange versus cues of aggression). No
one doubts that the mind contains some
adaptive specializations that execute (rel-
atively) domain-general computations
(e.g., Brase et al. 1998, Rode et al. 1999,
Gallistel & Gibbon 2000, Gigeren-
zer & Selten 2001). But these cannot
produce adaptive behavior unless they
interact with a large number of expert
systems that are domain specialized and
content rich (e.g., Pinker 1997, 2010;
Cosmides & Tooby 2001; Cosmides et al.
2010). True blank slates—architectures
that are content free except for a few
hedonic reinforcers—lack the compu-
tational properties necessary to produce
behavior that tracks fitness (Cosmides
& Tooby 1987, Tooby et al. 2005).
(For comprehensive introductions to the
conceptual foundations of evolutionary
psychology, which include detailed
arguments for each point listed above,
along with controversies and responses,
see Tooby & Cosmides 1992, 2005.)
Knowing that natural selection produces
computational systems that solve adaptive
problems reliably, quickly, and efficiently al-
lows evolutionary psychologists to approach
the study of the mind like an engineer. One
starts with a good specification of an adap-
tive information-processing problem and does
a task analysis of that problem. This allows one
to see what properties a program would have to
have in order to solve that problem well. This
approach generates testable hypotheses about
the structure of the programs that compose
the mind—a point we hope to illustrate in this
From the earliest days of the field, evolu-
tionary psychologists have used sexual selection
theory to explore the psychology of mating rela-
tionships in humans and other animals (Trivers
1972, Symons 1979, Daly & Wilson 1988, Buss
1989). They have already produced a massive
literature on this topic, opening up an area of
study that had been neglected by the psycholog-
ical sciences (for recent reviews, see Buss 2005,
part III, and Roney 2009).
It is less obvious how knowledge and
principles from evolutionary biology can
guide research in more traditional areas of the
cognitive sciences. So we have chosen exam-
ples from visual attention, spatial cognition,
categorization, reasoning, learning, and moti-
vation. In each case, the theoretical framework
provided by evolutionary psychology led to
new questions and surprising results—ones
suggesting the existence of content-specialized
procedures. Through these cases, we hope to
illustrate key features of evolutionary psychol-
ogy: the importance of considering ancestral
environments; how hunter-gatherer studies
and models from behavioral ecology, such as
optimal foraging theory, can lead research in
strange new directions; the role of evolutionary
game theory in generating specific hypotheses
about cognitive design; how adaptationist
204 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
hypotheses are tested against alternatives; the
importance of content-specificity; and the need
to be computational, even when researching
motivation and emotion.
Visual attention is an umbrella term for a suite
of operations that select some portions of a
scene, rather than others, for more extensive
processing. Most research in this area has
explored how attention is deployed in response
to either (a) low-level visual features (color,
intensity, orientation, contrast) or (b) informa-
tion that is personally or task relevant, given a
volitionally chosen goal. This reflects a usually
implicit assumption of the field: that the func-
tion of attention is to enhance the processing
of features necessary for building accurate
knowledge of what exists in the world and the
performance of goals chosen by the individual.
Very few studies have considered the
possibility that there are evolved systems
designed to deploy attention in response to
particular categories of information, in a way
that is independent of volitional goals. Voli-
tional attention is important for a tool-using
species, but focusing too exclusively on a
single task can be very costly. An evolutionary
perspective suggests there should be systems
that incidentally scan the environment for
opportunities and dangers; when there are
sufficient cues that a more pressing adaptive
problem is at hand—an angry antagonist, a
stalking predator, a mating opportunity—this
should trigger an interrupt circuit on volitional
attention and activate programs specialized for
processing information about the new problem
in an adaptive manner. According to this view,
attention is a complex system with interacting
components, some serving object perception,
some deployed volitionally, and some moni-
toring the environment in an ongoing manner
for adaptively important situations. These
monitoring systems are likely to be category
driven because their function is to detect the
presence of situations defined over high-level
objects (e.g., people, animals, antagonists,
cooperators), which can rarely be identified on
the basis of low-level features alone.
Faces regulate social interaction, so initial
attempts to look for category-specific atten-
tional systems started with the human face. In
short order, systems were found that preferen-
tially attend to human faces (Ro et al. 2001)
and that snap attention to the location at which
a pair of eyes is gazing (Friesen & Kingstone
2003). But is this caused by an adaptive special-
ization that evolved for attending to faces—one
with design features that were functionally or-
ganized by natural selection for that purpose?
Because faces are important now, as well as an-
cestrally, a skeptic could argue that preferential
attention to faces is caused by a domain-general
expertise system—an evolved system to be sure,
but one that will cause preferential attention to
any perceptual cue that, if attended, would en-
hance performance on current tasks.
In principle, evidence from developmental
disorders could rule out this expertise hypoth-
esis, just as it did in the debate about face
recognition (Duchaine et al. 2006). Autism, for
example, may selectively disrupt attention to
faces (Chawarska et al. 2010, Remington et al.
2012). But a different way of approaching the
question is to compare attention to faces with
attention to stimulus classes with which people
have much less experience, but that were im-
portant ancestrally. Could there be a category-
specific attentional system analogous to the
appendix: there because it was adaptive in our
evolutionary past, but relatively useless now?
Animal Monitoring: An Appendix
in Visual Attention?
The survival of the typical undergraduate re-
search subject might depend on how vigilantly
she monitors cars and trucks as she drives or
crosses the street, but not on her ability to spot
edible turtles, avoid an ornery warthog, or judge
whether the lion laying sated in the grass has just
started attending to potential prey. The oppo-
site is true of the hunter-gatherers from whom
she is descended. For our foraging ancestors,
nonhuman animals presented either dangers Evolutionary Psychology 205
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
(e.g., predators) or opportunities for hunting
(e.g., prey). Snakes and spiders—ancestral
dangers of little consequence in modern cities
and suburbs—do capture attention in what
appears to be a parallel search process ( ¨
et al. 2001). But what about other animals,
including ones that may not be fear relevant?
And what about attentional monitoring, as
distinct from attentional capture? Because
animals can change their behavior and location
quickly, the lives and livelihoods of our ances-
tors turned on their ability to monitor them
for changes in their state and location. Does
our visual attention system harbor a mech-
anism that monitors animals in an ongoing
It does. New et al. (2007a) addressed this
question using the change detection protocol,
in which subjects are asked to spot the difference
between two rapidly alternating photos that are
almost identical, but include one difference.
This paradigm is famous for eliciting change
blindness—a condition in which observers are
unaware that an element of the scene is chang-
ing (e.g., whole buildings can repeatedly appear
and disappear without the subject noticing). In
this protocol, the only task subjects are given is
to detect changes, so they are free to follow their
own inclinations in attending to different enti-
ties in photos of complex natural scenes. This
allows one to see whether the attentional system
monitors animals more than other objects.
It turns out that change blindness is limited
largely to inanimate objects. Changes to nonhu-
man animals (and to people) are detected faster
and more accurately than changes to plants,
buildings, tools, and even vehicles. For exam-
ple, changes to a small bird at the periphery of a
complex natural scene were detected faster and
more accurately than changes to a large build-
ing at a scene’s center. The time course of re-
sponses revealed that animals not only capture
attention, but they are monitored in an ongoing
manner for changes in their state and location.
A series of control tasks showed that the at-
tentional advantage found for animals was not
due to lower-level visual features, expectation of
motion, task demands, properties of the back-
ground scene, or how interesting the targets
were judged to be. The monitoring system re-
sponsible appears to be category driven, that is,
it is automatically activated by any target the
visual recognition system has categorized as an
What is the origin of this animal-monitoring
system? It could have been built into visual at-
tention because of its benefits over evolutionary
time, regardless of its current utility. Another
possibility is that the visual system does not start
out biased to monitor some categories of infor-
mation over others; but it might be designed to
create category-specific monitoring systems as
an expertise, for any class of stimuli that are fre-
quently encountered and important to monitor.
New et al. (2007a) tested between these phylo-
genetic and ontogenetic accounts by comparing
change detection for nonhuman animals to that
for vehicles and humans.
Vehicles versus animals. Ontogenetically,
monitoring vehicles for sudden changes in
their states and locations is a highly trained skill
of life-and-death importance to car-driving,
street-crossing research subjects, but it was of
no importance phylogenetically. In contrast,
monitoring animals was important phyloge-
netically, but having your attention drawn to
pigeons and squirrels is merely a distraction in
modern cities and suburbs. If the ontogenetic
expertise hypothesis were true, one would
expect people to develop a category-specialized
system for monitoring vehicles, such that
changes to vehicles are detected as well as—or,
indeed, better than—changes to nonhuman
animals. But the reverse was true: Speed and
accuracy at noticing changes were far greater
for nonhuman animals than for vehicles. This
is what one would expect if animal monitoring
arises from an animal-specific evolved system
rather than content-free learning.
People are an interesting stimulus class
because they were important and frequently
encountered phylogenetically and ontogenet-
ically. As such, they represent an upper bound-
ary on the effects that expertise can create,
over and above any evolved bias. How well
206 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
are nonhuman animals monitored, compared to
People versus nonhuman animals. If
preferential attention to a given category is
acquired by domain-general learning processes
alone, then category-driven differences in
expertise would have to arise from differences
in the frequency of experience with a stimulus
class and differences in the current utility of
monitoring it. This would predict that
preferential attention to people should be
much stronger than preferential attention to
nonhuman animals.
From infancy, we are immersed in immense
numbers of important transactions with other
humans, which could have driven the acqui-
sition of human-oriented attentional expertise
without invoking any evolved bias toward hu-
mans. Moreover, the amount of experience
subjects living in American cities and suburbs
have with the human species is greater, by
many orders of magnitude, than their experi-
ence with taxa such as birds, turtles, fish, insects,
or African mammals. Do these vast differences
in exposure rates and current utility translate
into vast differences in the extent to which peo-
ple are monitored compared to nonhuman an-
No. New and his associates found that the
differences between attention to humans and
to other species were marginal at best; indeed,
changes to nonhuman animals were detected
just as well as changes to people in some of
their experiments. Finding similar performance
when exposure rates differ by orders of magni-
tude is inconsistent with any acquisition theory
that invokes domain-general expertise with no
evolved biases.
People are more important and more fre-
quently seen than virtually any other category
of stimuli, yet they recruit little more atten-
tion than nonhuman animals. This suggests that
domain-general expertise systems, if they ex-
ist, do little more than fine tune an evolved
system for monitoring people. At the same time,
the fact that nonhuman animals were moni-
tored about as well as people—and much bet-
ter than vehicles and other objects—implies the
existence of a content-specialized system that
was shaped by ancestral selection pressures, not
general learning processes.
Automatic Regulation of Attention
by High-Level Social Cues
Paleoanthropology and studies of modern
hunter-gatherers show that our ancestors
evolved as a group-living species, in small,
face-to-face bands consisting of 25–200 men,
women, and children. Because most of one’s
interactions will be with ingroup members in
this social ecology, a default setting that allo-
cates more attention to ingroup than outgroup
members would be functional. Not surpris-
ingly, people find it easier to visually distinguish
and recall individuating information about
ingroup members than outgroup members.
This phenomenon—ingroup heterogeneity
paired with outgroup homogeneity—is well-
established in social psychology (Anthony et al.
1992, Ostrom & Sedikides 1992). It is often
mentioned in conjunction with the cross-race
recognition deficit: For some perceivers,
people of a different race “all look alike”—
another case of outgroup homogeneity (for an
interesting discussion, see Levin 2000). Are
there circumstances that modulate attention in
a way that reverses this bias?
It may be safe to ignore outgroup members
in most circumstances, but not when there are
cues that they might intend to harm you. This
line of reasoning led Ackerman et al. (2006) to
propose a system that automatically upregulates
attention to outgroup faces in response to cues
of aggressive intent. To test this hypothesis, the
researchers used race as a proxy for group mem-
bership, and tested recognition memory for in-
group and outgroup faces when their expression
was neutral versus angry. The faces of black and
white men were briefly shown to white subjects,
who were then given an old/new recognition
test. When expressions on the faces were
neutral, subjects were better at recognizing the
faces of white than black men, replicating the
well-known outgroup homogeneity effect. But Evolutionary Psychology 207
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
when the faces were angry, subjects were just as
good—and sometimes better—at recognizing
the faces of black men. Not only did outgroup
homogeneity disappear for angry faces, but
it was reversed when subjects were operating
under processing constraints. When exposure
times were brief and distractor photos were
present, the subjects were better at recognizing
the faces of angry black men than angry white
men, demonstrating outgroup heterogeneity.
This striking result reverses well-established
effects from two large literatures: outgroup
homogeneity and the cross-race recognition
deficit. Yet it is precisely what one would
expect on an evolutionary-functional account.
The anger of an ingroup member is less likely
to erupt into aggression because ingroup mem-
bers participate in a network of cooperative
relationships that afford other opportunities for
resolving disagreements (see below). Because
outgroup members are outside this network,
aggression may be the only bargaining tool
open to them, making their anger more
dangerous. Importantly, angry expressions in-
creased attention to the individuating features
of outgroup members only; ingroup faces were
recognized just as well when they were neutral
as when they were angry.
Activating evolutionarily important goals,
such as self-protection and mating, can modu-
late attention and other cognitive processes in
functional ways (Maner et al. 2005, Becker et al.
2010, Kenrick et al. 2010). But the upregulation
of attention to angry outgroup faces found by
Ackerman et al. was not in response to any
instruction or explicitly represented goal state.
It occurred spontaneously in response to an an-
cestrally relevant threat cue: the species-typical
facial expression associated with anger. More
interestingly, attention was upregulated only
when this species-typical threat expression was
on the face of an outgroup member. Ingroup
members elicited attention regardless of their
emotional state; but when processing limita-
tions forced a trade-off between angry ingroup
and outgroup members, the system preferen-
tially attended to angry outgroup members.
This pattern suggests a system that is func-
tionally specialized for adaptively regulating
attention in response to high-level social cues.
Evolutionary psychology is concerned with
the evolved architecture of the mind. Many
computational mechanisms within this archi-
tecture will be the same in males and females
(i.e., sexually monomorphic), and others will
be different (i.e., sexually dimorphic). Sex
differences in behavior can arise in either
case. When boys and men encounter different
social feedback, environments, and experiences
than girls and women, sexually monomorphic
mechanisms can generate sex differences in
behavior. Sexually dimorphic mechanisms can
also generate sex differences in behavior. The
design of a computational system in women
might differ from the design of the homologous
system in men: different inferences, decision
rules, signal detection thresholds, preferences,
and motivational systems can cause women and
men to make different decisions based on the
same information. But a sexually dimorphic de-
sign could also lead men and women to seek out
and remember different kinds of information,
social feedback, environments, and experi-
ences. For this reason, the mere discovery that
men and women have different experiences is
not sufficient to support the hypothesis that sex
differences in their behavior were generated by
a sexually monomorphic psychology.
Evolutionary psychology provides a frame-
work for predicting the presence and absence of
sex differences in the design of computational
systems. No sex differences are expected in
mechanisms that evolved to solve problems that
were the same for ancestral men and women.
The evolved architecture of mechanisms should
differ between the sexes only when the adaptive
problems faced by ancestral males and females
were systematically different over long periods
of evolutionary time. This principle can slice
domains very finely indeed, as we illustrate us-
ing research on the presence—and absence—of
sex differences in spatial cognition.
208 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Spatial Specializations for Foraging
In cognitive tests tapping spatial cognition
and navigation, men often outperform women
(Voyer et al. 1995). Psychologists have long
assumed that this male advantage is general,
holding across spatial problems and domains.
It is not. Evolutionary psychologists have dis-
covered that women outperform men in certain
spatial tasks. This female advantage was pre-
dicted in advance of any data, based on a care-
ful analysis of how the spatial and navigational
problems associated with foraging for plant
foods differ from those associated with hunting.
There is a sexual division of labor in hunter-
gatherer societies, with men specializing in
hunting and women specializing in gathering
sessile resources, such as plant foods (Marlowe
2007). These tasks place different demands on
spatial cognition. Animals move from place
to place, and they do their best to evade their
predators. Tracking an animal can take a hunter
into unknown territory, requiring a certain
amount of dead reckoning to return to camp in
an energy-efficient way (spatial tasks showing
a male advantage, such as mental rotation, are
thought to tap this navigational skill; Silverman
et al. 2000). Plants, by contrast, stay in one
place. But when a forager encounters a plant,
the berries may need another week to ripen,
the twining vine may be too young to have pro-
duced a mature tuber, and it may be a month be-
fore mature nuts appear on the mongongo tree.
A hunter who is opportunistically harvesting
plants as he tracks animals does not need to re-
member the location of plants that are of no im-
mediate use. But being able to relocate a plant
at a later time, when it has become harvestable,
is important for a forager who specializes in
gathering edible plant foods. And, although
foraging women occasionally hunt small ani-
mals when the opportunity arises, they typically
specialize in gathering plants and other sessile
To return to these sessile resources when
they are harvestable, a forager needs to remem-
ber their location at two scales. The first scale
is within a patch: this requires encoding an
edible plant’s position relative to other plants
and landmarks in a tangled bank of vegetation.
Based on this adaptive problem, Silverman
& Eals (1992) had looked for and found a
female advantage in object-location memory,
which was content-general—as it should be,
given the need to encode the position of
edible plants relative to rocks, trees, and other
objects. The second scale is one that supports
navigation back to the resource patch at a
later time. Navigation at this scale requires
encoding the resource’s absolute location
within a represented environment—a quite
different task, more similar to dead reckoning.
Reasoning that navigational specializations in
women at this larger scale should be triggered
by the presence of plant resources, New et al.
(2007b) conducted their first study at a farmers’
After taking people around and having them
taste and rate foods at different stands, the re-
searchers brought subjects back to a place where
they could not see any of the stands and asked
them to point to where each of the foods had
been—a task that taps both spatial memory
and the kind of vector integration necessary
for efficient navigation. Women outperformed
men—an advantage that held even when the
researchers controlled for a variety of experi-
ential variables, including visits to the farmer’s
market and particular stalls, and how often
each food is eaten and liked (none of which
predicted any variance in pointing accuracy).
This is not because the researchers happened
to find a particularly gifted sample of women:
Men scored higher than the women did on
a general not-plant-related sense-of-direction
test (one that also predicted unique variance
in pointing accuracy across sexes). Consistent
with the idea that pointing accuracy taps naviga-
tion between patches, women succeeded with-
out the relational cues that Silverman & Eals
(1992) identified as important for finding a re-
source within a tangled bank. Follow-up stud-
ies by Krasnow and colleagues (2011) showed
that this female spatial advantage is not caused
by differences in women’s ability to remem-
ber the identity of food resources they have Evolutionary Psychology 209
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
seen; it reflects better spatial memory for the
absolute location of plant foods within a spa-
tial frame. This female spatial memory advan-
tage was highly domain specific: It was elicited
only by fruiting trees. No spatial sex differ-
ences were found for other categories tested,
including buildings, animals, tools, or gender-
stereotyped objects, including jewelry and
An adaptationist approach to sex differences
can slice a domain with remarkable precision,
a point nicely illustrated by a second finding
from the farmer’s market study. Although the
sexual division of labor among hunter-gatherers
suggested the hypothesis that women will have
an advantage in remembering the location of
gatherable resources, other factors can affect
spatial memory as well. Some of these—such as
the nutritional quality of the resource—should
be relevant to both sexes.
Optimal foraging theory was developed by
behavioral ecologists to predict and explain
which species foragers will harvest (Schoener
1971). Not surprisingly, its formal models have
identified a food’s caloric density as one impor-
tant predictor of whether foragers will spend
time searching for it. This led the farmer’s
market researchers to ask what would other-
wise be a very strange question for a traditional
cognitive psychologist: All else equal, are peo-
ple more accurate at pointing to the location of
foods of higher caloric density? Is their spatial/
navigational performance better for, say,
almonds and avocados than for cucumbers and
The answer is yes: Calories count for both
sexes. There was a robust correlation between a
food’s caloric density and accuracy at pointing
to its location. This was not because people pre-
ferred the taste of high-calorie foods: Subjects
had rated how much they liked each food dur-
ing the initial tasting-and-rating phase of the
experiment, but there was no correlation what-
soever between how much they liked each food
and their pointing accuracy. Moreover, the ex-
tent to which calories improved spatial memory
was independent of sex and had a similar effect
size for women and men.
Content matters. A psychologist expecting to
find spatial and navigational processes that op-
erate independently of content would never
think to look for any of these effects. Nor
were they stumbled upon during more than
50 years of intuition-driven research in spatial
That calories count for navigation—that
high-caloric-density foods activate better
spatial memory and processes for returning
to their location—is a result undreamt of in
the philosophy of most cognitive scientists.
So is the discovery that women have a spatial
advantage that is not found for most objects,
but emerges when the task involves gatherable
plant resources. These surprising results imply
that content matters deeply: Embedded within
the computational systems that govern spatial
memory and navigation are elements that
respond differentially to plant resources and to
the caloric density of foods.
Game theory is a tool for analyzing strategic
social behavior—how agents will behave when
they are interacting with others who can antici-
pate and respond to their behavior. Economists
use it to analyze how people will respond to
incentives present in the immediate situation.
Their models typically assume rational actors,
who calculate the payoffs of alternative options
(anticipating that other players will do likewise)
and choose the one most likely to maximize
their short-term profits (but see Hoffman et al.
Evolutionary biologists also adopted game
theory as an analytic tool (Maynard Smith
1982), because the behavior of other people
can be as relentless a selection pressure as
predators and foraging. In contrast to eco-
nomics, evolutionary game theory requires
no assumptions about rationality; indeed, it
can be usefully applied to cooperation among
bacteria or fighting in spiders. It is used to
model interactions among agents endowed
with well-defined decision rules that produce
210 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
situationally contingent behavior. Although
these decision rules are sometimes called strate-
gies by evolutionary biologists, no conscious
deliberation by bacteria (or humans) is implied
(or ruled out) by this term. Whether the
decision rules being analyzed are designed to
regulate foraging, fighting, or cooperating, the
immediate payoffs of these decisions, in food
or resources, are translated into the currency
of offspring produced by the decision-making
agent, and these offspring inherit their parents’
decision rule. In evolutionary game theory, a
decision rule or strategy that garners higher
payoffs leaves more copies of itself in the
next generation than alternatives that garner
lower payoffs. By analyzing the reproductive
consequences of alternative decision rules over
generations, evolutionary biologists can deter-
mine which strategies natural selection is likely
to favor and which are likely to be selected
The evolution of cooperation has been vig-
orously investigated using game theory. We il-
lustrate the method below and then show how
it has led to the discovery of domain-specialized
concepts and reasoning procedures.
The Evolution of Two-Party
Cooperation: Constraints
from Game Theory
The evolution of adaptations for cooperation
is tricky, even when only two individuals are
involved and they can interact repeatedly.
Two-party cooperation, also known as recipro-
cal altruism, reciprocation, or social exchange,
is often modeled as a repeated Prisoner’s
dilemma game. In each round of the game, the
player must decide whether to cooperate or
defect—to provide a benefit of magnitude Bto
the other player (at cost Cto oneself) or refrain
from doing so. In these games, BC>0for
both players. In these environments, strategies
that always cooperate—no matter how their
partners respond—are outcompeted by strate-
gies that always defect, and eventually disappear
from the population (see sidebar Unconditional
Cooperation Is Not Evolutionarily Stable).
Imagine a population of agents participating in a series of
prisoners’ dilemma games. Each agent is equipped with one
of two possible decision rules: “always cooperate” or “always
defect.” “Always cooperate” causes unconditional cooperation:
agents with this design incur cost Cto provide their partner with
benefit B, regardless of how their partner behaves in return.
The other decision rule, “always defect,” accepts benefits from
others but never provides them, so it never suffers cost C. When
two unconditional cooperators interact, their payoff is positive,
because BC>0. When two defectors interact, they get
nothing—they are no better or worse off than if they had not
interacted at all. But every time a cooperator interacts with a
defector, the cooperator suffers a net loss (because it pays cost
Cwith no compensating benefit) and the defector gets B(the
benefit provided by the cooperator) while incurring no cost.
Now imagine that the agents are randomly sorted into pairs
for each new round, there are nrounds during a generation, and
the probability of being paired with a cooperator versus a defec-
tor is pversus (1 p), a function of their relative proportions in
the population. The “always defect” rule never suffers a cost, but
it earns Bevery time it is paired with an agent who always coop-
erates, which is nptimes; thus npBis the total payoff earned by
each defector that generation. In contrast, the “always cooperate”
rule suffers cost Cin every round, for a total cost of nC. It earns
Bonly from the nprounds in which it meets another cooperator,
for a total benefit of npB. Hence, n(pB C) is the total payoff
earned by each cooperator that generation. These payoffs deter-
mine the relative number of offspring each design produces. Be-
cause npB >npB nC, the “always defect” design will leave more
copies of itself in the next generation than the “always cooperate”
design. As this continues over generations, unconditional coop-
erators will eventually disappear from the population, and only
defectors will remain. “Always defect” is an evolutionarily stable
strategy in an environment where the only alternative is a design
that always cooperates. “Always cooperate” is not an evolution-
arily stable strategy—a population of unconditional cooperators
can be invaded and displaced by designs that always defect.
Although unconditional cooperation fails,
agent-based simulations show that decision
rules that cause cooperation can evolve and
be maintained by natural selection if they
implement a strategy for cooperation that is Evolutionary Psychology 211
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
phenotypic mechanism
designed by natural
selection; its features
are functionally
organized to solve
ancestral problems of
survival and
a strategy that
cooperates on round n
of a repeated PD
game, unless its
partner defected on
round n-1
conditional—a strategy that not only recog-
nizes and remembers its history of interaction
with other agents, but also uses that infor-
mation to cooperate with other cooperators
and defect on defectors (Tit-for-Tat is an
example; Axelrod & Hamilton 1981, Axelrod
1984). Conditional cooperators remember acts
of cooperation and cooperate in response, so
they provide benefits to one another, earning
apayoffof(BC) every time they interact.
Because the cooperation of one elicits future
cooperation from the other, they cooperate re-
peatedly, and these positive payoffs accumulate
over rounds. In this, they are like unconditional
cooperators. The difference is that conditional
cooperators limit their losses to defectors. The
first time a conditional cooperator interacts
with a particular defector, it suffers a one-time
loss, C, and the defector earns a one-time
benefit, B. But the next time these two individ-
uals meet, the conditional cooperator defects,
and it does not resume cooperation unless its
partner responds by cooperating. As a result,
designs that defect cannot continue to prosper
at the expense of designs that cooperate
conditionally. Nor can they harvest gains in
trade from interacting with one another. Over
generations, conditional cooperators outre-
produce defectors because they harvest gains
in trade from interacting repeatedly with one
Defectors are often referred to as “cheaters”
in two-party reciprocation or social exchange.
The results of evolutionary game theory tell us
that cognitive adaptations for participating in
social exchange can be favored and maintained
by natural selection, but only if they imple-
ment some form of conditional cooperation. To
do so, they require design features that detect
and respond to cheaters (see Reasoning section,
Collective Action
A similar analysis applies to collective actions:
situations in which three or more individu-
als cooperate to achieve a common goal and
then share the resulting benefits. Defectors in
this form of group cooperation are called “free
There are many situations, such as common
defense, in which the benefits of group cooper-
ation will be shared by everyone in the group,
regardless of how much they contributed to
producing them. When this is true, those who
contribute to the common goal at high lev-
els are at a selective disadvantage compared to
those who contribute at low levels (or not at
all). The benefits of collective action will be
reaped by high and low contributors alike, but
the costs of contribution fall disproportionately
on the high contributors. Consequently, the
low contributors—the free riders—experience
higher net payoffs than the high contributors,
and those who contribute nothing do best of all.
Because net payoffs are translated into offspring
produced, decision rules that cause free riding
will leave more copies of themselves in the next
generation than those that always contribute at
high levels. This will continue over generations.
Eventually the population will consist entirely
of agents who free ride; as a result, no one in
this population will contribute to collective ac-
tions. Indeed, the total population may end up
smaller than it was originally, because it is now
composed entirely of agents who do not benefit
from resources that can be harvested only by co-
operating with others and sharing the resulting
As in two-party cooperation, adaptations for
participating in collective action can be selected
for only if they cause contributors to cooperate
conditionally. In social exchange, a cooperator
can avoid cheaters by switching partners when
alternatives are available. This is more difficult
in collective actions, because withdrawing from
free riders means withdrawing from the group.
A better solution is to keep the group, and either
exclude free riders from it or else punish them
to incentivize higher contributions in the future
(Boyd & Richerson 1992, Hauert et al. 2002,
Panchanathan & Boyd 2004, Tooby et al. 2006,
Tooby & Cosmides 2010).
With these selection pressures in mind, we
turn to concepts, categorization, and reasoning.
212 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Are our brains designed to reliably develop con-
cepts as specific as “cheater” or “free rider,”
which categorize people using criteria that sat-
isfy constraints from evolutionary game theory?
The idea seems eccentric. Indeed, the study of
concepts and categorization started from the as-
sumption that categorization is a unitary and
general process, driven by perceptual similar-
ity and shared features (e.g., Bruner et al. 1956,
Rips et al. 1973). Some categorization processes
do operate widely over many content domains
(for review, see Ashby & Maddox 2005), but
they coexist in the brain with a large num-
ber of content-rich, domain-specific inference
systems (for reviews, see Pinker 1997, Boyer
& Barrett 2005). These include the theory of
mind system (e.g., Baron-Cohen 1995, Leslie
et al. 2004, Onishi & Baillargeon 2005, Saxe
& Powell 2006), intuitive physics (e.g., Spelke
1990, Leslie 1994), and folk biology (Medin &
Atran 1999, Barrett 2005, Mahon & Caramazza
Each of these systems represents the world
using specialized concepts and embodies in-
ferences that, when applied to those concepts,
are well designed for solving a different adap-
tive problem faced by the ancestral hunter-
gatherers from whom we are descended. For
example, the theory of mind system uses cues
such as self-propelled motion and contin-
gent reactivity to distinguish “agents” from
other “objects”; proprietary concepts, such as
“belief,” “desire,” and “intention,” which are
attributed only to entities classified as “agents”;
and specialized reasoning circuits for inferring
these mental states and using them to predict
and explain behavior.
Broad concepts, such as “agents” whose ac-
tions reflect their “beliefs” and “desires,” and
physical “objects” that move only when acted
upon by an outside “force,” are necessary for
interacting with the world. But as systems for
regulating social interactions, they are blunt
Intricate rules of obligation, entitlement,
and moral violation regulate social interac-
tions, and these differ by domain—consider,
for example, the differences between coop-
erating with a team, negotiating rank in a
status hierarchy, courting a romantic partner,
trading favors with a friend, and helping a
sibling. Evolutionary biologists and behavioral
ecologists have developed sophisticated, game-
theoretic models of what counts as adaptive
social behavior in each of these domains. Im-
plementing these behavioral strategies requires
a number of fine-grained social categories and
nuanced moral concepts. As a first example,
we consider evidence for a domain-specialized
concept, “free rider,” designed for regulating
cooperation in collective actions.
Concepts for Collective Action:
Free Riders Versus Cooperators
Humans are almost unique in the extent to
which they participate in collective actions. It is
fundamental to the human propensity to work
in teams and form coalitions, so understanding
the psychological mechanisms that make this
zoologically unusual form of cooperation pos-
sible is fundamental to understanding organiza-
tional behavior, social systems, economics, and
even politics (Olson 1965, Brewer & Kramer
1986, Ostrom 1990, Price et al. 2002).
As discussed above, adaptations for con-
tributing to collective actions cannot evolve un-
less they are accompanied by a desire to exclude
or punish free riders: those with motivational
systems inclining them to avoid the costs of con-
tributing to a collective action while benefiting
from the contributions of others. Economists
argue similarly, that rational actors will with-
draw from collective actions when free riders
are present; indeed, while noting that ratio-
nal choice theory cannot explain it, behavioral
economists and social psychologists have re-
peatedly shown that people are willing to pun-
ish free riders, even when they incur a personal
cost to do so (Yamagishi 1986, Fehr & G¨
2000, Masclet et al. 2003). But what criteria Evolutionary Psychology 213
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
does the mind use to categorize someone as a
free rider?
Economists assume that individuals assess
incentives in the immediate situation and make
decisions that will maximize their short-term
profit. This view suggests that participants in
a collective action will classify anyone who
has contributed less than themselves (or oth-
ers) as a free rider (e.g., Masclet et al. 2003).
Evolutionary game theory asks a different ques-
tion: given the structure of ancestral environ-
ments, which decision rule for categorizing
free riders will best promote its own repro-
duction over generations? Given the ecological
conditions faced by ancestral hunter-gatherers,
a concept that classifies everyone who under-
contributes as a free rider—to be excluded or
punished—is a losing strategy.
Error management. Two categorization
errors threaten the evolutionary stability of
conditional cooperation: mistakenly treating a
free rider as a cooperator (a miss) and mistak-
enly treating a cooperator as a free rider (a false
alarm). In foraging societies, false alarms are
more costly than misses. Based on estimates
of injury rates and variance in foraging success
among existing hunter-gatherers, our ances-
tors experienced frequent reversals of fortune
(Kaplan & Hill 1985, Gurven 2004, Sugiyama
2004). As a result, every individual endowed
with neurocognitive mechanisms that cause
conditional cooperation will sometimes fail to
contribute to a collective action due to errors,
accidents, bad luck, or injury. Categorizing
these conditional cooperators as free riders will
trigger cycles of mutual defection (you defect
on the supposed free rider, who defects on you
in return...). These cycles prevent both parties
from harvesting the benefits of repeated mu-
tual cooperation that occur when conditional
cooperators correctly recognize one another.
Without these benefits, decision rules that
cause cooperation are eventually outcompeted
by those that cause free riding, and collective
action disappears from the population.
This means false alarms were very costly fit-
ness errors for strategies that cooperate con-
ditionally. Misses were less costly: Mistakenly
cooperating with a free rider results in a one-
time loss, because strategies that cooperate con-
ditionally defect on partners who have defected
on them (see above; for agent-based simulations
demonstrating this point in the context of two-
person cooperation, see Delton et al. 2011).
When false alarms are more costly
than misses, natural selection should equip
categorization systems with criteria that mini-
mize them, even if that increases the frequency
of misses (on error management theory, see
Haselton & Buss 2000, Haselton & Nettle
2006). A categorization system that uses level of
contribution as its sole criterion for classifying
someone as a free rider does exactly the wrong
thing: It minimizes misses (the less costly error)
at the expense of generating many false alarms
(the more costly error). An evolutionarily
stable strategy for collective action requires a
“free rider” concept that distinguishes between
undercontributors, sorting them on the basis of
whether they show cues of cooperative versus
exploitive intent.
Exploitive designs. Using an unobtrusive
measure of social categorization based on re-
call errors (a “who did what?” protocol, anal-
ogous to the “who said what?” protocol devel-
oped by Taylor et al. 1978), Delton et al. (2012)
showed that people who try to contribute to
a collective action but fail are not categorized
as free riders—even when they contribute less
than others. To be categorized as a free rider,
the target must undercontribute in a way sug-
gesting exploitive intent—that is, a motivation
to benefit from the collective action without
incurring the costs of contributing to it (e.g.,
by consuming the resource they had promised
to contribute or by making no effort to pro-
cure the promised resource). These individuals
were subsequently judged to be less trustwor-
thy, more selfish, more deserving of punish-
ment, and less desirable as future cooperative
The discovery that targets with exploitive
intent are sorted into a distinct mental
category—and evaluated more negatively than
214 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
targets who try to contribute but fail due to
accident or bad luck—is robust. Moreover, it
cannot be accounted for by a domain-general
process that sifts for any behavioral difference
between targets and uses it to categorize them.
In a control condition using the same cooper-
ative targets that had previously been catego-
rized (when the other targets were free riders),
Delton and colleagues showed that targets who
try to contribute, but fail in one of two distinct
ways, are not sorted into separate categories.
Moral psychology and free riders. Although
these results suggest a domain-specialized
system for categorizing free riders, they could
also be accounted for if the mind has criteria
for distinguishing those who violate moral rules
from those who do not. To test this “moral vio-
lator” counter-hypothesis, Delton et al. (2012)
used the same unobtrusive method to show that
the mind distinguishes free riders from other
kinds of moral violators. As in the other “who
did what?” experiments, subjects saw that ev-
eryone who had agreed to participate in the
collective action contributed resources on three
of five days. But on the other two days, subjects
saw that some targets consumed a resource they
had promised to contribute to the group (free
riders), and others stole a resource owned by the
group. Every one of these targets was intention-
ally violating a moral rule—and illicitly taking
a benefit for themselves that was obligated to
the group. Nevertheless, subjects sharply dis-
tinguished them, as revealed by the categoriza-
tion measure and the response and character
ratings gathered subsequently.
That the mind slices the moral domain so
thinly is remarkable: stealing a resource from
the group and consuming a resource promised
to the group are so similar that most ap-
proaches to moral psychology would not distin-
guish them. These experiments suggest that our
minds really are prepared to notice and remem-
ber which individuals are free riders on collec-
tive actions, making very subtle distinctions be-
tween free riders and people who commit other,
very similar kinds of moral violations.
reasoning about
with the format “If P
then Q
We take for granted that two parties can
make themselves better off than they were be-
fore by exchanging things each values less for
things each values more (help, favors, goods,
services). This form of cooperation for mu-
tual benefit—social exchange—does not exist
in many species, but it is as characteristic of
human life as language and tool use. From evo-
lutionary game theory, we saw that adaptations
for two-party cooperation can be favored and
maintained by selection only if they implement
a strategy that cooperates conditionally.
Conditional cooperation requires cognitive
systems that not only recognize different indi-
viduals, but also remember whether they had
cooperated or defected in the past. Memory
research shows that faces of cooperators and
cheaters are remembered better than faces
of individuals who did neither (Bell et al.
2010), and faces of cooperators activate reward
centers in the brain (Singer et al. 2004). In-
deed, participating in social exchange activates
reward centers more than nonsocial activities
that provide the same payoffs (Elliott et al.
2006). It also triggers a very specialized form of
conditional reasoning (reviewed in Cosmides
& Tooby 2005, 2008a,b).
Conditional Reasoning
and Social Exchange
The study of conditional reasoning was pio-
neered by Peter Wason, who began his in-
quiries with a simple question: Does the brain
contain a reasoning system that implements
first-order logic? (Wason & Johnson-Laird
1972). First-order logic is very useful: It has
rules of inference that generate true conclu-
sions from true premises—a very specialized
function. But its procedures are blank—free of
content—so they can operate uniformly on in-
formation from any domain. Logic’s domain
generality is a good feature if your goal is to
acquire valid knowledge about the world, no
matter what subject you are studying. But this
design feature is a bug for a system designed to
reason adaptively about social exchange. Evolutionary Psychology 215
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Logical violation of a
conditional: cases of
Pand not-Qviolate “If
Pthen Q”
When participating in social exchange,
you agree to deliver a benefit conditionally
(conditional on the other person doing what
you required in return). This contingency can
be expressed as a “social contract”, a condi-
tional rule that fits the following template: “If
you accept benefit Bfrom me, then you must
satisfy my requirement R.” The social contract
is offered because the individual providing the
benefit expects to be better off if its conditions
are satisfied [e.g., if the theater owner receives
the price of a ticket (“requirement R”) in return
for access to the symphony (“benefit B”)]. The
target accepts these terms only if the benefit
provided more than compensates for any losses
he incurs by satisfying the requirement (e.g.,
if hearing the symphony is worth the cost of
the ticket to him). This mutual provisioning
of benefits, each conditional on the others’
compliance, is what is meant by social exchange
or reciprocation (Cosmides 1985, 1989; Tooby
& Cosmides 1996). Understanding it requires
a form of conditional reasoning. But the
inferential rules required do not conform to
the inferential rules of truth-preserving logics
(Cosmides & Tooby 1989, 2008a).
For example, there is no logical inference
by which “If Pthen Q” implies “If Qthen P
(e.g., “If a person is a biologist, then he enjoys
camping,” does not imply “If a person enjoys
camping, then he is a biologist”). But what if P
and Qrefer to benefits and requirements, and
the conditional rule expresses a social exchange
between two parties?
Because conditional cooperation makes
entitlement to benefits contingent on satisfying
obligations, it is natural to infer that “If you
accept benefit Bfrom me, then you must satisfy
my requirement R” also implies “If you satisfy
my requirement Rthen you are entitled to
receive benefit Bfrom me” (e.g., when I say,
“If you borrow my car, then you have to fill
my tank with gas,” I also mean “If you fill my
tank with gas, then you may borrow my car”).
Logic forbids this inference, but reasoning
procedures designed for social exchange
require it. An evolutionarily stable strategy for
reasoning about conditional rules involving
social exchange requires functionally special-
ized inference rules like these, which operate
on abstract yet content-specific conceptual
elements, such as “agent,” “benefit,” “require-
ment,” “obligation,” and “entitlement—what
Cosmides & Tooby (1989, 1992, 2008a) call
social contract algorithms.
To implement decision rules for conditional
cooperation, social contract algorithms also re-
quire an information search function, designed
to look for cheaters. “Cheaters” are individuals
with a disposition to violate social contracts
by taking the benefit offered without satisfy-
ing the requirement on which it was made
Investigations with the Wason
Selection Task
The hypothesis that the brain contains social
contract algorithms, which include a subrou-
tine for detecting cheaters, predicts a dissoci-
ation in reasoning performance by content: a
sharply enhanced ability to reason adaptively
about conditional rules when those rules spec-
ify a social exchange.
Peter Wason’s four-card selection task is
a standard tool for investigating conditional
reasoning (see Supplemental Figure 1; fol-
low the Supplemental Material link from the
Annual Reviews home page at http://www. Subjects are given a con-
ditional rule of the form “If Pthen Q”andasked
to identify possible violations of it—a format
that easily allows one to see how performance
varies as a function of the rule’s content. Wason
developed this task to see if we humans are nat-
ural falsificationists: if we spontaneously apply
first-order logic to look for cases that might vi-
olate a conditional rule. It turns out that people
perform poorly on this task: For most condi-
tional rules, only 5% to 30% of normal sub-
jects respond with the logically correct answer,
even when the rule describes familiar content
drawn from everyday life—such as a disease
causing a particular symptom or a rule describ-
ing people’s preferences or habits (Wason 1983,
Cosmides & Tooby 2008a).
216 Cosmides ·Tooby
Supplemental Material
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Content matters, however. When the
conditional rule involves social exchange
and detecting a violation corresponds to
looking for cheaters, 65% to 80% of subjects
correctly detect violations on the Wason
selection task (see Supplemental Figure 2;
follow the Supplemental Material link from
the Annual Reviews home page at http://www. They succeed even when
the rule specifies a wildly unfamiliar social con-
tract (e.g., “If you get a tattoo on your face, then
I’ll give you cassava root”). The ability to detect
cheaters on social contracts is already present
by age 3–4 (N´
nez & Harris 1998, Harris et al.
2001), and it is found cross-culturally—not just
in industrialized market economies, but also
among Shiwiar hunter-horticulturalists of the
Ecuadorian Amazon (Sugiyama et al. 2002).
This is not because social contracts acti-
vate the inferences of first-order logic. Look-
ing for cheaters requires one to investigate
two classes of individuals: those who have ac-
cepted the benefit offered in the social con-
tract rule (to see if they failed to satisfy the
requirement) and those who have not satisfied
the requirement (to see if they took the ben-
efit anyway). In many Wason selection tasks,
these choices are (by coincidence) logically cor-
rect. But it is simple to create a social contract
problem where investigating the same individ-
uals is logically incorrect (see Supplemental
Figure 3; follow the Supplemental Material
link from the Annual Reviews home page at When this
is done, people do not respect the rules
of logic; they look for cheaters instead
(Cosmides 1989, Gigerenzer & Hug 1992).
The claim that the mind contains reasoning
procedures specialized for detecting cheaters
was (and is) very controversial (Cosmides &
Tooby 2005, 2008a,b; Fodor 2008). A common
response is that people are good at detecting
violations of any conditional rule that is deontic
(i.e., that expresses permission, obligation,
entitlement, or prohibition), whether it is a
social contract or not (e.g., Cheng & Holyoak
1985, Manktelow & Over 1991, Sperber et al.
1995, Fodor 2000). In support of this, they
note—correctly—that people are good at
detecting violations of (deontic) precautionary
rules, such as “if you work with toxic gases, then
you must wear a gas mask” (Cheng & Holyoak
1989, Fiddick et al. 2000). But two separate
systems regulate reasoning about these two
domains: Brain damage can selectively impair
a person’s ability to detect cheaters on social
contracts while leaving intact their ability to
detect violations of precautionary rules (Stone
et al. 2002)—a neural dissociation that is
supported by brain imaging studies (Fiddick
et al. 2005, Ermer et al. 2006, Reis et al. 2007).
Most tellingly, however, good violation
detection is not found for deontic rules that
are neither precautions nor social contracts
(Cosmides & Tooby 2008a,b). Indeed, social
contracts themselves do not elicit violation
detection unless this can reveal potential
cheaters—individuals with a disposition to
illicitly benefit by violating the rule (for ex-
planation, see section Concepts for Collective
Action: Free Riders Versus Cooperators).
Using the same social contract rule—which
was given the same deontic interpretation in
all problems—Cosmides et al. (2010) paramet-
rically varied three cues relevant to detecting
cheaters. Violation detection was high when
potential violators were acting intentionally,
would get the benefit regulated by the rule,
and the situation allowed violations. But
removing any one of these cues independently
(and additively) down-regulated detection of
social contract violations (see Supplemental
Figure 4; follow the Supplemental Material
link from the Annual Reviews home page
at Perfor-
mance was lower (a) when violations reflected
innocent mistakes (rather than intentional
actions), (b) when the violators would not get
the benefit regulated by the rule, and (c)when
the situation made cheating difficult (when
violations are unlikely, the search for them
is unlikely to reveal those with a disposition
to cheat). These results indicate that the
reasoning mechanism involved is not designed
to look for general rule violators, or deontic
rule violators, or violators of social contracts, Evolutionary Psychology 217
Supplemental Material
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
or even cases in which someone has been
cheated; it does not deign to look for people
who violated a social exchange by mistake—not
even when they have accidentally benefited
by doing so. Instead, this reasoning system is
monomaniacally focused on looking for social
contract rule violations when this is likely
to lead to detecting “cheaters”—defined as
agents who obtain a rationed benefit while
intentionally not meeting the requirement.
During the first wave of the cognitive revolu-
tion, researchers focused on the design of sys-
tems that evolved for knowledge acquisition,
not motivation. Some areas, such as vision sci-
ence, made progress despite the use of intuitive
and informal ideas about their adaptive func-
tion. Not all biological functions correspond
so transparently to our intuitions, however.
Notable among these are adaptive problems
relevant to value and motivation.
The adaptive regulation of behavior re-
quires systems that value alternative behavioral
choices in ways that tracked their fitness con-
sequences. But what counts? For example, in
making choices, how much weight should an
organism place on the welfare of a sibling com-
pared to its own? How intimidated should one
individual be by the threat posed by another?
When does the value of forging relationships
with strangers offset the risk of exposure to new
pathogens? Computational systems that make
trade-offs like these should exist, and their study
can benefit from systematic analyses of the bio-
logical problems that selected for their design.
Biologically speaking, computing value
requires more than a general system for
maximizing “utility,” as some economists (and
psychologists) conceive it. What constitutes
biologically successful valuation (i.e., values
enabling choices that promote fitness) differs
from one domain to the next. In many cases, the
criteria for computing value are fundamentally
incommensurable across domains (Cosmides
& Tooby 1987, Tooby et al. 2005). There is
no general set of cross-domain choice criteria
whose uniform application can adaptively guide
food choice, mate choice, group affiliation,
response to mate infidelity, incest avoidance,
predator avoidance, friend-directed altruism,
free rider punishment, cheater avoidance,
infant nursing, sexual arousal, food sharing,
aggression titrating, contagion avoidance,
and so on. When valuation systems require
distinct and incommensurable criteria to solve
motivational problems (e.g., food choice versus
mate choice versus predator avoidance), each
incommensurable domain will require (at least)
one functionally distinct, domain-specialized
Internal Regulatory Variables
By starting with models of specific adaptive
problems drawn from evolutionary biology and
evolutionary anthropology, evolutionary psy-
chologists have identified candidate problems
for which there should be evolved motivational
specializations. When this is done, it becomes
clear that computational strategies capable of
solving these problems require elements that
have no counterpart in traditional models of
motivation. For example, what drive is reduced
by helping family members? By sexual jealousy?
By avoiding incest? By punishing free riders?
By favoring the ingroup? By expressing anger?
From what general goal could they be backward
The kind of programs necessary to solve
motivational problems require computational
elements that are not exactly concepts, beliefs,
representations of goal states, desires, prefer-
ences, or drives, but something else: internal
regulatory variables (along with evolved spe-
cializations that compute them and deliver
them to evolved decision-making systems)
(Lieberman et al. 2007, Ermer et al. 2008,
Tooby et al. 2008, Sell et al. 2009a). Each
regulatory variable evolved to track a narrow,
targeted property of the body, the social en-
vironment, or the physical environment—such
as aggressive formidability, relative status,
218 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
reliability as a cooperator, or kinship—whose
computation provided the necessary inputs
to evolved decision rules. These regulatory
variables store magnitudes (or parameters),
which either express value or provide input to
mechanisms that compute value.
Below we illustrate with research exploring
the architecture of the kin detection system.
Evidence so far indicates that this system pro-
duces kinship indexes—variables that regulate
motivational systems governing altruism and
sexual attraction/aversion.
Genetic Relatedness and Motivation:
Siblings, Incest, and Altruism
One family of adaptive problems involving val-
uation arises from asking how genetic related-
ness should affect social behavior. Prior to the
integration of the evolutionary sciences with
psychology, questions like this were rarely con-
sidered. Yet adaptive problems posed by relat-
edness are nonintuitive, biologically real, and
have large fitness consequences.
Kin-directed altruism. The theory of kin
selection (Hamilton 1964) was a fundamental
advance in the theory of natural selection,
which follows from replicator dynamics. A
gene can cause its own spread not only by
increasing the reproduction of the individual
it is in, but also by increasing the reproduction
of others who are more likely to carry the
same gene than a random member of the
population—that is, close genetic relatives.
This means that natural selection can favor the
evolution of motivational designs that, under
the right envelope of conditions, cause the
individual to sacrifice his or her own welfare
to increase the welfare of a genetic relative.
There is evidence supporting the predictions
of kin selection theory in species in which close
genetic relatives frequently interact, including
amoebas, plants, shrimp, social insects, rodents,
and primates (reviewed in Lieberman et al.
2007). Given that our ancestors lived in small
bands with close genetic relatives, kin selection
theory predicts that human motivational sys-
tems governing welfare trade-offs (including
altruism) should take kinship into account.
Inbreeding avoidance. A second adaptive
problem that arises from proximity to close
genetic relatives is inbreeding depression.
Recessive alleles that are harmless when
matched with a healthy allele can be injurious
when matched with duplicates of themselves.
Because all people carry many unexpressed
deleterious recessives they received from their
parents, zygotes produced when two close rel-
atives mate are far more likely to carry defec-
tive alleles that match than zygotes produced
by individuals who are not related to one an-
other. This leads to a sharp increase in the num-
ber of genetic diseases expressed in children
produced by incestuous matings—costs that
may be further aggravated by parasites differen-
tially exploiting more genetically homogeneous
hosts (Charlesworth & Charlesworth 1999,
Lieberman et al. 2007). This makes incest a
major fitness error, like approaching preda-
tors, eating gravel, or killing your children.
Consequently, computational designs that cost-
effectively reduce inbreeding depression by
avoiding mating with close genetic relatives
outcompete variants in which mating decisions
are unaffected by relatedness. Hence, the hu-
man psychological architecture should contain
evolved systems designed to inhibit incest.
A Kin Detection System
These two adaptive problems—inbreeding
avoidance and kin-directed altruism—both
require a kin detection system: a neurocompu-
tational system that is well engineered (given
the structure of ancestral environments) for
computing which individuals in one’s social
environment are close genetic relatives. By
analyzing the adaptive problem, Lieberman
et al. (2007) derived a model of this architecture
(see Figure 1) and conducted a series of tests
of predictions drawn from the model.
According to this theory, the kin detection
system uses ancestrally reliable cues to com-
pute and update a continuous variable, a kinship Evolutionary Psychology 219
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Figure 1
Architecture of a kin detection system, and its relationship to motivational systems regulating altruism (welfare trade-offs) and sexual
attraction. The system includes several internal regulatory variables. A high kinship index (KIj) between self and individual jshould
lower the sexual value of jto self (SVj), and increase how much weight one places on j’s welfare when making choices (WTRj).
index, KIij, for each familiar other, j. The mag-
nitude of KIij embodies the system’s pairwise
estimate of genetic relatedness between self
(i) and other ( j). Kinship indexes are evolved
regulatory variables that should serve as in-
put to neural programs regulating altruism by
itoward jand, separately, to programs regu-
lating i’s sexual attraction to j. When KIij is
high, it should up-regulate motivations to pro-
vide aid to j, and it should down-regulate sexual
attraction by activating disgust at the possibility
of sex with j.
Ancestrally reliable cues to genetic related-
ness. The adaptive problem of detecting re-
latedness is hard to solve because genetic re-
latedness cannot be directly observed. Instead,
the system must infer it, based on cues that
predict genetic relatedness. A domain-general
learning system cannot, by itself, identify and
then use whatever transient cues best predict
relatedness in the local environment. Discover-
ing which novel cues are best would require the
system to already know the genetic relatedness
of others—the exact problem the kin detection
system needs to solve. Instead, the kin detection
system must contain within its evolved design
a specification of the core cues that it will use
to determine relatedness—cues picked out over
evolutionary time by natural selection because
they reliably tracked genetic relatedness in the
ancestral social world.
For our hunter-gatherer ancestors, a reliable
cue to relatedness is provided by the close asso-
ciation between mother and infant that begins
with birth and is maintained by maternal attach-
ment. Maternal perinatal association (MPA)
provides an effective psychophysical foundation
for the mutual kin detection of mother and
child. It also provides a foundation for sibling
detection. Among our ancestors, when an indi-
vidual observed an infant in an enduring care-
taking association with the observer’s mother,
that infant was likely to be the observer’s sib-
ling. To use this high-quality information, the
kin detection system would need a monitoring
subsystem specialized for registering MPA.
Although MPA allows older offspring to de-
tect their younger siblings, it cannot be used
by younger siblings because they did not exist
when their older siblings were born and nursed.
This implies that the kin detection system’s
psychophysical front end must monitor at least
one additional cue to relatedness. The cumula-
tive duration of coresidence between two chil-
dren, summed over the full period of parental
220 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
care until late adolescence, is a cue that could be
used to predict genetic relatedness—an expan-
sion and modification of the early ethological
proposal about imprinting during early child-
Hunter-gatherer bands fission and fuse over
time, as their members forage and visit other
bands; this means individuals frequently spent
short periods of time with unrelated or distantly
related persons. However, hunter-gatherer
parents (especially mothers) maintained close
association with their dependent children in
order to care for them. Siblings, therefore,
maintained a higher-than-average cumulative
association with each other within the band
structure. As association is summed over
longer periods of time, it monotonically
becomes an increasingly good cue to genetic
relatedness. This invites the hypothesis that
the kin detection system has a system for mon-
itoring duration of coresidence between iand j
during i’s childhood, and that its output allows
younger offspring to detect their older siblings.
Does a kin detection system regulate
sibling altruism and sexual aversion? To
compute the kinship index, the kin detection
system requires (a) monitoring circuitry de-
signed to register cues to relatedness (MPA,
coresidence during childhood, possibly others)
and (b) a computational device, the kinship
estimator, whose procedures have been tuned
by a history of selection to take these registered
inputs and transform them into a kinship
index—the regulatory variable that evolved to
track genetic relatedness.
If these cues are integrated into a single
kinship index—that is, if the kinship index
for each familiar individual is a real compu-
tational element of human psychology—then
two distinct motivational systems should be
regulated by the same pattern of input cues.
For example, when iis younger than j,i’s
kinship index toward jshould be higher the
longer they coresided during i’s childhood.
As a result, i’s levels of altruism and sexual
aversion toward jwill be predicted by their
duration of childhood coresidence.
Lieberman et al. (2007) tested these hy-
potheses about the computational architecture
of human kin detection by quantitatively
matching naturally generated individual
variation in two predicted cues of genetic
relatedness—maternal perinatal associa-
tion and duration of coresidence during
childhood—to individual variation in altruism
directed toward a given sibling and opposition
to incest with that sibling. When the MPA
cue was absent (as it always is for youngers
detecting older siblings), duration of childhood
coresidence with a specific sibling predicted
measures of altruism and sexual aversion
toward that sibling, with similar effect sizes.
When the MPA cue was present (which is
possible only for olders detecting younger
siblings), measures of altruism and sexual
aversion toward the younger sibling were high,
regardless of childhood coresidence.
These results support the model in
Figure 1. At least two cue-monitoring systems
must be present, because motivational out-
comes were regulated by both MPA and cores-
idence duration. Moreover, these two inputs
regulated two different motivational outputs,
altruism and sexual aversion, representing two
entirely independent adaptive problems (titrat-
ing kin-directed altruism and avoiding incest).
The fact that two different motivational systems
are regulated in parallel by the same cues to ge-
netic relatedness implicates a single underlying
computational variable—a kinship index—that
is accessed by both motivational systems. Fi-
nally, the kinship estimator must be a part of the
architecture because, where both cues are avail-
able, the more reliable cue—maternal perinatal
association—trumps coresidence duration. The
two cues interact in a noncompensatory way,
rather than being additive, meaning there is un-
likely to be a direct path from the input variables
(cues) to the motivational systems they regulate.
This entire computational system appears to
operate nonconsciously and independently of
conscious beliefs. When beliefs about genetic
relatedness conflict with the cues this system
uses (as they do when people have coresided
with step-siblings), the motivational outputs Evolutionary Psychology 221
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
(caring, sexual disgust) are shaped by the cues,
not the beliefs. It is worth noting that this sys-
tem influences moral sentiments as well: Those
who have opposite-sex siblings associated with
a high kinship index are more morally opposed
to incestuous relationships, even between third
parties (Lieberman et al. 2003, 2007; Fessler &
Navarete 2004).
Finally, it is important to recognize that the
kin detection system constitutes a learning sys-
tem: It is designed to learn the genetic relat-
edness of familiar others. What makes it espe-
cially interesting is that it does not resemble the
general-purpose learning systems psychologists
are used to positing as explanations for behav-
ior. Instead, it is a proprietary learning system,
with a dedicated function, whose complex ar-
chitecture incorporates content-inflected com-
putational elements (e.g., mother, neonate,
coresidence). The computational problem-
solving strategy evolved to exploit the particu-
lar relationships of the ancestral world (foraging
patterns, mother-infant association) in order to
successfully acquire information it was designed
to learn. It is an open question how much of hu-
man learning is carried out by domain-specific
or content-sensitive devices, and how much by
more general-purpose systems.
The case of kin detection (including
incest avoidance and kin-directed altruism)
is instructive in that it provides an example
of what the computational architectures of
evolved motivational adaptations are likely
to look like. It suggests that the architecture
of human motivation is full of registers for
evolved variables: kinship indexes, sexual
value indexes, coresidence measures, welfare
trade-off ratios, and others. These variables
acquire their properties and meaning by the
evolved behavior-controlling and motivation-
generating procedures that compute and access
them (Tooby et al. 2008). That is, each has a lo-
cation embedded in the input–output relations
of our evolved programs, and their function
inheres in the role they play in the control
architecture of these programs. The kinship
index is located downstream of input cues, it
tracks relatedness, and it is accessed by down-
stream sexual valuation and welfare trade-off
On this view, there is a hidden, previ-
ously unmapped layer of neurocomputational
procedures and representations, consisting of
(a) internal regulatory variables (e.g., a kinship
index), (b) procedures that compute them (e.g.,
the kinship estimator), (c) psychophysical front
ends that monitor cues that serve as inputs to
the procedures that compute these variables
(e.g., maternal perinatal association monitor-
ing), and (d) the entities that these variables
feed, such as decision rules, motivational
intensities, emotion programs, and conscious
feelings (e.g., disgust at the idea of sexual
contact with a sibling). These systems embody
an evolved functional logic that reflects the
adaptive problems they evolved to solve, and
they cause us to experience specific motiva-
tions, value specific outcomes, and express
specific behaviors given certain inputs. Given
the complexity of these systems, one would
never discover them using blind empiricism.
However, models of adaptive problems give
strong guidance in how to construct experi-
mental programs to detect these systems and
to map their computational structure.
In Figure 1, the kinship index feeds systems
that compute at least two other regulatory
variables: (a) the sexual value index (SVi,j), a reg-
ulatory variable whose magnitude represents i’s
assessment of j’s value as a sexual partner, and
(b) a welfare trade-off ratio (WTRi,j), a variable
whose magnitude represents the weight iputs
on j’s welfare when making decisions that
impact them both. Ample literature supports
the view that many ancestrally reliable cues of
a potential sexual partner’s health and fertility
are integrated to form a sexual value index,
whose magnitude is experienced as the person’s
sexual attractiveness (for review, see Sugiyama
2005). But why posit the existence of welfare
trade-off ratios?
222 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Welfare Trade-Offs
In social species, actions undertaken by one
individual sometimes impact the welfare of
others. Although biologists had previously as-
sumed that selection favored choices that were
completely selfish, over the past five decades,
evolutionary biologists have identified a series
of selection pressures that favor taking the wel-
fare of others into account. That is, there are
classes of situations in which the most fitness-
promoting choice is not simply to maximize
one’s own welfare. Instead, under various con-
ditions selection favors placing weight on the
welfare of another as well, so that decisions re-
flect a trade-off between the welfare of self and
other. Adaptive problems that favor calibrating
such trade-offs include kin selection (Hamilton
1964), reciprocation or exchange (Trivers 1971,
Axelrod & Hamilton 1981), aggression and
the asymmetric war of attrition (Hammerstein
& Parker 1982), and externalities (Tooby &
Cosmides 1996).
Most evolutionary models tend to dissect
strategic games involving kinship, aggressive
formidability, reciprocity, and so on as inde-
pendent, isolated problems. However, real or-
ganisms facing real choices cannot. Humans are
often playing a number of games at once with
the same individuals. Each act or choice is an
expression of the weight the actor places on
the target’s welfare, and so a single act cannot
express inconsistent weights at the same time.
Yet different games will rarely converge on the
same weighting function for a specific target: A
target may be a sibling, for example, yet have
cheated recently in a dyadic reciprocation. A
person cannot simultaneously give aid (e.g., if
the other is indexed as close kin) and withhold
aid (e.g., if the other is indexed as someone who
has cheated in their reciprocity relationship).
For the neural architecture to be able to de-
cide which self-favoring or other-favoring acts
to choose at any given moment, the brain needs
a motivational architecture that registers the
factors that, taken individually, might call for
different weightings, and integrates them (at
any given time) into a single welfare trade-off
ratio (Tooby et al. 2008). Multiple converging
lines of evidence support the hypothesis
that WTRs are not just post hoc theoretical
constructs, but exist as real neurocognitive
elements within the human motivational
architecture—elements that play a role in de-
cision making. A number of empirical studies
now support the view that WTRs are person-
specific stored values that display remarkable
consistency across large numbers of benefit
allocations between self and other (Delton
Anger and the Recalibration of
Welfare Trade-Off Ratios
If WTRs actually exist in human brains, then
selection can design adaptations whose func-
tion is to recalibrate their magnitude in oneself
and in others. Gratitude, guilt, and anger may
be social emotions that evolved to recalibrate
WTRs (Tooby & Cosmides 2008). Evidence
supports the view that the gratitude emotion
program is activated when someone puts an un-
expectedly high weight on your welfare. The
gratitude program then recalibrates your WTR
toward the generous person upward—serving
the function of consolidating cooperative rela-
tionships (Lim 2012). Guilt is hypothesized to
be activated when a person discovers they have
placed too low a weight on someone else’s wel-
fare. It also triggers an upward recalibration of
the transgressor’s WTR toward the victim, so
that future treatment is less exploitive (Sznycer
Within this framework, the greatest amount
of research has been done on the recalibrational
theory of anger (Sell 2005, Tooby et al. 2008,
Sell et al. 2009a). In this view, anger is the ex-
pression of a neurocomputational system that
evolved to adaptively regulate behavior in the
context of resolving conflicts of interest in fa-
vor of the angry individual. The anger system
is triggered by actions indicating that the other
party is placing too little weight on the welfare
of the actor (i.e., when their actions express a
WTR that is too low). For example, an action
by jthat imposes a given cost on self is more Evolutionary Psychology 223
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
likely to activate one’s anger program when j
did it to gain a small benefit than a large one,
because this indicates j’s WTR toward oneself
is low (Sell 2005, Lim 2012).
Once triggered, the anger system deploys
bargaining tactics designed to incentivize the
target of the anger to place greater weight on
the welfare of the angry individual in the future.
It does this by activating one of two social
negotiating tactics: threatening to inflict costs
(aggression) or threatening to withdraw ex-
pected benefits (i.e., lowering one’s own WTR
toward the target of the anger). Acts or signals
of anger (such as the anger face) communicate
that, unless the target starts to place more
weight on the angry individual’s welfare, the
angry individual will inflict costs on the target
(in noncooperative relationships) or with-
draw benefits from the target (in cooperative
relationships)—that is, the angry individual will
lower her WTR toward the target unless the
target raises his WTR toward her. Experiments
show that choices revealing a low WTR trigger
anger in cooperative relationships, which in
turn leads the angry individuals to lower their
own WTR toward their partner (the predicted
bargaining response). The magnitude of the
subject’s anger predicts the magnitude of the
subject’s WTR recalibration (Lim 2012).
Because interpersonal bargaining power
arises from the relative ability to inflict costs or
to confer/withhold benefits, the recalibrational
model of anger predicts that individuals with
enhanced abilities to deploy these tactics will
anger more easily, will feel entitled to better
treatment (to a higher WTR from others), and
will prevail more in conflicts of interest. Their
greater ability to inflict costs or withdraw
benefits translates into greater leverage in
bargaining—meaning that anger is more likely
to be successful for them than for others with
less leverage. This suggests that there should
be two regulatory variables that summarize
these dimensions of social power and feed into
the anger system: a formidability index that
encodes the individual’s self-assessment of his
ability to inflict costs (fighting ability) and a
conferral index that encodes the individual’s
self-assessment of his ability to confer/withhold
benefits. For our male ancestors, upper-body
strength was a major component of the ability
to inflict costs on others (by injuring or killing
them). Hence, greater strength should set a
man’s formidability index higher. Even now,
people can accurately assess men’s strength
from sparse visual or vocal cues, and they
spontaneously base their assessment of oth-
ers’ fighting ability on it (Sell et al. 2009b,
2010; Fessler et al. 2012). The ability to
confer/withhold benefits has many sources,
but one factor that is easy to operationalize is
sexual attractiveness in women. As predicted
by the recalibrational theory of anger, Sell
et al. (2009a) found that men with greater
upper-body strength were more prone to
anger, felt more entitled to better treatment,
and prevailed more in conflicts of interest than
men with less upper-body strength. In women,
attractiveness produced these same effects.
Evolutionary psychology is an organizing
framework that can be applied to any topic in
the psychological sciences. Discovering the de-
sign of the mind is easier when evolutionary
biology tells us what we might find.
The authors are not aware of any affiliations, memberships, funding, or financial holdings that
might be perceived as affecting the objectivity of this review.
Ackerman JM, Shapiro JR, Neuberg SL, Kenrick DT, Becker DV, et al. 2006. They all look the same to me
(unless they’re angry): from out-group homogeneity to out-group heterogeneity. Psychol. Sci. 17:836–40
Anthony T, Copper C, Mullen B. 1992. Cross-racial facial identification: a social cognitive integration. Pers.
Soc. Psychol. Bull. 18:296–301
224 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Ashby FG, Maddox WT. 2005. Human category learning. Annu. Rev. Psychol. 56:149–78
Axelrod R. 1984. The Evolution of Cooperation. New York: Basic Books
Axelrod R, Hamilton W. 1981. The evolution of cooperation. Science 211:1390–96
Introduces the field of
psychology; provides
many empirical
Barkow JH, Cosmides L, Tooby J, eds. 1992. The Adapted Mind: Evolutionary Psychology and the
Generation of Culture. New York: Oxford Univ. Press
Baron-Cohen S. 1995. Mind Blindness: An Essay on Autism and Theory of Mind. Cambridge, MA: MIT Press
Barrett HC. 2005. Adaptations to predators and prey. See Buss 2005, pp. 200–23
Becker DV, Anderson US, Neuberg SL, Maner JK, Shapiro JR, et al. 2010. More memory bang for the
attentional buck: Self-protection goals enhance encoding efficiency for potentially threatening males.
Soc. Psychol. Pers. Sci. 1:182–89
Bell R, Buchner A, Musch J. 2010. Enhanced old–new recognition and source memory for faces of cooperators
and defectors in a social-dilemma game. Cognition 117:261–75
Boyd R, Richerson PJ. 1992. Punishment allows the evolution of cooperation (or anything else) in sizable
groups. Ethol. Sociobiol. 13:171–95
Boyer P, Barrett HC. 2005. Domain specificity and intuitive ontology. See Buss 2005, pp. 96–188
Brase GL, Cosmides L, Tooby J. 1998. Individuation, counting, and statistical inference: the role of frequency
and whole-object representations in judgment under uncertainty. J. Exp. Psychol. Gen. 127:3–21
Brewer MB, Kramer RM. 1986. Choice behavior in social dilemmas: effects of social identity, group size, and
decision framing. J. Pers. Soc. Psychol. 50:543–49
Bruner JS, Goodnow JJ, Austin GA. 1956. A Study of Thinking.NewYork:Wiley
Buss DM. 1989. Sex differences in human mate preferences: evolutionary hypotheses tested in 37 cultures.
Behav. Brain Sci. 12:1–14
Buss DM, ed. 2005. The Handbook of Evolutionary Psychology. Hoboken, NJ: Wiley
Charlesworth B, Charlesworth D. 1999. The genetic basis of inbreeding depression. Genet. Res. 74:329–40
Chawarska K, Volkmar F, Klin A. 2010. Limited attentional bias for faces in toddlers with autism spectrum
disorders. Arch. Gen. Psychiatry 67:178–85
Cheng PW, Holyoak KJ. 1985. Pragmatic reasoning schemas. Cogn. Psychol. 17:391–416
Cheng PW, Holyoak KJ. 1989. On the natural selection of reasoning theories. Cognition 33:285–313
Cosmides L. 1985. Deduction or Darwinian algorithms? An explanation of the “elusive” content effect on the Wason
selection task. PhD thesis. Harvard Univ.
Cosmides L. 1989. The logic of social exchange: Has natural selection shaped how humans reason? Studies
with the Wason selection task. Cognition 31:187–276
Cosmides L, Barrett HC, Tooby J. 2010. Adaptive specializations, social exchange, and the evolution of human
intelligence. Proc. Natl. Acad. Sci. USA 107:9007–14
Cosmides L, Tooby J. 1987. From evolution to behavior: evolutionary psychology as the missing link. In The
Latest on the Best: Essays on Evolution and Optimality, ed. J Dupr´
e, pp. 277–306. Cambridge, MA: MIT
Cosmides L, Tooby J. 1989. Evolutionary psychology and the generation of culture, part II. Case study: a
computational theory of social exchange. Ethol. Sociobiol. 10:51–97
Cosmides L, Tooby J. 1992. Cognitive adaptations for social exchange. See Barkow et al. 1992, pp. 163–228
Cosmides L, Tooby J. 2001. Unraveling the enigma of human intelligence: evolutionary psychology and the
multimodular mind. In The Evolution of Intelligence, ed. RJ Sternberg, JC Kaufman, pp. 145–98. Hillsdale,
NJ: Erlbaum
Cosmides L, Tooby J. 2005. Neurocognitive adaptations designed for social exchange. See Buss 2005, pp. 584–
Cosmides L, Tooby J. 2008a. Can a general deontic logic capture the facts of human moral reasoning? How
the mind interprets social exchange rules and detects cheaters. See Sinnott-Armstrong 2008, pp. 53–119
Cosmides L, Tooby J. 2008b. When falsification strikes: a reply to Fodor. See Sinnott-Armstrong 2008,
pp. 143–64
Daly M, Wilson M. 1988. Homicide.NewYork:AldinedeGruyter
Discusses replicator
dynamics; complex
functional design is a
consequence of natural
Dawkins R. 1982. The Extended Phenotype. Oxford, UK: Oxford Univ. Press
Delton AW. 2010. A psychological calculus for welfare tradeoffs. PhD thesis, Univ. Calif., Santa Barbara Evolutionary Psychology 225
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Delton AW, Cosmides L, Guemo M, Robertson TE, Tooby J. 2012. The psychosemantics of free riding:
dissecting the architecture of a moral concept. J. Pers. Soc. Psychol. 102:1252–70
Delton AW, Krasnow MM, Cosmides L, Tooby J. 2011. The evolution of direct reciprocity under uncertainty
can explain human generosity in one-shot encounters. Proc. Natl. Acad. Sci. USA 108:13335–40
Duchaine BC, Yovel G, Butterworth EJ, Nakayama K. 2006. Prosopagnosia as an impairment to
face-specific mechanisms: elimination of the alternative hypotheses in a developmental case. Cogn.
Neuropsychol. 23:714–47
Elliott R, V ¨
ollm B, Drury A, McKie S, Richardson P, Deakin JF. 2006. Co-operation with another player in a
financially rewarded guessing game activates regions implicated in theory of mind. Soc. Neurosci. 1:385–95
Ermer E, Cosmides L, Tooby J. 2008. Relative status regulates risky decision making about resources in men:
evidence for the co-evolution of motivation and cognition. Evol. Hum. Behav. 29:106–18
Ermer E, Guerin SA, Cosmides L, Tooby J, Miller MB. 2006. Theory of mind broad and narrow: Reasoning
about social exchange engages ToM areas, precautionary reasoning does not. Soc. Neurosci. 1:196–219
Fehr E, G¨
achter S. 2000. Cooperation and punishment in public goods experiments. Am. Econ. Rev. 90:980–94
Fessler DMT, Holbrook C, Snyder JK. 2012. Weapons make the man (larger): Formidability is represented
as size and strength in humans. PLoS ONE 7:e32751
Fessler DMT, Navarrete CD. 2004. Third-party attitudes toward sibling incest: evidence for Westermarck’s
hypotheses. Evol. Hum. Behav. 25:277–94
Fiddick L, Cosmides L, Tooby J. 2000. No interpretation without representation: the role of domain-specific
representations and inferences in the Wason selection task. Cognition 77:1–79
Fiddick L, Spampinato MV, Grafman J. 2005. Social contracts and precautions activate different neurological
systems: an fMRI investigation of deontic reasoning. Neuroimage 28:778–86
Fodor J. 2000. Why we are so good at catching cheaters. Cognition 75:29–32
Fodor J. 2008. Comment on Cosmides and Tooby. See Sinnott-Armstrong 2008, pp. 137–42
Friesen CK, Kingstone A. 2003. Abrupt onsets and gaze direction cues trigger independent reflexive attentional
effects. Cognition 87:B1–10
Gallistel CR. 1995. The replacement of general purpose theories with adaptive specializations. In The Cognitive
Neurosciences, ed. MS Gazzaniga, pp. 1255–67. Cambridge, MA: MIT Press
Gallistel CR, Gibbon J. 2000. Time, rate, and conditioning. Psychol. Rev. 107:289–344
Gigerenzer G, Hug K. 1992. Domain-specific reasoning: social contracts, cheating, and perspective change.
Cognition 43:127–71
Gigerenzer G, Selten R, eds. 2001. Bounded Rationality: The Adaptive Toolbox. Cambridge, MA: MIT Press
Gurven M. 2004. To give and to give not: the behavioral ecology of human food transfers. Behav. Brain Sci.
Hamilton WD. 1964. The genetical evolution of social behaviour, I and II. J. Theor. Biol. 7:1–52
Hammerstein P, Parker GA. 1982. The asymmetric war of attrition. J. Theor. Biol. 96:647–82
Harris P, N´
nez M, Brett C. 2001. Let’s swap: early understanding of social exchange by British and Nepali
children. Mem. Cogn. 29:757–64
Haselton MG, Buss DM. 2000. Error management theory: a new perspective on biases in cross-sex mind
reading. J. Pers. Soc. Psychol. 78:81–91
Haselton MG, Nettle D. 2006. The paranoid optimist: an integrative evolutionary model of cognitive biases.
Pers. Soc. Psychol. Rev. 10:47–66
Hauert C, De Monte S, Hofbauer J, Sigmund K. 2002. Replicator dynamics for optional public good games.
J. Theor. Biol. 218:187–94
Herrnstein RJ. 1977. The evolution of behaviorism. Am. Psychol. 32:593–603
Hoffman E, McCabe KA, Smith VL. 1998. Behavioral foundations of reciprocity: experimental economics
and evolutionary psychology. Econ. Inq. 36:335–52
Kaplan H, Hill K. 1985. Food sharing among Ache foragers: tests of explanatory hypotheses.
Curr. Anthropol. 26:223–46
Kenrick DT, Neuberg SL, Griskevicius V, Becker DV, Schaller M. 2010. Goal-driven cognition and functional
behavior: the fundamental motives framework. Curr. Dir. Psychol. Sci. 19:63–67
Krasnow MM, Truxaw D, Gaulin SJC, New J, Ozono H, et al. 2011. Cognitive adaptations for gathering-
related navigation in humans. Evol. Hum. Behav. 32:1–12
226 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Leslie AM. 1994. ToMM, ToBy, and agency: core architecture and domain specificity. In Mapping the Mind:
Domain Specificity in Cognition and Culture, ed. LA Hirschfeld, SA Gelman, pp. 119–48. New York:
Cambridge Univ. Press
Leslie AM, Friedman O, German TP. 2004. Core mechanisms in “theory of mind.” Trends Cogn. Sci. 8:528–33
Levin DT. 2000. Race as a visual feature: using visual search and perceptual discrimination tasks to understand
face categories and the cross-race recognition deficit. J. Exp. Psychol.: Gen. 129:559–74
Lieberman D, Tooby J, Cosmides L. 2003. Does morality have a biological basis? An empirical test of the
factors governing moral sentiments relating to incest. Proc. R. Soc. Lond. B Biol. Sci. 270:819–26
Lieberman D, Tooby J, Cosmides L. 2007. The architecture of human kin detection. Nature 445:727–31
Lim J. 2012. Welfare tradeoff ratios and emotions: psychological foundations of human reciprocity. PhD thesis. Univ.
Calif., Santa Barbara
Mahon BZ, Caramazza A. 2009. Concepts and categories: a cognitive neuropsychological perspective. Annu.
Rev. Psychol. 60:27–51
Maner JK, Kenrick DT, Becker DV, Robertson TE, Hofer B, et al. 2005. Functional projection: how funda-
mental social motives can bias interpersonal perception. J. Pers. Soc. Psychol. 88:63–78
Manktelow KI, Over DE. 1991. Social roles and utilities in reasoning with deontic conditionals. Cognition
Marlowe FW. 2007. Hunting and gathering: the human sexual division of foraging labor. Cross-Cult. Res.
Masclet D, Noussair C, Tucker S, Villeval MC. 2003. Monetary and nonmonetary punishment in the voluntary
contributions mechanism. Am. Econ. Rev. 93:366–80
Uses game theory to
model evolution of
social behavior.
Maynard Smith J. 1982. Evolution and the Theory of Games. Cambridge, UK: Cambridge Univ. Press
Medin D, Atran S, eds. 1999. Folkbiology. Cambridge, MA: MIT Press
New J, Cosmides L, Tooby J. 2007a. Category-specific attention for animals reflects ancestral priorities, not
expertise. Proc. Natl. Acad. Sci. USA 104:16598–603
New J, Krasnow MM, Truxaw D, Gaulin SJC. 2007b. Spatial adaptations for plant foraging: Women excel
and calories count. Proc. R. Soc. Lond. B Biol. Sci. 274:2679–84
nez M, Harris PL. 1998. Psychological and deontic concepts: separate domains or intimate connection?
Mind Lang. 13:153–70
Ohman A, Flykt A, Esteves F. 2001. Emotion drives attention: detecting the snake in the grass. J. Exp. Psychol.:
Gen. 130:466–78
Olson M. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA: Harvard
Univ. Press
Onishi KH, Baillargeon R. 2005. Do 15-month-old infants understand false beliefs? Science 308:255–58
Ostrom E. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge, UK:
Cambridge Univ. Press
Ostrom TM, Sedikides C. 1992. Out-group homogeneity effects in natural and minimal groups. Psychol. Bull.
Panchanathan K, Boyd R. 2004. Indirect reciprocity can stabilize cooperation without the second-order free
rider problem. Nature 432:499–502
Pinker S. 1997. How the Mind Works. New York: Norton
Pinker S. 2010. The cognitive niche: coevolution of intelligence, sociality, and language. Proc. Natl. Acad. Sci.
USA 107:8993–99
Price ME, Cosmides L, Tooby J. 2002. Punitive sentiment as an anti-free rider psychological device. Evol.
Hum. Behav. 23:203–31
Reis DL, Brackett MA, Shamosh NA, Kiehl KA, Salovey P, Gray JR. 2007. Emotional intelligence predicts
individual differences in social exchange reasoning. Neuroimage 35:1385–91
Remington A, Campbell R, Swettenham J. 2012. Attentional status of faces for people with autism spectrum
disorder. Autism 16:59–73
Rips LJ, Shoben EJ, Smith EE. 1973. Semantic distance and the verification of semantic relations.
J. Verbal Learn. Verbal Behav. 12:1–20
Ro T, Russell C, Lavie N. 2001. Changing faces: a detection advantage in the flicker paradigm. Psychol. Sci.
12:94–99 Evolutionary Psychology 227
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Rode C, Cosmides L, Hell W, Tooby J. 1999. When and why do people avoid unknown probabilities in
decisions under uncertainty? Testing some predictions from optimal foraging theory. Cognition 72:269–
Roney JR. 2009. The role of sex hormones in the initiation of human mating relationships. In Endocrinology of
Social Relationships, ed. PT Ellison, PB Gray, pp. 246–69. Cambridge, MA: Harvard Univ. Press
Saxe R, Powell LJ. 2006. It’s the thought that counts. Psychol. Sci. 17:692–99
Schoener TW. 1971. Theory of feeding strategies. Annu. Rev. Ecol. Syst. 2:369–404
Sell A. 2005. Regulating Welfare-Tradeoff Ratios: Three Tests of an Evolutionary-Computational Model of Human
Anger. PhD thesis. Univ. Calif., Santa Barbara
Sell A, Bryant GA, Cosmides L, Tooby J, Sznycer D, et al. 2010. Adaptations in humans for assessing physical
strength from the voice. Proc. R. Soc. Lond. B Biol. Sci. 277:3509–18
Sell A, Cosmides L, Tooby J, Sznycer D, von Rueden C, Gurven M. 2009b. Human adaptations for the
visual assessment of strength and fighting ability from the body and face. Proc. R. Soc. Lond. B Biol. Sci.
Sell A, Tooby J, Cosmides L. 2009a. Formidability and the logic of human anger. Proc. Natl. Acad. Sci. USA
Silverman I, Choi J, Mackewn A, Fisher M, Moro J, Olshansky E. 2000. Evolved mechanisms underlying
wayfinding: further studies on the hunter-gatherer theory of spatial sex differences. Evol. Hum. Behav.
Silverman I, Eals M. 1992. Sex differences in spatial abilities: evolutionary theory and data. See Barkow et al.
1992, pp. 533–49
Singer T, Kiebel SJ, Winston JS, Dolan RJ, Frith CD. 2004. Brain responses to the acquired moral status of
faces. Neuron 41:653–62
Sinnott-Armstrong W, ed. 2008. Moral Psychology. The Evolution of Morality: Adaptations and Innateness.
Cambridge, MA: MIT Press
Spelke ES. 1990. Principles of object perception. Cogn. Sci. 14:29–56
Sperber D, Cara F, Girotto V. 1995. Relevance theory explains the selection task. Cognition 57:31–95
Stone VE, Cosmides L, Tooby J, Kroll N, Knight RT. 2002. Selective impairment of reasoning about social
exchange in a patient with bilateral limbic system damage. Proc. Natl. Acad. Sci. USA 99:11531–36
Sugiyama LS. 2004. Illness, injury, and disability among Shiwiar forager-horticulturists: implications of health-
risk buffering for the evolution of human life history. Am. J. Phys. Anthropol. 123:371–89
Sugiyama LS. 2005. Physical attractiveness in adaptationist perspective. See Buss 2005, pp. 292–343
Sugiyama LS, Tooby J, Cosmides L. 2002. Cross-cultural evidence of cognitive adaptations for social exchange
among the Shiwiar of Ecuadorian Amazonia. Proc. Natl. Acad. Sci. USA 99:11537–42
Symons D. 1979. The Evolution of Human Sexuality. New York: Oxford Univ. Press
Argues that organisms
do not have motives to
“maximize fitness”;
behavior is generated by
mechanisms that
promoted fitness in the
Symons D. 1992. On the use and misuse of Darwinism in the study of human behavior. See Barkow
et al. 1992, pp. 137–59
Sznycer D. 2010. Cognitive adaptations for calibrating welfare tradeoff motivations, with special reference to
the emotion of shame. PhD thesis. Univ. Calif., Santa Barbara
Taylor SE, Fiske ST, Etcoff NL, Ruderman AJ. 1978. Categorical and contextual bases of person memory
and stereotyping. J. Pers. Soc. Psychol. 36:778–93
Discusses why the
genetic basis of complex
adaptations must be
universal (quantitative
traits freer to vary).
Tooby J, Cosmides L. 1990a. On the universality of human nature and the uniqueness of the individual:
the role of genetics and adaptation. J. Pers. 58:17–67
Tooby J, Cosmides L. 1990b. The past explains the present: emotional adaptations and the structure of
ancestral environments. Ethol. Sociobiol. 11:375–424
Describes defects of
blank-slate psychology
and the standard social
science model; argues
for domain-specificity
of many evolved
Tooby J, Cosmides L. 1992. The psychological foundations of culture. See Barkow et al. 1992, pp. 19–
Tooby J, Cosmides L. 1996. Friendship and the banker’s paradox: other pathways to the evolution of adap-
tations for altruism. In Evolution of Social Behaviour Patterns in Primates and Man, ed. WG Runciman, J
Maynard Smith, RIM Dunbar, pp. 119–43. New York: Oxford Univ. Press
Tooby J, Cosmides L. 2005. Conceptual foundations of evolutionary psychology. See Buss 2005, pp. 5–67
228 Cosmides ·Tooby
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64CH08-Cosmides ARI 11 November 2012 8:40
Tooby J, Cosmides L. 2008. The evolutionary psychology of the emotions and their relationship to internal
regulatory variables. In Handbook of Emotions, ed. M Lewis, JM Haviland-Jones, LF Barrett, pp. 114–37.
New York: Guilford. 3rd ed.
Tooby J, Cosmides L. 2010. Groups in mind: the coalitional roots of war and morality. In Human Morality and
Sociality: Evolutionary and Comparative Perspectives, ed. H Høgh-Olesen, pp. 191–234. New York: Palgrave
Tooby J, Cosmides L, Barrett HC. 2005. Resolving the debate on innate ideas: learnability constraints and
the evolved interpenetration of motivational and conceptual functions. In The Innate Mind: Structure and
Content, ed. P Carruthers, S Laurence, S Stitch, pp. 305–37. New York: Oxford Univ. Press
Tooby J, Cosmides L, Price ME. 2006. Cognitive adaptations for n-person exchange: the evolutionary roots
of organizational behavior. Manag. Decis. Econ. 27:103–29
Tooby J, Cosmides L, Sell A, Lieberman D, Sznycer D. 2008. Internal regulatory variables and the design of
human motivation: a computational and evolutionary approach. In Handbook of Approach and Avoidance
Motivation, ed. AJ Elliot, pp. 251–71. Mahwah, NJ: Erlbaum
Trivers RL. 1971. Evolution of reciprocal altruism. Q. Rev. Biol. 46:35–57
Trivers RL. 1972. Parental investment and sexual selection. In Sexual Selection and the Descent of Man, 1871–
1971, ed. B Campbell, pp. 136–79. Chicago: Aldine-Atherton
Voyer D, Voyer S, Bryden MP. 1995. Magnitude of sex differences in spatial abilities: a meta-analysis and
consideration of critical variables. Psychol. Bull. 117:250–70
Wason PC. 1983. Realism and rationality in the selection task. In Thinking and Reasoning: Psychological Ap-
proaches, ed. J St BT Evans, pp. 44–75. London: Routledge & Kegan Paul
Wason PC, Johnson-Laird P. 1972. Psychology of Reasoning: Structure and Content. Cambridge, MA: Harvard
Univ. Press
Discusses gene as unit
of selection; stresses
importance of design
evidence for
recognizing adaptations.
Williams GC. 1966. Adaptation and Natural Selection. Princeton, NJ: Princeton Univ. Press
Yamagishi T. 1986. The provision of a sanctioning system as a public good. J. Pers. Soc. Psychol. 51:110–16 Evolutionary Psychology 229
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64-FrontMatter ARI 15 November 2012 14:20
Annual Review of
Volume 64, 2013 Contents
Shifting Gears: Seeking New Approaches for Mind/Brain Mechanisms
Michael S. Gazzaniga pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp1
Biological Bases of Behavior
The Endocannabinoid System and the Brain
Raphael Mechoulam and Linda A. Parker ppppppppppppppppppppppppppppppppppppppppppppppppppp21
Jamie Ward ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp49
Scene Perception, Event Perception, Object Recognition
Visual Aesthetics and Human Preference
Stephen E. Palmer, Karen B. Schloss, and Jonathan Sammartino ppppppppppppppppppppppppp77
Attention and Performance
Detecting Consciousness: A Unique Role for Neuroimaging
Adrian M. Owen pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp109
Executive Functions
Adele Diamond pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp135
Animal Learning and Behavior
The Neuroscience of Learning: Beyond the Hebbian Synapse
C.R. Gallistel and Louis D. Matzel pppppppppppppppppppppppppppppppppppppppppppppppppppppppp169
Evolutionary Psychology
Evolutionary Psychology: New Perspectives on Cognition
and Motivation
Leda Cosmides and John Tooby ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp201
Origins of Human Cooperation and Morality
Michael Tomasello and Amrisha Vaish pppppppppppppppppppppppppppppppppppppppppppppppppppp231
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64-FrontMatter ARI 15 November 2012 14:20
Language and Communication
Gesture’s Role in Speaking, Learning, and Creating Language
Susan Goldin-Meadow and Martha Wagner Alibali ppppppppppppppppppppppppppppppppppppp257
Nonverbal and Verbal Communication
The Antecedents and Consequences of Human Behavioral Mimicry
Tanya L. Chartrand and Jessica L. Lakin ppppppppppppppppppppppppppppppppppppppppppppppppp285
Intergroup Relations, Stigma, Stereotyping, Prejudice, Discrimination
Sexual Prejudice
Gregory M. Herek and Kevin A. McLemore pppppppppppppppppppppppppppppppppppppppppppppp309
Social Neuroscience
A Cultural Neuroscience Approach to the Biosocial Nature
of the Human Brain
Shihui Han, Georg Northoff, Kai Vogeley, Bruce E. Wexler,
Shinobu Kitayama, and Michael E.W. Varnum pppppppppppppppppppppppppppppppppppppppp335
Organizational Climate/Culture
Organizational Climate and Culture
Benjamin Schneider, Mark G. Ehrhart, and William H. Macey pppppppppppppppppppppppp361
Industrial Psychology/Human Resource Management
Employee Recruitment
James A. Breaugh ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp389
Learning and Performance in Educational Settings
Self-Regulated Learning: Beliefs, Techniques, and Illusions
Robert A. Bjork, John Dunlosky, and Nate Kornell ppppppppppppppppppppppppppppppppppppppp417
Teaching of Subject Matter
Student Learning: What Has Instruction Got to Do With It?
Hee Seung Lee and John R. Anderson pppppppppppppppppppppppppppppppppppppppppppppppppppppp445
Health Psychology
Bringing the Laboratory and Clinic to the Community: Mobile
Technologies for Health Promotion and Disease Prevention
Robert M. Kaplan and Arthur A. Stone ppppppppppppppppppppppppppppppppppppppppppppppppppp471
Research Methodology
Multivariate Statistical Analyses for Neuroimaging Data
Anthony R. McIntosh and Bratislav Miˇsi´cppppppppppppppppppppppppppppppppppppppppppppppppp499
Contents vii
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
PS64-FrontMatter ARI 15 November 2012 14:20
Social Network Analysis: Foundations and Frontiers on Advantage
Ronald S. Burt, Martin Kilduff, and Stefano Tasselli ppppppppppppppppppppppppppppppppppppp527
Cumulative Index of Contributing Authors, Volumes 54–64 ppppppppppppppppppppppppppp549
Cumulative Index of Chapter Titles, Volumes 54–64 ppppppppppppppppppppppppppppppppppp554
An online log of corrections to Annual Review of Psychology articles may be found at
viii Contents
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
Connect With Our Experts
Connect With Our Experts
Tel: 800.523.8635 (us/can) | Tel: 650.493.4400 | Fax: 650.424.0910 | Email:
New From Annual Reviews:
Annual Review of Vision Science • Volume 1 • November 2015
Co-Editors: J. Anthony Movshon, New York University and Brian A. Wandell, Stanford University
The Annual Review of Vision Science reviews progress in the visual sciences, a cross-cutting set of disciplines that intersect
psychology, neuroscience, computer science, cell biology and genetics, and clinical medicine. The journal covers a broad
range of topics and techniques, including optics, retina, central visual processing, visual perception, eye movements, visual
development, vision models, computer vision, and the mechanisms of visual disease, dysfunction, and sight restoration. The
study of vision is central to progress in many areas of science, and this new journal will explore and expose the connections that
link it to biology, behavior, computation, engineering, and medicine.
Adaptive Optics Ophthalmoscopy, Austin Roorda,
Jacque L. Duncan
Angiogenesis in Eye Disease, Yoshihiko Usui,
Peter D. Westenskow, Salome Murinello, Michael I. Dorrell,
Leah Scheppke, Felicitas Bucher, Susumu Sakimoto,
Liliana P Paris, Edith Aguilar, Martin Friedlander
Color and the Cone Mosaic, David H. Brainard
Control and Functions of Fixational Eye Movements,
Michele Rucci, Martina Poletti
Deep Neural Networks A New Framework for Modeling
Biological Vision and Brain Information Processing,
Nikolaus Kriegeskorte
Development of Three-Dimensional Perception in Human
Infants, Anthony M. Norcia, Holly E. Gerhard
Functional Circuitry of the Retina, Jonathan B. Demb,
Joshua H. Singer
Image Formation in the Living Human Eye, Pablo Artal
Imaging Glaucoma, Donald C. Hood
Mitochondria and Optic Neuropathy, Janey L. Wiggs
Neuronal Mechanisms of Visual Attention, John Maunsell
Optogenetic Approaches to Restoring Vision, Zhuo-Hua
Pan, Qi Lu, Anding Bi, Alexander M. Dizhoor, Gary W. Abrams
Organization of the Central Visual Pathways Following Field
Defects Arising from Congenital, Inherited, and Acquired
Eye Disease, Antony B. Morland
Contributions of Retinal Ganglion Cells to Subcortical
Visual Processing and Behaviors, Onkar S. Dhande,
Benjamin K. Staord, Jung-Hwan A. Lim,
Andrew D. Huberman
Ribbon Synapses and Visual Processing in the Retina,
Leon Lagnado, Frank Schmitz
The Determination of Rod and Cone Photoreceptor Fate,
Constance L. Cepko
A Revised Neural Framework for Face Processing,
Brad Duchaine, Galit Yovel
Visual Adaptation, Michael A. Webster
Visual Functions of the Thalamus, W. Martin Usrey,
Henry J. Alitto
Visual Guidance of Smooth Pursuit Eye Movements,
Stephen Lisberger
Visuomotor Functions in the Frontal Lobe, Jerey D. Schall
What Does Genetics Tell Us About Age-Related
Macular Degeneration? Felix Grassmann, Thomas Ach,
Caroline Brandl, Iris M. Heid, Bernhard H.F. Weber
Zebrash Models of Retinal Disease, Brian A. Link,
Ross F. Collery
Access all Annual Reviews journals via your institution at
FREE online access to Volume 1 will be available until November 2016.
Annu. Rev. Psychol. 2013.64:201-229. Downloaded from
Access provided by on 08/21/16. For personal use only.
... This prescriptive and gene-centered neo-Darwinian view has been challenged by recent advances in biology and genetics (Müller 2007;Danchin et al. 2019). Meanwhile, in cognitive sciences, adaptationism remains the predominant explanation for cognitive and cultural processes on the basis of their possible optimization of adaptative value (Durham 1990;Cosmides and Tooby 2013;Lacquaniti et al. 2013). In other words, according to the prescriptive view, cognition and culture are considered to be largely the result of a selection process that guarantees optimization of adaptive value. ...
... This view is prescriptive as only the genes that determine traits optimizing the odds of survival and reproduction by their associated functions are selected. This prescriptive view has naturally led to consider that each organ in the body evolved to serve a specific function and is retained or discarded by natural selection depending on whether they optimally solve problems that affect global fitness/reproduction (Gregory 2008;Cosmides and Tooby 2013). Fitness should be understood here in its broadest sense, as the ability of organisms to survive and reproduce in the environment in which they find themselves (Orr 2009). ...
Full-text available
Since Darwin’s theory of evolution, adaptationism is frequently invoked to explain cognition and cultural processes. Adaptationism can be described as a prescriptive view, as phenotypes that do not optimize fitness should not be selected by natural selection. From an epistemological perspective, the principle of a prescriptive definition of adaptation seems incompatible with recent advances in epigenetics, evolutionary developmental biology, ethology, and genomics. From these challenges, a proscriptive view of adaptation has emerged, postulating that phenotypes that are not deleterious will be evolutionary maintained. In this epistemological investigation, we examine how the shift from adaptationism to a proscriptive view changes our view of cognition and culture. We argue that, while adaptationism leads to cognitivism and a view of culture as strategies to optimize overall fitness, the proscriptive definition favors embodied theories of cognition and a view of culture as the cumulative diffusion of behaviors allowed by the constraints of reproduction.
... Many evolutionary psychologists assert that much of human cognitive behavior can be explained by appealing to a set of adaptive psychological mechanisms, which have evolved to solve recurrent problems or challenges in human ancestral environments and are therefore the functional products of natural selection (Barkow, Cosmides, and Tooby 1992;Buller 2005). 4 Evolutionary psychologists thus typically maintain that many human social behaviors and even individual traits/abilities occur universally across all cultures (Cosmides and Tooby 2013). Examples of such traits/abilities include abilities to infer others' emotions (Al-Shawaf et al. 2016), identify kinship (Jones 2003), and cooperate with others (Johnson and Bering 2006, for a helpful review). ...
Full-text available
We start by introducing the idea of echo chambers. Echo chambers are social and epistemic structures in which opinions, leanings, or beliefs about certain topics are amplified and reinforced due to repeated interactions within a closed system; that is, within a system that has a rather homogeneous sample of sources or people, which all share the same attitudes towards the topics in question. Echo chambers are a particularly dangerous phenomena because they prevent the critical assessment of sources and contents, thus leading the people living within them to deliberately ignore or exclude opposing views. In the second part of this paper, we argue that the reason for the appearance of echo chambers lies in the adoption of what we call 'epistemic vices'. We examine which vices might be responsible for their emergence, and in doing so, we focus on a specific one; 'epistemic violence'. In assessing and evaluating the role of this epistemic vice, we note that it can be triggered by epistemic contexts characterized by high stakes that may turn ordinary intellectual virtues (such as skepticism) into vices (such as denialism). In the third part of this contribution, we suggest a way to deal with echo chambers. The solution focuses on advocating a responsibilist pedagogy of virtues and vices that-we claim-might be capable of preventing their emergence.
... We argue that effort itself is perceived as a costly signal of moral character, even when one's efforts are devoted to tasks that do not have direct moral or economic consequences. Someone who invests effort in one task may be seen as a preferable partner for future cooperative tasks-as one who will not take a "free ride" on the backs of others' collective efforts (Cosmides & Tooby, 2013). Thinking about effort moralization as a "deeply rational" heuristic process, one that facilitates social decision-making irrespective of its economic irrationality (Kenrick et al., 2009), may provide stronger theoretical foundations for understanding why displays of effort increase perceptions of one's moral character-even in situations where one's efforts are not materially beneficial or necessary. ...
Full-text available
People believe that effort is valuable, but what kind of value does it confer? We find that displays of effort signal moral character. Eight studies (N = 5,502) demonstrate the nature of these effects in the domains of paid employment, personal fitness, and charitable fundraising. The exertion of effort is deemed morally admirable (Studies 1-6) and is monetarily rewarded (Studies 2-6), even in situations where effort does not directly generate additional product, quality, or economic value. Convergent patterns of results emerged in South Korean and French cross-cultural replications (Studies 2b-c). We contend that the seeming irrationality of valuing effort for its own sake, such as in situations where one's efforts do not directly increase economic output (Studies 3-6), reveals a "deeply rational" social heuristic for evaluating potential cooperation partners. Specifically, effort cues engender broad moral trait ascriptions, and this moralization of effort influences donation behaviors (Study 5) and cooperative partner choice decision-making (Studies 4 and 6). In situating our account of effort moralization into past research and theorizing, we also consider the implications of these effects for social welfare policy and the future of work.
This article examines Skinner’s often neglected ideas about evolution, which he returns to in his final academic paper. I attempt to square Skinner’s advocacy for evolutionary explanation, including his efforts to reconcile biological, individual, and cultural adaptation, with how he is framed and critiqued by a school of evolutionary psychologists who attribute to Skinner a blank slate, or so-called standard social science model, view of the mind. I argue that characterizing Skinner in this manner is inconsistent with his evolutionary writings and ignores Skinner’s explicit disavowals of such interpretations. I then discuss Skinner’s evolutionary views in light of contemporary evolutionary theories of human psychology. I also compare the reception to evolutionary psychology and Skinner within the field more generally and conclude by discussing the proposal that evolutionary psychology should be considered a new paradigm for psychology, a claim that seems to follow from evolutionary psychologists’ caricature of Skinner.
Full-text available
Many researchers assume that until 10-12,000 years ago, humans lived in small, mobile, relatively egalitarian bands. This "nomadic-egalitarian model" suffuses the social sciences. It informs evolutionary explanations of behavior and our understanding of how contemporary societies differ from those of our evolutionary past. Here, we synthesize research challenging this model and articulate an alternative, the diverse Pleistocene model, to replace it. We review the limitations of using recent foragers as models of Late Pleistocene societies and the considerable social variation among foragers commonly considered small-scale, mobile, and egalitarian. We review ethnographic and archaeological findings covering 34 world regions showing that non-agricultural peoples often live in groups that are more sedentary, unequal, large, politically stratified, and capable of large-scale cooperation and resource management than is normally assumed. These characteristics are not restricted to extant Holocene hunter-gatherers but, as suggested by archaeological findings from 27 Middle Stone Age sites, likely characterized societies throughout the Late Pleistocene (until c. 130 ka), if not earlier. These findings have implications for how we understand human psychological adaptations and the broad trajectory of human history.
Successful gene transmission requires sex, but sex requires staying alive long enough to have some of it. For humans this requires, at a minimum, growing and surviving through more than a decade of sexual immaturity. In other words, the ‘mating machine’ depends on a functioning ‘survival machine’. And more survival for our ancestors (as in all animals) generally meant more sex. Evolution, therefore (as in all animals), has given us a survival instinct. In humans this involves motivations and behaviours for garnering a wide range of particular resources (e.g. physiological necessities, shelter, social life, and alliances), and for avoiding a wide range of health and mortality risks (injury and disease). In this chapter, we explore how many of our likes and dislikes today, our joys and our fears (and cultures that represent them), were shaped by the evolution of a ‘Survival Drive’ in Pleistocene Africa (Curtis 2013; Orians 2014). As Kenrick and Griskevicius (2013) put it, ‘The humans who became our ancestors were those who protected themselves from enemies and predators, avoided infection and disease, got along with other people in their tribe, and gained the respect of their fellow tribe members’.
Multiple lines of evidence suggest that there are two major dimensions of social perception, often called warmth and competence, and that warmth is prioritized over competence in multiple types of social decision-making. Existing explanations for this prioritization argue that warmth is more consequential for an observer's welfare than is competence. We present a new explanation for the prioritization of warmth based on humans' evolutionary history of cooperative partner choice. We argue that the prioritization of warmth evolved because ancestral humans faced greater variance in the warmth of potential cooperative partners than in their competence but greater variance in competence over time within cooperative relationships. These each made warmth more predictive than competence of the future benefits of a relationship, but because of differences in the distributions of these traits, not because of differences in their intrinsic consequentiality. A broad, synthetic review of the anthropological literature suggests that these conditions were characteristic of the ecologies in which human social cognition evolved, and agent-based models demonstrate the plausibility of these selection pressures. We conclude with future directions for the study of preferences and the further integration of social and evolutionary psychology.
The survival processing effect is enhanced memory for the items processed in relation to a survival scenario. According to the adaptive memory framework, a similar mnemonic advantage can be expected to emerge when items are processed in a reproduction context. The present study investigated the effects of survival processing and infidelity-based reproductive processing on recall memory, using scenarios that were matched in terms of the number of problems and the perceived threat. In four within-subject conditions (survival, sexual infidelity, emotional infidelity and pleasantness), 134 participants rated words on how helpful they are to survive, to detect infidelity, and how pleasant they are. The results showed a survival processing effect. None of the infidelity conditions led to enhanced recall memory, regardless of the participants’ sex. Taken together with the previous studies, the current findings suggest that there is no infidelity-based reproductive processing effect that is comparable to the survival processing advantage.
Scientists have long employed puppets in research with young children; this essay explores the validity of this practice. After considering what puppets are, their main types and history, I note the different ways puppets have been employed in research. One of these uses raises the issue of whether and when children apply their theory of mind to puppets. After exploring this issue, I consider if children believe puppets actually are animate and sentient, like humans, and whether children participating in experiments with puppets are pretending (in the sense of pretend play) that puppet stimuli are human. Children aged three years and older and infants are discussed separately, as different definitions of puppets have been used in the research across these age levels and different patterns of results have been obtained. I end with ideas regarding further research.
Full-text available
How and why do our minds generate different levels of attraction to others? This chapter integrates life history, evolutionary psychology, and human biology approaches to address this question. Biological adaptations regulate a vast number of life history trade-offs that affect how we look, smell, sound, and behave. Selection produced adaptations that evaluate these cues and regulate our degree of attraction to others based on their relative probable social value to us in different contexts. This chapter outlinesthe alternative evolutionary explanations for the emergence of an attraction,basic components necessary for attraction systems to evolve, and sources of variation in attractiveness assessment. It identifies different domains of social value for which attractiveness assessment evolved, reviews evidence for some of the hypothesized attractiveness-assessment adaptations in those domains, and highlights avenues calling for increased attention. Finally, it calls for greater integration of evolutionary psychology, human biological research, and data from small-scale foraging societies to generate predictions about these domains of social value, the cues or signals associated with them, adaptations selected to regulate attraction to them, and the life history trade-offs involved in these processes. New research on body shape attractiveness is presented to illustrate these points.
Full-text available
One of the most familiar empirical phenomena associated with face recognition is the cross-race (CR) recognition deficit whereby people have difficulty recognizing members of a race different from their own. Most researchers assume that the CR deficit is caused by failure to generalize perceptual encoding expertise from same-race (SR) faces to CR faces. However, this explanation ignores critical differences in the social cognitions and feature coding priorities associated with SR and CR faces. On the basis of data from visual search and perceptual discrimination tasks, it appears that the deficit occurs because people emphasize visual information specifying race at the expense of individuating information when recognizing CR faces. In particular, it is possible to observe a paradoxical improvement in both detection and perceptual discrimination accuracy for CR faces that is limited to those who recognize them poorly. These findings support a new explanation for the CR recognition deficit based on feature coding differences between CR and SR faces, and appear incompatible with similarity-based models of face categories.
Full-text available
While substantial evidence supports the existence of Westermarck's [Westermarck, E. (1891). The history of human marriage. London: Macmillan & Co.] hypothesized inbreeding avoidance mechanism, questions remain. We examined the Westermarck hypothesis using third-party reactions to a fictional case of sibling incest, a method paralleling that of Lieberman et al. [Proc. R. Soc. Lond., B Biol. Sci. 273 (2003) 819]. Controlling for cultural attitudes, a history of cosocialization with an opposite-sex individual was associated with increased disgust at, and decreased tolerance of, others' incestuous behavior. Consistent with parental investment theory, this effect was stronger in females than in males. In males, each additional cosocialized sibling increased the strength of the response; in females, there was a nonsignificant trend in the same direction. Indirect measures failed to reveal a time-limited sensitive period during which cosocialization has maximal effect. These results both bolster the evidence in favor of Westermarck's inbreeding avoidance hypothesis and support his neglected explanation of the origin of incest taboos.
Participants searched for discrepant fear-relevant pictures (snakes or spiders) in grid-pattern arrays of fear-irrelevant pictures belonging to the same category (flowers or mushrooms) and vice versa. Fear-relevant pictures were found more quickly than fear-irrelevant ones. Fear-relevant, but not fear-irrelevant, search was unaffected by the location of the target in the display and by the number of distractors, which suggests parallel search for fear-relevant targets and serial search for fear-irrelevant targets. Participants specifically fearful of snakes but not spiders (or vice versa) showed facilitated search for the feared objects but did not differ from controls in search for nonfeared fear-relevant or fear-irrelevant, targets. Thus, evolutionary relevant threatening stimuli were effective in capturing attention, and this effect was further facilitated if the stimulus was emotionally provocative.