Emerging Trends in The Social and Behavioral Sciences

Genetics and Social Behavior

Item

Title

Genetics and Social Behavior

Author

Harpending, Henry

Cochran, Gregory

Research Area

Special Areas of Interdisciplinary Study

Topic

Genetics, the Individual and Society

Abstract

We focus on the effects of gene differences on social and behavioral differences among individuals and among larger groups of individuals. Many specific genetic markers are known that influence aspects of personality and behavior. The focus on single genes and groups of genes is giving way to quantitative genetics, the statistical study of transmission of characteristics viewed as the outcome of the effects of very large numbers of genes. While traditional social science largely ignores the effects of genetically transmitted influences, the subject persists and grows in importance. Classical quantitative genetic methods may give much insight into human behavioral diversity and they provide the “right” way to measure and assess variation in rates of threshold traits. We discuss examples, trends, and possibilities for the incorporation of genetic data and models in the social and behavioral sciences without advocating major changes in practice.

Related Essays

Telomeres (Psychology), Nancy Adler and Aoife O'Donovan

Social Epigenetics: Incorporating Epigenetic Effects as Social Cause and Consequence (Sociology), Douglas L. Anderton and Kathleen F. Arcaro

Kin‐Directed Behavior in Primates (Anthropology), Carol M. Berman

The Sexual Division of Labor (Anthropology), Rebecca Bliege Bird and Brian F. Codding

Genetics and the Life Course (Sociology), Evan Charney

Genetic and Environmental Approaches to Political Science (Political Science), Zoltán Fazekas and Peter K. Hatemi

Food Sharing (Anthropology), Michael Gurven and Adrian V. Jaeggi

An Evolutionary Perspective on Developmental Plasticity (Psychology), Sarah Hartman and Jay Belsky

Grandmothers and the Evolution of Human Sociality (Anthropology), Kristen Hawkes and James Coxworth

The Neurobiology and Physiology of Emotions: A Developmental Perspective (Psychology), Sarah S. Kahle and Paul D. Hastings

Herd Behavior (Psychology), Tatsuya Kameda and Reid Hastie

Genetic Foundations of Attitude Formation (Political Science), Christian Kandler et al.

Cultural Neuroscience: Connecting Culture, Brain, and Genes (Psychology), Shinobu Kitayama and Sarah Huff

Reconciliation and Peace‐Making: Insights from Studies on Nonhuman Animals (Anthropology), Sonja E. Koski

Cooperative Breeding and Human Evolution (Anthropology), Karen L. Kramer

Niche Construction: Implications for Human Sciences (Anthropology), Kevin N. Laland and Michael O'Brien

Culture, Diffusion, and Networks in Social Animals (Anthropology), Janet Mann and Lisa Singh

Evolutionary Perspectives on Animal and Human Personality (Anthropology), Joseph H. Manson and Lynn A. Fairbanks

Neural and Cognitive Plasticity (Psychology), Eduardo Mercado III

Evolutionary Theory and Political Behavior (Political Science), Michael Bang Petersen and Lene Aarøe

Gestural Communication in Nonhuman Species (Anthropology), Simone Pika

Darwinism as a Decryption Key for the Human Mind (Psychology), Csaba Pléh and Ottilia Boross

DNA Revolution and the Social and Behavioral Sciences (Psychology), Maciej Trzaskowski and Robert Plomin

Born This Way: Thinking Sociologically about Essentialism (Sociology), Kristen Schilt

Vocal Communication in Primates (Anthropology), Katie E. Slocombe

Primate Allomaternal Care (Anthropology), Stacey Tecot and Andrea Baden

How Form Constrains Function in the Human Brain (Psychology), Timothy D. Verstynen

A Gene‐Environment Approach to Understanding Youth Antisocial Behavior (Psychology), Rebecca Waller et al.

Identifier

etrds0145

extracted text

Genetics and Social Behavior
HENRY HARPENDING and GREGORY COCHRAN

Abstract
We focus on the effects of gene differences on social and behavioral differences
among individuals and among larger groups of individuals. Many specific genetic
markers are known that influence aspects of personality and behavior. The focus
on single genes and groups of genes is giving way to quantitative genetics, the
statistical study of transmission of characteristics viewed as the outcome of the
effects of very large numbers of genes. While traditional social science largely
ignores the effects of genetically transmitted influences, the subject persists and
grows in importance. Classical quantitative genetic methods may give much insight
into human behavioral diversity and they provide the “right” way to measure
and assess variation in rates of threshold traits. We discuss examples, trends, and
possibilities for the incorporation of genetic data and models in the social and
behavioral sciences without advocating major changes in practice.

INTRODUCTION
There are two distinct ways we could approach the task of reviewing genetics
and human social behavior. One is to review aspects that are genetic in the
sense of hard-wired in the human genome and universal. For example, we
all learned spoken language early in our lives and most agree that the trait
“learn to speak the language(s) in your social environment” is such a universal, while the specific language(s) we learned were entirely dependent on our
rearing environment. In spite of the existence of people who never learned to
speak, we do not hesitate to regard learning speech as “genetic.” However,
things are not so simple: speech does not have to be taught while literacy
does. We are all literate, but many of us struggled to learn, say, calculus, in a
way we never struggled with literacy. This would lead to a consideration of
the nature of cognitive modules, a lively area of current research.
Another way to proceed is to consider how genetic differences among people or among groups of people lead to or have been generated by individual
or group differences in social behavior. Are individual differences in social
behavior determined by gene differences? Do group level gene differences
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.

1

2

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

modulate group differences in social behavior? A century ago, the response
of most educated people would have been “of course” to questions such as
these, whereas today, the response of many educated people and scientists is
“of course not.” We suggest that this change has gone too far, motivated in
many cases by political undercurrents, and that an important task for social
science it to understand and investigate gene differences.
In the past century, many anthropologists were interested in temperament,
for example, and culture and personality was a prominent subfield. It is
striking that in all that literature, the possibility of such things reflecting gene
differences was never on the table. Roughly between 1900 and 1950, there
had emerged a new cluster of academic disciplines, social science. The vision
was that this would be a third kind of knowledge, somewhere between the
humanities and the sciences, a branch that would stand between and perhaps unite human knowledge. The optimism and ambition of practitioners
of the new disciplines were great and they were contagious so, for example,
Yale University created the Institute of Human Relations in the 1930,
“directly concerned with the problems of man’s individual and group conduct” (Morawski, 1986). Harvard University created what was practically a
whole new division of the university with its Department of Social Relations
in 1946, a hybrid of sociology, anthropology, and psychology.
A simple, indeed simplistic, statement of the origin of social science in its
early years was that it insisted on the separation of the body and soul, the soul
being the locus of social behavior and personality. If social science was to be
the scientific study of the human soul, it was necessary that there must be a
universal human soul separate from the human body. The result was a denial
of human biological diversity’s relevance to social behavior and, in its strong
form in Anthropology due to Boas and his followers, a denial of ongoing
evolution in humans of that soul. There was in universities something like a
truce between the biological sciences and the social sciences so that even the
great biologist Ernst Mayr could say that “We cannot escape the conclusion
that man’s evolution toward manness suddenly came to a halt” (Mayr, 1963).
FOUNDATIONAL RESEARCH
There have always been cracks in the strong social science stance but in the
past decade or so it has been noticeably weakening. A prominent example
has been psychiatry converting essentially to biological psychiatry. Advances
in biochemistry such as radioimmunoassay, allowing assays of hormones
and other chemicals at reasonable cost, led to a lively literature on the effects
of hormones and other chemicals on behavior.
Much of the study of genetic influences on behavior and personality relied
for decades on the similarity between genetic relatives, especially identical

Genetics and Social Behavior

3

and fraternal twins. Twins separated at birth were of special interest because
it was presumed that they did not share a common environment and their
resemblance reflected shared genes. A not unreasonable criticism of such
studies was that (i) they shared a common prenatal uterine environment and
(ii) foster families were not random families but were often chosen to match
the adoptees in terms of race, social class, and so on.
In the past several decades, new technology from genomics has allowed
genotyping of hundreds of thousands to millions of genetic markers in
humans, directly measuring similarity (i.e., kinship) between individuals.
This means, for example, we can look at similarities between third cousins
and compare those similarities to those between fifth cousins, cryptic cousins
as the individuals are complete strangers to each other. The possible errors
from looking at close genealogical relatives in quantitative genetic studies
are now bypassed (Purcell et al., 2009; Visscher, 2010), as the new methods
for estimating heritability have replicated the old estimates. An old nagging
source of doubt and controversy in behavior genetic studies is gone.
CUTTING EDGE RESEARCH
SINGLE LOCUS TRAITS
Much of human behavior is universal and adaptive, or used to be in past environments. We eat when hungry, fear spiders and snakes, love our children.
Individuals without these behaviors failed to reproduce, or to reproduce at
rates comparable to those with the behaviors. Such patterns can be thought
of as strategies, patterns or rules of conduct that, on average, led to reproductive success in the human past. However, many aspects of our behavior vary.
The differences show up early in life and do not seem to be affected much by
parental rearing style.
What are the sources of these differences between individuals? Genetic differences must contribute, as personality is heritable, in humans as in other
species (Kendler & Greenspan, 2006). To say that some trait is heritable is to
say that it is genetically transmissable to offspring. More precisely, heritability is a fraction between 0 and 1 that specifies the amount of trait diversity in
a population that reflects gene differences.
Environmental insults contribute as well, such as neurotropic viruses,
toxins, and other influences. However, here we are going to talk genetics.
Genetic influences on behavior fall into two fundamental categories: those
that are the product of natural selection, and those that are not.
To what extent is existing variation in human behavior adaptive? For
example, some people are scrupulous and hyper-moral while, at the other
extreme, some people are without apparent conscience. (We call them

4

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

sociopaths.) There is apparently a continuum from one extreme to the other,
with most people falling somewhere between the two extremes. Why should
not there be a single best strategy, that is, why is not everyone the same?
Often there is a single optimum, but not always. It depends on
whether the payoff in terms of fitness of a particular course of action
is frequency-dependent—in other words, whether it depends on the actions
of other individuals. Running from a forest fire pays off whether anyone
else does or not, but the payoff of running for tribal chief decreases as the
number of candidates increases. If everyone is passive (a “dove”), aggressive
individuals (hawks) prosper, but as hawks becomes more common, they
increasingly run into and fight other hawks, so their payoff decreases. If the
cost of fighting among hawks is high enough, the equilibrium solution is a
mixed state consisting of both hawks and doves (Smith & Price, 1973). Thus
there is no single best solution, no single optimal behavior. This is often the
case with social interactions. It is worth remembering that hawks do not
benefit the species as a whole: The species would do better if everyone just
got along. However, when the strategy benefits individual hawks and the
genes they carry, they increase in frequency.
A related issue is whether adaptive variation in behavior and personality
is fixed or flexible, heritable or not. Sometimes an individual can pick one of
several possible life strategies based on exterior clues. If a female bee larva is
fed royal jelly, she becomes a queen; otherwise, she becomes a sterile worker.
The capability to assume those different roles is adaptive and presumably a
result of natural selection, but it is not noticeably heritable, because all female
bees have it. In other cases, like those pesky fire ants that exhibit single-queen
and Los Angeles-style multiqueen colonies, the two morphs are determined
by the two alleles of a single supergene (Wang et al., 2013), and the behavior
difference is completely heritable. In the former case of queens and workers
the two outcomes reflect environment effects with no contribution from gene
differences while in the latter case of colony structure the outcomes reflect
simple gene differences with no contribution from the environment.
As we are arguably a lot smarter than ants or bees, one might think that
most adaptive personality variation in humans would be learned (a response
to exterior cues) rather than heritable. Certainly some is, but much variation
looks heritable. People do not seem to learn to be aggressive or shy—they just
are, and in those tendencies resemble their biological parents. There are models that may explain this. One is that jacks of all trades are masters of none:
If you play the same role all the time, you will be better at it than someone
who keeps switching personalities. It could be the case that such switching is
physiologically difficult and/or expensive. Moreover, in at least some cases,
being predictable has social value. Someone who is known to be implacably aggressive will always win at chicken (Kahn, Aligic˘a, & Weinstein, 2009).

Genetics and Social Behavior

5

Being known as the sort of guy who would rush into a burning building to
save ugly strangers may pay off, even though actually running into that blaze
does not.
In addition, if a particular role or personality type only became viable relatively recently—say, 10,000 years ago, in the early Neolthic—a mutation that
induces that personality may have become fairly common, but would not
yet be part of a precise and flexible system. The required modifier genes that
would turn those tendencies on when they pay and off when they do not,
would take longer to evolve. Evolution is ongoing, so many new adaptations
are likely to be imperfect and incomplete.
John Tooby and Leda Cosmides, leading founders of evolutionary psychology, have argued strongly against the possibility of heritable, adaptive
behavioral variation in humans (Tooby & Cosmides, 1990). Their arguments
imply that no such thing should exist in any species, but we have found
genetic morphs in lizards, birds, crustaceans, ants, and butterflies. They
are widespread in nature. However, one of their arguments is especially
interesting: they argue that adaptations, including behavioral adaptations,
are usually generated by complex, coadapted sets of genes (true), and that
such gene complexes would be broken up by sexual reproduction (also
true): Even if the dad had a set of genes that made him a natural blacksmith
or tap dancer, his kids would never inherit that whole set and the talent
should disappear.
They are on to something with this argument, but nature seems to have
found solutions. In some cases, a set of genes that determine the alternative
phenotype are arranged as a supergene, a group of genes that are physically
close and strongly linked. That keeps the gene complex from being broken
up. In other cases, the alternate forms are determined by a single gene that
acts as a switch. Different alleles of that gene specify different morphs. This
kind of mechanism is also largely unaffected by recombination.
Perhaps the heritable variant is in some way a simple trait. For example,
what if simply losing a particular complex adaptation was adaptive when
rare? It is easy to see that a complex behavioral adaptation could be stopped
cold by a single mutation. Alternatively, for that matter, what if greatly intensifying a particular behavior or drive—turning up the volume knob—was
adaptive when rare? It is possible to imagine a simple mutation that turns up
the volume in some way? Surely. In addition, of course, such simple initial
changes can be gradually refined by natural selection.
We know of many clear examples of distinct behavioral morphs or strategies in other species. We mentioned the two different kinds of fire-ant society.
Many of the examples are of male morphs with different reproductive strategies. One of the most interesting and well-known example is Uta stansburiana,
a species of lizard studied by Barry Sinervo (Sinervo & Lively, 1996). The

6

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

species has three male morphs with distinct color patterns: orange, yellow,
and blue. Orange males are large, aggressive, and control a territory with
several females. They dominate blue males, but yellow males, which mimic
females, often cuckold them. Blue males have a smaller territory with one
female, which they can effectively defend against yellow males. Orange beats
blue, blue beats yellow, and yellow beats orange. It is scissors-paper-rock.
Within-group adaptive behavioral variation is possible, as we know of
many examples in other species. Behavioral differences between geographically separated populations of the same species are also possible, as selection
pressures often vary with location. That means that genetically induced
adaptive behavioral variation could exist in humans—but does it?
The usual measures of personality all show significant heritability
(Turkheimer, 2011), which is consistent with adaptive personality variation
but also with nonadaptive genetic influences, such as mutational load. For
example, aggressive individuals could be present in a population as evolved
players of a “hawk” strategy in their interpersonal relations, or they could
simply be unfortunate bearers of deleterious genetic mutations leading to
their aggressive behavior. The case that would be clearest, easiest to prove,
would be one in which different variants of a single gene have a significant
influence on behavior. If such variants existed, they would almost certainly
be the product of adaptive evolution. Mutational pressure would create a
very different picture, with rare deleterious mutations of many different
genes, rather than two or three common variants of a single locus.
The most promising possible case of adaptive genetic variation in humans
is in MAOA, an enzyme that degrades several neutrotransmitters. The gene,
located on the X chromosome, contains a 30-base sequence that is repeated
different numbers of times in different MAOA variants—2R, 3R, 3.5R, 4R,
and 5R. These repeats are in a regulatory region and affect gene activity.
We know that complete loss of MAOA activity has a strong impact on
human behavior. A null mutation of MAOA has been seen in large Dutch
kindred (Brunner, Nelen, Breakefield, Ropers, & Van Oost, 1993). Males with
this mutation show impulsive aggressiveness and mild mental retardation.
Low levels of the gene products (rather than any particular variant) are
associated with trait aggression in men (Alia-Klein et al., 2008).
In mouse models, inactivation of MAOA causes increased aggressiveness
(Cases et al., 1995; Scott, Bortolato, Chen, & Shih, 2008). In rhesus monkeys,
there is an association between low activity MAOA alleles that is dependent
on early environment. There seems to a similar interaction effects in humans:
low activity MAOA alleles (the 3R and 2R variants) seem to increase vulnerability to environmental stresses such as abuse in childhood. In this case, the
environmental effect of being abused as a child, like the jelly to the female bee
larva, leads to behavior change in adulthood only if a specific genetic variant

Genetics and Social Behavior

7

is present. With this level of complexity, spreading of a trait as learned or not
loses its meaning.
Despite the special difficulties of performing these investigations in
humans, it appears there really are genetically induced behavioral variants
that are the product of selection. MAOA is not the only locus for which we
have evidence of this, but it is probably the best-understood, and makes the
point.
There are important aspects of human behavioral variation that are almost
certainly maladaptive, such as mental retardation and mental illness. They
too are heritable, and to a significant extent genetic in origin. That is not to
say that environmental insults and pathogens do not play a role. Prenatal
starvation doubled the incidence of schizophrenia in the cohorts affected by
the Dutch famine of 1944 and the Chinese famine associated with the Great
Leap Forward (St Clair et al., 2005; Susser & Lin, 1992). Neurosyphilis once
accounted for about half of the patients in psychiatric hospitals (Barondes,
1990). However, we have reduced the impact of syphilis and other pathogens
over the past century, as well as famine. Genetic causes have not diminished,
and therefore account for a larger fraction of cases than they once did.
Personality variations generally do not drastically reduce reproductive
fitness, but mental illness and mental retardation do. For example, in
schizophrenia, fitness is drastically lowered in men (about 25% of normal)
and significantly lowered in women as well, about 50% of normal (Bassett,
Bury, Hodgkinson, & Honer, 1996). As schizophrenia is quite heritable
(Tsuang, 2000) there is an apparent paradox: how can schizophrenia (and
other forms of mental illness) persist over time at moderately high frequency,
0.5–1%? (Saha, Chant, Welham, & McGrath, 2005).
Increasingly, it looks as if the cause is ongoing mutation. Our understanding of mutation in humans and its negative consequences for human health is
developing rapidly, because of advances in sequencing technology. It is now
possible to identify mutations by high accuracy whole-genome sequencing
of family triads: if you see a sequence in the child that does not exist in his
parents, it is a new mutation. (Of course it could also be parental assignment
error, but that is easily detected because there would be thousands of mismatches between the parental genes and those in the child.)
There are a number of recent reports that have found evidence for de novo
mutations as an important cause of schizophrenia and autism (Awadalla
et al., 2010; Girard et al., 2011). A high rate of ongoing mutations also explains
why a significant fraction of schizophrenia cases are sporadic: they appear
in families that do not have previous cases of schizophrenia (Xu et al., 2011).
These techniques have also confirmed that most mutations originate in
the father, and that the mutation rate increases linearly with post-pubertal
father’s age.

8

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Many problems are more common in the children of older fathers, but disorders that result from impaired brain function, such as autism, schizophrenia,
and reduced intelligence, are particularly common (Kondrashov, 2012). This
seems to be a consequence of the great complexity of brain development and
function, making use of the majority of all genes.
Schizophrenia must be made up of many different mutational disorders.
At the same time, a given mutation often seems to increase risk for several different mental disorders. (Cross-Disorder Group of the Psychiatric
Genomes Consortium, 2013). Clearly, progress in this area has been hampered by simultaneously lumping together causally unrelated syndromes
and splitting a syndrome with a single cause into several different categories.
On the other hand, if this new picture is correct, it may suggest new and
effective approaches for fighting mental illness. In the short run, finding ways
to reduce the mutation rate might pay off, especially if any factors other than
paternal age are found to materially influence that rate. In the longer run,
gene therapy may offer hope.
QUANTITATIVE TRAITS
Many regard the origin of evolutionary biology as a science to be the
publication by R.A. Fisher (1918) of a paper titled “The correlation between
relatives under the supposition of Mendelian inheritance.” In it he showed
that if a quantitative trait, such as height or blood pressure that is measured
on a continuous scale, was determined by an environmental effect and
a large number of genes, each of small effect, the trait should follow a
Normal or Gaussian distribution in the population. The model predicts
correlations between relatives of different degrees. Before this publication
there was widespread doubt about Darwin’s theory and conflict between
two traditions—Mendelians, who sought to understand variation using
Mendel’s principles, and the Biometricians such as Galton and Pearson, who
doubted that Mendel’s insights were relevant to quantitative traits. With the
two schools reconciled, biologists stopped throwing stones at each other,
and by the 1930s modern evolutionary theory had essentially taken shape.
Fisher defined the heritability of a quantitative trait as the fraction of the
population variance of the trait was due to genetic variation, in particular
to additive genetic variation, additive variation being variation that is transmitted from parent to offspring. By convention, heritability is written as h2 .
The model led to the “breeder’s equation” giving the predicted response of a
trait to selection as r = h2 s where r is the response to selection, the difference
between the average value in the population before selection and the average value of the offspring of selected parents. This equation is the workhorse
of quantitative genetics as it relates the intensity of selection on a trait to the

Genetics and Social Behavior

9

rate of change in the character that would result. It seems too simple, yet in
general it works very well.
A general finding is that nearly everything one can measure in humans is
partially heritable, from stature and IQ on the high end (∼0.80) to personality
traits from questionnaires on the low (∼0.25) (Turkheimer, 2011). Such estimates are obtained from correlations between relatives, most famously from
the extent of similarity between pairs of identical twins. The implication from
the generally positive findings is that many of the differences among us that
have generally been assumed to be entirely learned, such as how religious
we are or whether we are politically liberal or conservative, reflect in part
gene differences. This suggests that group differences without our species
also reflect in part gene differences, but there is little direct exploration of
that possibility yet in the literature.
Heritability studies in humans have always been subject to doubts and criticisms. For example, similarities between identical twins separated early in
life are a staple of the literature, but there remained doubt about whether their
environments were really different. Further, even if separated at birth they
had shared a uterine environment during gestation and this shared environment might account for similarities. These shadows of doubt, together with
the old idea that gene differences are not the proper domain for the social sciences, have kept the impact of heritability studies low, or at least lower than
they ought to be.
Most of these doubts and reservations are now resolved with the use of SNP
(single nucleotide polymorphism) chips that provide for any individual his
or her genotype at several hundred thousand or more genetic loci. These loci
are not necessarily or even usually functionally significant, but these chips
allow us to measure overall genetic similarity, called kinship, between any
two individuals. “Kinship” in common usage means degree of relationship
computed from a pedigree, while in genetics, kinship often means simply
overall genetic similarity. The meanings are equivalent in practice, but kinship from chips is more accurate. For example, from a pedigree my kinship
with my child is 1/4 or 0.25, but in fact my kinship with my child might be
as low as 0.2 or so or as high as 0.30 or more. (Social scientists may be more
familiar with what is called the coefficient of relationship rather than kinship:
roughly relationship is twice kinship.)
In a series of papers, Peter Visscher and colleagues have computed kinship
between all pairs of individuals from large longitudinal studies and looked
at distant kinship and interesting quantitative traits such as stature (Yang
et al., 2010), and intelligence (Davies et al., 2011) and schizophrenia and
bipolar disorder (Purcell et al., 2009). In order to avoid the possibility of
shared environments any pair more closely related than third cousins were
not used. They showed that the new method using computed genetic

10

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

similarity yielded that same heritability estimates that had been known for
years, essentially falsifying the doubts and quibbles that had plagued human
quantitative genetics for decades. These papers also showed quite clearly
that Fisher’s original model of 1918 was in fact the right description of the
inheritance mechanism: for both stature and intelligence our differences are
caused by the small effects of a very large number of genes scattered over
the genome, ending hopefully the futile search for “genes for” one thing and
another. The details are somewhat technical: a lucid explanation is given by
Visscher (2010) in a commentary on his earlier paper.
Epigenetic transmission has received much attention in the past several
years. In the simplest case, some environmental effect causes DNA to be
modified so that a gene is turned off or turned on, and this modified gene
is transmitted to later generations. The modification is not permanent and
the modification can be reversed (reset) in the future. We have so far no clear
picture of its importance in human social behavior. The new SNP technology will certainly be an important tool in following this up, as a correlation
between, say, parent and child would be inflated by shared epigenetics while
correlations between more distant relatives would not. The SNP approach
informs about correlations between, for example, sixth cousins, who have
no idea of their relationship and who would not share epigenetic effects as
parent and child or siblings would (Tal, Kisdi, & Jablonka, 2010).
An important category of model is the threshold model in which there is
an underlying Normal distribution of some unobserved trait. Individuals
with more than the threshold value of the underlying trait exhibit the phenotype under study. In an important series of papers in the 1970s Robert
Cloninger and others proposed that sociopathy, defined by rigid explicit criteria, followed such a model in which the thresholds were different in males
and females (Cloninger, Reich, & Guze, 1975a, 1975b; Reich, Cloninger, &
Guze, 1975): the threshold for sociopathic behavior was higher in females.
It required a higher concentration of the underlying trait to cause outright
sociopathy in women.
Threshold models provide useful insight even without assuming anything
about a transmission mechanism. For example, Eisner (2001) discusses the
decline in homicide victimization rates in Europe from the Middle Ages to
the present. A typical comparison in his data is from 50 homicides per 100,000
population in the year 1200 to 2 per 100,000 today, a 25-fold decline. This
seems to be a big decline, but is it really? How does it compare to a decline
from 100 to 4 per 100,000?
A model, perhaps a poor model but better than no model initially, is
to imagine that homicide is a reflection of some underlying normally
distributed genetic trait that has varied over time, perhaps because of
selection against violence. A rate of 50/100,000 implies that this fraction of

Genetics and Social Behavior

11

the population, 0.05%, is homicidal. If the underlying trait has a standard
normal distribution, 0.05% corresponds to the fraction of the population
that is >3.3 standard deviations from the mean. A decline to 2/100,000,
0.002%, corresponds to 4.1 standard deviations from the mean. Assuming
that selection has simply shifted the underlying distribution of violence
proneness to the left without changing the variance, which is a consequence
of Fisher’s model and often seen in experiments with animals, the amount
of shift is 4.1–3.3 or 0.8 standard deviations. In more familiar terms this 0.8
standard deviations would correspond to about 2 inches in height or 12 IQ
points, not a very impressive change at all over 900 years.
What if were studying another social indicator, 100 times as common, that
changed from in the same time from 50 to 2 per 1000 rather than 100,000. With
the same assumptions about an unchanging variance the implied change in
the underlying distribution is 1.3 standard deviations over 900 years, nearly
double the change in proneness to homicide, corresponding to a change in
height of 3.5 inches or a change in IQ of nearly 20 points. If this were due
to selection, the implied selection intensity against the latter trait is much
greater.
None of this is to claim that homicide or drunk driving or anything else is
“genetic.” If we had more information we could investigate the heritability of
the trait(s) and build more accurate and more useful models. Even so, thinking in terms of a threshold and an underlying Normal distribution provides
an important and useful way to understand changes in discrete traits over
time.
ISSUES FOR FUTURE RESEARCH
Genetics and genetic approaches will become more and more important in
social science research and understanding. Specific loci such as MAOA that
we discussed earlier, and numerous others, are of limited interest as few of
these loci are very significant in determining everyday human differences.
This means that the practice of looking for “genes for” one thing and another
will lose salience in social science. While the occasional locus may remain
salient and important, it is increasingly apparent that most of our diversity
reflects our small differences at large and very larger numbers of genetic
loci, the effect of each locus so small as to be essentially undetectable. While
the old paradigm of looking for “genes for” one thing and another will
persist it will surely fade in the face of the even older quantitative genetic
models as they come back for their place in the social sciences. The recent
use of SNP chips has essentially resolved old doubts and controversies
about biases and inaccuracies in heritability estimates. The standard model

12

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

of genetic influence being the combined small effects of a large number of
genes scattered throughout the genome, fits most data quite well.

REFERENCES
Alia-Klein, N., Goldstein, R. Z., Kriplani, A., Logan, J., Tomasi, D., Williams, B., … ,
Fowler, J. S. (2008). Brain monoamine oxidase-A activity predicts trait aggression.
Journal of Neuroscience 28, 5099–5104.
Awadalla, P., Gauthier, J., Myers, R. A., Casals, F., Hamdan, F. F., Griffing, A. R., …
Tarabeux, J. (2010). Direct measure of the de novo mutation rate in autism and
schizophrenia cohorts. The American Journal of Human Genetics 87, 316–324.
Barondes, S. H. (1990). The biological approach to psychiatry: History and prospects.
The Journal of Neuroscience, 10, 1707–1710.
Bassett, A. S., Bury, A., Hodgkinson, K. A., & Honer, W. G. (1996). Reproductive
fitness in familial schizophrenia. Schizophrenia Research, 21, 151–160.
Brunner, H. G., Nelen, M., Breakefield, X. O., Ropers, H. H., & Van Oost, B. A. (1993).
Abnormal behavior associated with a point mutation in the structural gene for
monoamine oxidase A. Science, 262(5133), 578–580.
Cases, O., Seif, I., Grimsby, J., Gaspar, P., Chen, K., Pournin, S., … Shih, J. C. (1995).
Aggressive behavior and altered amounts of brain serotonin and norepinephrine
in mice lacking MAOA. Science 268, 1763.
Cloninger, C. R., Reich, T., & Guze, S. B. (1975a). The multifactorial model of disease
transmission: III. Familial relationship between sociopathy and hysteria (Briquet’s
syndrome). British Journal of Psychiatry, 127, 23–32.
Cloninger, R. C., Reich, T., & Guze, S. B. (1975b). The multifactorial model of disease transmission: II. Sex differences in the familial transmission of sociopathy
(antisocial personality). British Journal of Psychiatry, 127, 11–22.
Cross-Disorder Group of the Psychiatric Genomes Consortium (2013). Identification
of risk loci with shared effects on five major psychiatric disorders: A genome-wide
analysis. The Lancet, 381(9875), 1371–1379.
Davies, G., Tenesa, A., Payton, A., Yang, J., Harris, S. E., Liewald, D., … Deary, I.
J. (2011). Genome-wide association studies establish that human intelligence is
highly heritable and polygenic. Molecular Psychiatry, 16(10), 996–1005.
Eisner, M. (2001). Modernization, self-control and lethal violence. The long-term
dynamics of European homicide rates in theoretical perspective. British Journal of
Criminology, 41(4), 618–638.
Fisher, R. A. (1918). The correlation between relatives on the supposition of
Mendelian inheritance. Transactions of the Royal Society of Edinburgh, 52(2), 399–433.
Girard, S. L., Gauthier, J., Noreau, A., Xiong, L., Zhou, S., Jouan, L., … Diallo, O.
(2011). Increased exonic de novo mutation rate in individuals with schizophrenia.
Nature Genetics 43, 860–863.
Kahn, H., Aligic˘a, P. D., & Weinstein, K. R. (2009). The essential Herman Kahn: In
Defense of thinking. Lanham, MD: Lexington Books.

Genetics and Social Behavior

13

Kendler, M. D., & Greenspan, P. D. (2006). The nature of genetic influences on
behavior: Lessons from “Simpler” organisms. American Journal of Psychiatry, 163,
1683–1694.
Kondrashov, A. (2012). Genetics: The rate of human mutation. Nature, 488, 467–468.
Mayr, E. (1963). Animal species and evolution. Cambridge, England: Harvard University Press.
Morawski, J. G. (1986). Organizing knowledge and behavior at Yale’s Institute of
human relations. Isis, 77(2), 219–242.
Purcell, S. M., Wray, N. R., Stone, J. L., Visscher, P. M., O’Donovan, M. C., Sullivan,
P. F., … Morris, D. W. (2009). Common polygenic variation contributes to risk of
schizophrenia and bipolar disorder. Nature, 460(7256), 748–752.
Reich, R., Cloninger, C. R., & Guze, S. B. (1975). The multifactorial model of disease
transmission: I. Description of the model and its use in psychiatry. British Journal
of Psychiatry, 127, 1–10.
Saha, S., Chant, D., Welham, J., & McGrath, J. (2005). A systematic review of the
prevalence of schizophrenia. PLoS Medicine, 2, e141.
Scott, A. L., Bortolato, M., Chen, K., & Shih, J. C. (2008). Novel monoamine oxidase A
knock out mice with human-like spontaneous mutation. Neuroreport, 19, 739–743.
Sinervo, B., & Lively, C. M. (1996). The rock-paper-scissors game and the evolution
of alternative male strategies. Nature, 380, 240–243.
Smith, J. M., & Price, G. R. (1973). The logic of animal conflict. Nature, 246, 15.
St Clair, D., Xu, M., Wang, P., Yu, Y., Fang, Y., Zhang, F., … Sham, P. (2005). Rates
of adult schizophrenia following prenatal exposure to the Chinese famine of
1959–1961. JAMA, the Journal of the American Medical Association 294, 557–562.
Susser, E. S., & Lin, S. P. (1992). Schizophrenia after prenatal exposure to the Dutch
Hunger Winter of 1944–1945. Archives of General Psychiatry, 49, 983.
Tal, O., Kisdi, E., & Jablonka, E. (2010). Epigenetic contribution to covariance between
relatives. Genetics, 184(4), 1037–1050.
Tooby, J., & Cosmides, L. (1990). On the universality of human nature and the uniqueness of the individual: The role of genetics and adaptation. Journal of Personality,
58, 17–67.
Tsuang, M. (2000). Schizophrenia: Genes and environment. Biological Psychiatry, 47,
210–220.
Turkheimer, E. (2011). Still missing. Research in Human Development, 8(3–4), 227–241.
Visscher, P. M. (2010). A commentary on ‘common SNPs explain a large proportion
of the heritability for human height’ by Yang et al. (2010). Twin Research and Human
Genetics, 13(6), 517.
Wang, J., Wurm, Y., Nipitwattanaphon, M., Riba-Grognuz, O., Huang, Y.-C., Shoemaker, D., & Keller, L. (2013). A Y-like social chromosome causes alternative
colony organization in fire ants. Nature, 493, 664–668.
Xu, B., Roos, J. L., Dexheimer, P., Boone, B., Plummer, B., Levy, S., ... Karayiorgou, M. (2011). Exome sequencing supports a de novo mutational paradigm for
schizophrenia. Nature Genetics, 43, 864–868.

14

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Yang, J., Benyamin, B., McEvoy, B. P., Gordon, S., Henders, A. K., Nyholt, D. R., …
Visscher, P. M. (2010). Common SNPs explain a large proportion of the heritability
for human height. Nature Genetics, 42(7), 565–569.

FURTHER READING
Reich, R., Cloninger, C. R., & Guze, S. B. (1975). The multifactorial model of disease
transmission: I. Description of the model and its use in psychiatry. British Journal
of Psychiatry, 127, 1–10.
Turkheimer, E. (2011). Still missing. Research in Human Development, 8(3–4), 227–241.
Visscher, P. M. (2010). A commentary on ‘common SNPs explain a large proportion
of the heritability for human height’ by Yang et al. (2010). Twin Research and Human
Genetics, 13(6), 517.

HENRY HARPENDING SHORT BIOGRAPHY
Henry Harpending is Professor of Anthropology at the University of Utah.
His academic webpage is http://harpending.humanevo.utah.edu. He has
spent several long fieldtrips in the northern Kalahari studying genetics and
family organization of foraging groups there as well as ranchers. He has also
had a parallel career working on human biological and cultural diversity,
human genetics, and modern human origins. Cochran and Harpending published The Ten Thousand Year Explosion (Basic Books) in 2010.
GREGORY COCHRAN SHORT BIOGRAPHY
Gregory Cochran is a physicist and an adjunct Professor of Anthropology
at the University of Utah. His early career was in optical physics working
in our aerospace industry. A decade or so ago, his interests turned to theoretical biology and anthropology. He has written about an infectious model
of human male homosexuality, immune system diversity, and immune system evolution among human continental populations, and acceleration of the
rate of evolution in humans, especially since the end of the ice ages and the
origins of agriculture.
RELATED ESSAYS
Telomeres (Psychology), Nancy Adler and Aoife O’Donovan
Social Epigenetics: Incorporating Epigenetic Effects as Social Cause and
Consequence (Sociology), Douglas L. Anderton and Kathleen F. Arcaro
Kin-Directed Behavior in Primates (Anthropology), Carol M. Berman
The Sexual Division of Labor (Anthropology), Rebecca Bliege Bird and Brian
F. Codding

Genetics and Social Behavior

15

Genetics and the Life Course (Sociology), Evan Charney
Genetic and Environmental Approaches to Political Science (Political Science),
Zoltán Fazekas and Peter K. Hatemi
Food Sharing (Anthropology), Michael Gurven and Adrian V. Jaeggi
An Evolutionary Perspective on Developmental Plasticity (Psychology),
Sarah Hartman and Jay Belsky
Grandmothers and the Evolution of Human Sociality (Anthropology), Kristen
Hawkes and James Coxworth
The Neurobiology and Physiology of Emotions: A Developmental Perspective (Psychology), Sarah S. Kahle and Paul D. Hastings
Herd Behavior (Psychology), Tatsuya Kameda and Reid Hastie
Genetic Foundations of Attitude Formation (Political Science), Christian
Kandler et al.
Cultural Neuroscience: Connecting Culture, Brain, and Genes (Psychology),
Shinobu Kitayama and Sarah Huff
Reconciliation and Peace-Making: Insights from Studies on Nonhuman
Animals (Anthropology), Sonja E. Koski
Cooperative Breeding and Human Evolution (Anthropology), Karen L.
Kramer
Niche Construction: Implications for Human Sciences (Anthropology), Kevin
N. Laland and Michael O’Brien
Culture, Diffusion, and Networks in Social Animals (Anthropology), Janet
Mann and Lisa Singh
Evolutionary Perspectives on Animal and Human Personality (Anthropology), Joseph H. Manson and Lynn A. Fairbanks
Neural and Cognitive Plasticity (Psychology), Eduardo Mercado III
Evolutionary Theory and Political Behavior (Political Science), Michael Bang
Petersen and Lene Aarøe
Gestural Communication in Nonhuman Species (Anthropology), Simone Pika
Darwinism as a Decryption Key for the Human Mind (Psychology), Csaba
Pléh and Ottilia Boross
DNA Revolution and the Social and Behavioral Sciences (Psychology), Maciej
Trzaskowski and Robert Plomin
Born This Way: Thinking Sociologically about Essentialism (Sociology),
Kristen Schilt
Vocal Communication in Primates (Anthropology), Katie E. Slocombe
Primate Allomaternal Care (Anthropology), Stacey Tecot and Andrea Baden
How Form Constrains Function in the Human Brain (Psychology), Timothy
D. Verstynen
A Gene-Environment Approach to Understanding Youth Antisocial
Behavior (Psychology), Rebecca Waller et al.



Genetics and Social Behavior
HENRY HARPENDING and GREGORY COCHRAN

Abstract
We focus on the effects of gene differences on social and behavioral differences
among individuals and among larger groups of individuals. Many specific genetic
markers are known that influence aspects of personality and behavior. The focus
on single genes and groups of genes is giving way to quantitative genetics, the
statistical study of transmission of characteristics viewed as the outcome of the
effects of very large numbers of genes. While traditional social science largely
ignores the effects of genetically transmitted influences, the subject persists and
grows in importance. Classical quantitative genetic methods may give much insight
into human behavioral diversity and they provide the “right” way to measure
and assess variation in rates of threshold traits. We discuss examples, trends, and
possibilities for the incorporation of genetic data and models in the social and
behavioral sciences without advocating major changes in practice.

INTRODUCTION
There are two distinct ways we could approach the task of reviewing genetics
and human social behavior. One is to review aspects that are genetic in the
sense of hard-wired in the human genome and universal. For example, we
all learned spoken language early in our lives and most agree that the trait
“learn to speak the language(s) in your social environment” is such a universal, while the specific language(s) we learned were entirely dependent on our
rearing environment. In spite of the existence of people who never learned to
speak, we do not hesitate to regard learning speech as “genetic.” However,
things are not so simple: speech does not have to be taught while literacy
does. We are all literate, but many of us struggled to learn, say, calculus, in a
way we never struggled with literacy. This would lead to a consideration of
the nature of cognitive modules, a lively area of current research.
Another way to proceed is to consider how genetic differences among people or among groups of people lead to or have been generated by individual
or group differences in social behavior. Are individual differences in social
behavior determined by gene differences? Do group level gene differences
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.

1

2

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

modulate group differences in social behavior? A century ago, the response
of most educated people would have been “of course” to questions such as
these, whereas today, the response of many educated people and scientists is
“of course not.” We suggest that this change has gone too far, motivated in
many cases by political undercurrents, and that an important task for social
science it to understand and investigate gene differences.
In the past century, many anthropologists were interested in temperament,
for example, and culture and personality was a prominent subfield. It is
striking that in all that literature, the possibility of such things reflecting gene
differences was never on the table. Roughly between 1900 and 1950, there
had emerged a new cluster of academic disciplines, social science. The vision
was that this would be a third kind of knowledge, somewhere between the
humanities and the sciences, a branch that would stand between and perhaps unite human knowledge. The optimism and ambition of practitioners
of the new disciplines were great and they were contagious so, for example,
Yale University created the Institute of Human Relations in the 1930,
“directly concerned with the problems of man’s individual and group conduct” (Morawski, 1986). Harvard University created what was practically a
whole new division of the university with its Department of Social Relations
in 1946, a hybrid of sociology, anthropology, and psychology.
A simple, indeed simplistic, statement of the origin of social science in its
early years was that it insisted on the separation of the body and soul, the soul
being the locus of social behavior and personality. If social science was to be
the scientific study of the human soul, it was necessary that there must be a
universal human soul separate from the human body. The result was a denial
of human biological diversity’s relevance to social behavior and, in its strong
form in Anthropology due to Boas and his followers, a denial of ongoing
evolution in humans of that soul. There was in universities something like a
truce between the biological sciences and the social sciences so that even the
great biologist Ernst Mayr could say that “We cannot escape the conclusion
that man’s evolution toward manness suddenly came to a halt” (Mayr, 1963).
FOUNDATIONAL RESEARCH
There have always been cracks in the strong social science stance but in the
past decade or so it has been noticeably weakening. A prominent example
has been psychiatry converting essentially to biological psychiatry. Advances
in biochemistry such as radioimmunoassay, allowing assays of hormones
and other chemicals at reasonable cost, led to a lively literature on the effects
of hormones and other chemicals on behavior.
Much of the study of genetic influences on behavior and personality relied
for decades on the similarity between genetic relatives, especially identical

Genetics and Social Behavior

3

and fraternal twins. Twins separated at birth were of special interest because
it was presumed that they did not share a common environment and their
resemblance reflected shared genes. A not unreasonable criticism of such
studies was that (i) they shared a common prenatal uterine environment and
(ii) foster families were not random families but were often chosen to match
the adoptees in terms of race, social class, and so on.
In the past several decades, new technology from genomics has allowed
genotyping of hundreds of thousands to millions of genetic markers in
humans, directly measuring similarity (i.e., kinship) between individuals.
This means, for example, we can look at similarities between third cousins
and compare those similarities to those between fifth cousins, cryptic cousins
as the individuals are complete strangers to each other. The possible errors
from looking at close genealogical relatives in quantitative genetic studies
are now bypassed (Purcell et al., 2009; Visscher, 2010), as the new methods
for estimating heritability have replicated the old estimates. An old nagging
source of doubt and controversy in behavior genetic studies is gone.
CUTTING EDGE RESEARCH
SINGLE LOCUS TRAITS
Much of human behavior is universal and adaptive, or used to be in past environments. We eat when hungry, fear spiders and snakes, love our children.
Individuals without these behaviors failed to reproduce, or to reproduce at
rates comparable to those with the behaviors. Such patterns can be thought
of as strategies, patterns or rules of conduct that, on average, led to reproductive success in the human past. However, many aspects of our behavior vary.
The differences show up early in life and do not seem to be affected much by
parental rearing style.
What are the sources of these differences between individuals? Genetic differences must contribute, as personality is heritable, in humans as in other
species (Kendler & Greenspan, 2006). To say that some trait is heritable is to
say that it is genetically transmissable to offspring. More precisely, heritability is a fraction between 0 and 1 that specifies the amount of trait diversity in
a population that reflects gene differences.
Environmental insults contribute as well, such as neurotropic viruses,
toxins, and other influences. However, here we are going to talk genetics.
Genetic influences on behavior fall into two fundamental categories: those
that are the product of natural selection, and those that are not.
To what extent is existing variation in human behavior adaptive? For
example, some people are scrupulous and hyper-moral while, at the other
extreme, some people are without apparent conscience. (We call them

4

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

sociopaths.) There is apparently a continuum from one extreme to the other,
with most people falling somewhere between the two extremes. Why should
not there be a single best strategy, that is, why is not everyone the same?
Often there is a single optimum, but not always. It depends on
whether the payoff in terms of fitness of a particular course of action
is frequency-dependent—in other words, whether it depends on the actions
of other individuals. Running from a forest fire pays off whether anyone
else does or not, but the payoff of running for tribal chief decreases as the
number of candidates increases. If everyone is passive (a “dove”), aggressive
individuals (hawks) prosper, but as hawks becomes more common, they
increasingly run into and fight other hawks, so their payoff decreases. If the
cost of fighting among hawks is high enough, the equilibrium solution is a
mixed state consisting of both hawks and doves (Smith & Price, 1973). Thus
there is no single best solution, no single optimal behavior. This is often the
case with social interactions. It is worth remembering that hawks do not
benefit the species as a whole: The species would do better if everyone just
got along. However, when the strategy benefits individual hawks and the
genes they carry, they increase in frequency.
A related issue is whether adaptive variation in behavior and personality
is fixed or flexible, heritable or not. Sometimes an individual can pick one of
several possible life strategies based on exterior clues. If a female bee larva is
fed royal jelly, she becomes a queen; otherwise, she becomes a sterile worker.
The capability to assume those different roles is adaptive and presumably a
result of natural selection, but it is not noticeably heritable, because all female
bees have it. In other cases, like those pesky fire ants that exhibit single-queen
and Los Angeles-style multiqueen colonies, the two morphs are determined
by the two alleles of a single supergene (Wang et al., 2013), and the behavior
difference is completely heritable. In the former case of queens and workers
the two outcomes reflect environment effects with no contribution from gene
differences while in the latter case of colony structure the outcomes reflect
simple gene differences with no contribution from the environment.
As we are arguably a lot smarter than ants or bees, one might think that
most adaptive personality variation in humans would be learned (a response
to exterior cues) rather than heritable. Certainly some is, but much variation
looks heritable. People do not seem to learn to be aggressive or shy—they just
are, and in those tendencies resemble their biological parents. There are models that may explain this. One is that jacks of all trades are masters of none:
If you play the same role all the time, you will be better at it than someone
who keeps switching personalities. It could be the case that such switching is
physiologically difficult and/or expensive. Moreover, in at least some cases,
being predictable has social value. Someone who is known to be implacably aggressive will always win at chicken (Kahn, Aligic˘a, & Weinstein, 2009).

Genetics and Social Behavior

5

Being known as the sort of guy who would rush into a burning building to
save ugly strangers may pay off, even though actually running into that blaze
does not.
In addition, if a particular role or personality type only became viable relatively recently—say, 10,000 years ago, in the early Neolthic—a mutation that
induces that personality may have become fairly common, but would not
yet be part of a precise and flexible system. The required modifier genes that
would turn those tendencies on when they pay and off when they do not,
would take longer to evolve. Evolution is ongoing, so many new adaptations
are likely to be imperfect and incomplete.
John Tooby and Leda Cosmides, leading founders of evolutionary psychology, have argued strongly against the possibility of heritable, adaptive
behavioral variation in humans (Tooby & Cosmides, 1990). Their arguments
imply that no such thing should exist in any species, but we have found
genetic morphs in lizards, birds, crustaceans, ants, and butterflies. They
are widespread in nature. However, one of their arguments is especially
interesting: they argue that adaptations, including behavioral adaptations,
are usually generated by complex, coadapted sets of genes (true), and that
such gene complexes would be broken up by sexual reproduction (also
true): Even if the dad had a set of genes that made him a natural blacksmith
or tap dancer, his kids would never inherit that whole set and the talent
should disappear.
They are on to something with this argument, but nature seems to have
found solutions. In some cases, a set of genes that determine the alternative
phenotype are arranged as a supergene, a group of genes that are physically
close and strongly linked. That keeps the gene complex from being broken
up. In other cases, the alternate forms are determined by a single gene that
acts as a switch. Different alleles of that gene specify different morphs. This
kind of mechanism is also largely unaffected by recombination.
Perhaps the heritable variant is in some way a simple trait. For example,
what if simply losing a particular complex adaptation was adaptive when
rare? It is easy to see that a complex behavioral adaptation could be stopped
cold by a single mutation. Alternatively, for that matter, what if greatly intensifying a particular behavior or drive—turning up the volume knob—was
adaptive when rare? It is possible to imagine a simple mutation that turns up
the volume in some way? Surely. In addition, of course, such simple initial
changes can be gradually refined by natural selection.
We know of many clear examples of distinct behavioral morphs or strategies in other species. We mentioned the two different kinds of fire-ant society.
Many of the examples are of male morphs with different reproductive strategies. One of the most interesting and well-known example is Uta stansburiana,
a species of lizard studied by Barry Sinervo (Sinervo & Lively, 1996). The

6

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

species has three male morphs with distinct color patterns: orange, yellow,
and blue. Orange males are large, aggressive, and control a territory with
several females. They dominate blue males, but yellow males, which mimic
females, often cuckold them. Blue males have a smaller territory with one
female, which they can effectively defend against yellow males. Orange beats
blue, blue beats yellow, and yellow beats orange. It is scissors-paper-rock.
Within-group adaptive behavioral variation is possible, as we know of
many examples in other species. Behavioral differences between geographically separated populations of the same species are also possible, as selection
pressures often vary with location. That means that genetically induced
adaptive behavioral variation could exist in humans—but does it?
The usual measures of personality all show significant heritability
(Turkheimer, 2011), which is consistent with adaptive personality variation
but also with nonadaptive genetic influences, such as mutational load. For
example, aggressive individuals could be present in a population as evolved
players of a “hawk” strategy in their interpersonal relations, or they could
simply be unfortunate bearers of deleterious genetic mutations leading to
their aggressive behavior. The case that would be clearest, easiest to prove,
would be one in which different variants of a single gene have a significant
influence on behavior. If such variants existed, they would almost certainly
be the product of adaptive evolution. Mutational pressure would create a
very different picture, with rare deleterious mutations of many different
genes, rather than two or three common variants of a single locus.
The most promising possible case of adaptive genetic variation in humans
is in MAOA, an enzyme that degrades several neutrotransmitters. The gene,
located on the X chromosome, contains a 30-base sequence that is repeated
different numbers of times in different MAOA variants—2R, 3R, 3.5R, 4R,
and 5R. These repeats are in a regulatory region and affect gene activity.
We know that complete loss of MAOA activity has a strong impact on
human behavior. A null mutation of MAOA has been seen in large Dutch
kindred (Brunner, Nelen, Breakefield, Ropers, & Van Oost, 1993). Males with
this mutation show impulsive aggressiveness and mild mental retardation.
Low levels of the gene products (rather than any particular variant) are
associated with trait aggression in men (Alia-Klein et al., 2008).
In mouse models, inactivation of MAOA causes increased aggressiveness
(Cases et al., 1995; Scott, Bortolato, Chen, & Shih, 2008). In rhesus monkeys,
there is an association between low activity MAOA alleles that is dependent
on early environment. There seems to a similar interaction effects in humans:
low activity MAOA alleles (the 3R and 2R variants) seem to increase vulnerability to environmental stresses such as abuse in childhood. In this case, the
environmental effect of being abused as a child, like the jelly to the female bee
larva, leads to behavior change in adulthood only if a specific genetic variant

Genetics and Social Behavior

7

is present. With this level of complexity, spreading of a trait as learned or not
loses its meaning.
Despite the special difficulties of performing these investigations in
humans, it appears there really are genetically induced behavioral variants
that are the product of selection. MAOA is not the only locus for which we
have evidence of this, but it is probably the best-understood, and makes the
point.
There are important aspects of human behavioral variation that are almost
certainly maladaptive, such as mental retardation and mental illness. They
too are heritable, and to a significant extent genetic in origin. That is not to
say that environmental insults and pathogens do not play a role. Prenatal
starvation doubled the incidence of schizophrenia in the cohorts affected by
the Dutch famine of 1944 and the Chinese famine associated with the Great
Leap Forward (St Clair et al., 2005; Susser & Lin, 1992). Neurosyphilis once
accounted for about half of the patients in psychiatric hospitals (Barondes,
1990). However, we have reduced the impact of syphilis and other pathogens
over the past century, as well as famine. Genetic causes have not diminished,
and therefore account for a larger fraction of cases than they once did.
Personality variations generally do not drastically reduce reproductive
fitness, but mental illness and mental retardation do. For example, in
schizophrenia, fitness is drastically lowered in men (about 25% of normal)
and significantly lowered in women as well, about 50% of normal (Bassett,
Bury, Hodgkinson, & Honer, 1996). As schizophrenia is quite heritable
(Tsuang, 2000) there is an apparent paradox: how can schizophrenia (and
other forms of mental illness) persist over time at moderately high frequency,
0.5–1%? (Saha, Chant, Welham, & McGrath, 2005).
Increasingly, it looks as if the cause is ongoing mutation. Our understanding of mutation in humans and its negative consequences for human health is
developing rapidly, because of advances in sequencing technology. It is now
possible to identify mutations by high accuracy whole-genome sequencing
of family triads: if you see a sequence in the child that does not exist in his
parents, it is a new mutation. (Of course it could also be parental assignment
error, but that is easily detected because there would be thousands of mismatches between the parental genes and those in the child.)
There are a number of recent reports that have found evidence for de novo
mutations as an important cause of schizophrenia and autism (Awadalla
et al., 2010; Girard et al., 2011). A high rate of ongoing mutations also explains
why a significant fraction of schizophrenia cases are sporadic: they appear
in families that do not have previous cases of schizophrenia (Xu et al., 2011).
These techniques have also confirmed that most mutations originate in
the father, and that the mutation rate increases linearly with post-pubertal
father’s age.

8

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Many problems are more common in the children of older fathers, but disorders that result from impaired brain function, such as autism, schizophrenia,
and reduced intelligence, are particularly common (Kondrashov, 2012). This
seems to be a consequence of the great complexity of brain development and
function, making use of the majority of all genes.
Schizophrenia must be made up of many different mutational disorders.
At the same time, a given mutation often seems to increase risk for several different mental disorders. (Cross-Disorder Group of the Psychiatric
Genomes Consortium, 2013). Clearly, progress in this area has been hampered by simultaneously lumping together causally unrelated syndromes
and splitting a syndrome with a single cause into several different categories.
On the other hand, if this new picture is correct, it may suggest new and
effective approaches for fighting mental illness. In the short run, finding ways
to reduce the mutation rate might pay off, especially if any factors other than
paternal age are found to materially influence that rate. In the longer run,
gene therapy may offer hope.
QUANTITATIVE TRAITS
Many regard the origin of evolutionary biology as a science to be the
publication by R.A. Fisher (1918) of a paper titled “The correlation between
relatives under the supposition of Mendelian inheritance.” In it he showed
that if a quantitative trait, such as height or blood pressure that is measured
on a continuous scale, was determined by an environmental effect and
a large number of genes, each of small effect, the trait should follow a
Normal or Gaussian distribution in the population. The model predicts
correlations between relatives of different degrees. Before this publication
there was widespread doubt about Darwin’s theory and conflict between
two traditions—Mendelians, who sought to understand variation using
Mendel’s principles, and the Biometricians such as Galton and Pearson, who
doubted that Mendel’s insights were relevant to quantitative traits. With the
two schools reconciled, biologists stopped throwing stones at each other,
and by the 1930s modern evolutionary theory had essentially taken shape.
Fisher defined the heritability of a quantitative trait as the fraction of the
population variance of the trait was due to genetic variation, in particular
to additive genetic variation, additive variation being variation that is transmitted from parent to offspring. By convention, heritability is written as h2 .
The model led to the “breeder’s equation” giving the predicted response of a
trait to selection as r = h2 s where r is the response to selection, the difference
between the average value in the population before selection and the average value of the offspring of selected parents. This equation is the workhorse
of quantitative genetics as it relates the intensity of selection on a trait to the

Genetics and Social Behavior

9

rate of change in the character that would result. It seems too simple, yet in
general it works very well.
A general finding is that nearly everything one can measure in humans is
partially heritable, from stature and IQ on the high end (∼0.80) to personality
traits from questionnaires on the low (∼0.25) (Turkheimer, 2011). Such estimates are obtained from correlations between relatives, most famously from
the extent of similarity between pairs of identical twins. The implication from
the generally positive findings is that many of the differences among us that
have generally been assumed to be entirely learned, such as how religious
we are or whether we are politically liberal or conservative, reflect in part
gene differences. This suggests that group differences without our species
also reflect in part gene differences, but there is little direct exploration of
that possibility yet in the literature.
Heritability studies in humans have always been subject to doubts and criticisms. For example, similarities between identical twins separated early in
life are a staple of the literature, but there remained doubt about whether their
environments were really different. Further, even if separated at birth they
had shared a uterine environment during gestation and this shared environment might account for similarities. These shadows of doubt, together with
the old idea that gene differences are not the proper domain for the social sciences, have kept the impact of heritability studies low, or at least lower than
they ought to be.
Most of these doubts and reservations are now resolved with the use of SNP
(single nucleotide polymorphism) chips that provide for any individual his
or her genotype at several hundred thousand or more genetic loci. These loci
are not necessarily or even usually functionally significant, but these chips
allow us to measure overall genetic similarity, called kinship, between any
two individuals. “Kinship” in common usage means degree of relationship
computed from a pedigree, while in genetics, kinship often means simply
overall genetic similarity. The meanings are equivalent in practice, but kinship from chips is more accurate. For example, from a pedigree my kinship
with my child is 1/4 or 0.25, but in fact my kinship with my child might be
as low as 0.2 or so or as high as 0.30 or more. (Social scientists may be more
familiar with what is called the coefficient of relationship rather than kinship:
roughly relationship is twice kinship.)
In a series of papers, Peter Visscher and colleagues have computed kinship
between all pairs of individuals from large longitudinal studies and looked
at distant kinship and interesting quantitative traits such as stature (Yang
et al., 2010), and intelligence (Davies et al., 2011) and schizophrenia and
bipolar disorder (Purcell et al., 2009). In order to avoid the possibility of
shared environments any pair more closely related than third cousins were
not used. They showed that the new method using computed genetic

10

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

similarity yielded that same heritability estimates that had been known for
years, essentially falsifying the doubts and quibbles that had plagued human
quantitative genetics for decades. These papers also showed quite clearly
that Fisher’s original model of 1918 was in fact the right description of the
inheritance mechanism: for both stature and intelligence our differences are
caused by the small effects of a very large number of genes scattered over
the genome, ending hopefully the futile search for “genes for” one thing and
another. The details are somewhat technical: a lucid explanation is given by
Visscher (2010) in a commentary on his earlier paper.
Epigenetic transmission has received much attention in the past several
years. In the simplest case, some environmental effect causes DNA to be
modified so that a gene is turned off or turned on, and this modified gene
is transmitted to later generations. The modification is not permanent and
the modification can be reversed (reset) in the future. We have so far no clear
picture of its importance in human social behavior. The new SNP technology will certainly be an important tool in following this up, as a correlation
between, say, parent and child would be inflated by shared epigenetics while
correlations between more distant relatives would not. The SNP approach
informs about correlations between, for example, sixth cousins, who have
no idea of their relationship and who would not share epigenetic effects as
parent and child or siblings would (Tal, Kisdi, & Jablonka, 2010).
An important category of model is the threshold model in which there is
an underlying Normal distribution of some unobserved trait. Individuals
with more than the threshold value of the underlying trait exhibit the phenotype under study. In an important series of papers in the 1970s Robert
Cloninger and others proposed that sociopathy, defined by rigid explicit criteria, followed such a model in which the thresholds were different in males
and females (Cloninger, Reich, & Guze, 1975a, 1975b; Reich, Cloninger, &
Guze, 1975): the threshold for sociopathic behavior was higher in females.
It required a higher concentration of the underlying trait to cause outright
sociopathy in women.
Threshold models provide useful insight even without assuming anything
about a transmission mechanism. For example, Eisner (2001) discusses the
decline in homicide victimization rates in Europe from the Middle Ages to
the present. A typical comparison in his data is from 50 homicides per 100,000
population in the year 1200 to 2 per 100,000 today, a 25-fold decline. This
seems to be a big decline, but is it really? How does it compare to a decline
from 100 to 4 per 100,000?
A model, perhaps a poor model but better than no model initially, is
to imagine that homicide is a reflection of some underlying normally
distributed genetic trait that has varied over time, perhaps because of
selection against violence. A rate of 50/100,000 implies that this fraction of

Genetics and Social Behavior

11

the population, 0.05%, is homicidal. If the underlying trait has a standard
normal distribution, 0.05% corresponds to the fraction of the population
that is >3.3 standard deviations from the mean. A decline to 2/100,000,
0.002%, corresponds to 4.1 standard deviations from the mean. Assuming
that selection has simply shifted the underlying distribution of violence
proneness to the left without changing the variance, which is a consequence
of Fisher’s model and often seen in experiments with animals, the amount
of shift is 4.1–3.3 or 0.8 standard deviations. In more familiar terms this 0.8
standard deviations would correspond to about 2 inches in height or 12 IQ
points, not a very impressive change at all over 900 years.
What if were studying another social indicator, 100 times as common, that
changed from in the same time from 50 to 2 per 1000 rather than 100,000. With
the same assumptions about an unchanging variance the implied change in
the underlying distribution is 1.3 standard deviations over 900 years, nearly
double the change in proneness to homicide, corresponding to a change in
height of 3.5 inches or a change in IQ of nearly 20 points. If this were due
to selection, the implied selection intensity against the latter trait is much
greater.
None of this is to claim that homicide or drunk driving or anything else is
“genetic.” If we had more information we could investigate the heritability of
the trait(s) and build more accurate and more useful models. Even so, thinking in terms of a threshold and an underlying Normal distribution provides
an important and useful way to understand changes in discrete traits over
time.
ISSUES FOR FUTURE RESEARCH
Genetics and genetic approaches will become more and more important in
social science research and understanding. Specific loci such as MAOA that
we discussed earlier, and numerous others, are of limited interest as few of
these loci are very significant in determining everyday human differences.
This means that the practice of looking for “genes for” one thing and another
will lose salience in social science. While the occasional locus may remain
salient and important, it is increasingly apparent that most of our diversity
reflects our small differences at large and very larger numbers of genetic
loci, the effect of each locus so small as to be essentially undetectable. While
the old paradigm of looking for “genes for” one thing and another will
persist it will surely fade in the face of the even older quantitative genetic
models as they come back for their place in the social sciences. The recent
use of SNP chips has essentially resolved old doubts and controversies
about biases and inaccuracies in heritability estimates. The standard model

12

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

of genetic influence being the combined small effects of a large number of
genes scattered throughout the genome, fits most data quite well.

REFERENCES
Alia-Klein, N., Goldstein, R. Z., Kriplani, A., Logan, J., Tomasi, D., Williams, B., … ,
Fowler, J. S. (2008). Brain monoamine oxidase-A activity predicts trait aggression.
Journal of Neuroscience 28, 5099–5104.
Awadalla, P., Gauthier, J., Myers, R. A., Casals, F., Hamdan, F. F., Griffing, A. R., …
Tarabeux, J. (2010). Direct measure of the de novo mutation rate in autism and
schizophrenia cohorts. The American Journal of Human Genetics 87, 316–324.
Barondes, S. H. (1990). The biological approach to psychiatry: History and prospects.
The Journal of Neuroscience, 10, 1707–1710.
Bassett, A. S., Bury, A., Hodgkinson, K. A., & Honer, W. G. (1996). Reproductive
fitness in familial schizophrenia. Schizophrenia Research, 21, 151–160.
Brunner, H. G., Nelen, M., Breakefield, X. O., Ropers, H. H., & Van Oost, B. A. (1993).
Abnormal behavior associated with a point mutation in the structural gene for
monoamine oxidase A. Science, 262(5133), 578–580.
Cases, O., Seif, I., Grimsby, J., Gaspar, P., Chen, K., Pournin, S., … Shih, J. C. (1995).
Aggressive behavior and altered amounts of brain serotonin and norepinephrine
in mice lacking MAOA. Science 268, 1763.
Cloninger, C. R., Reich, T., & Guze, S. B. (1975a). The multifactorial model of disease
transmission: III. Familial relationship between sociopathy and hysteria (Briquet’s
syndrome). British Journal of Psychiatry, 127, 23–32.
Cloninger, R. C., Reich, T., & Guze, S. B. (1975b). The multifactorial model of disease transmission: II. Sex differences in the familial transmission of sociopathy
(antisocial personality). British Journal of Psychiatry, 127, 11–22.
Cross-Disorder Group of the Psychiatric Genomes Consortium (2013). Identification
of risk loci with shared effects on five major psychiatric disorders: A genome-wide
analysis. The Lancet, 381(9875), 1371–1379.
Davies, G., Tenesa, A., Payton, A., Yang, J., Harris, S. E., Liewald, D., … Deary, I.
J. (2011). Genome-wide association studies establish that human intelligence is
highly heritable and polygenic. Molecular Psychiatry, 16(10), 996–1005.
Eisner, M. (2001). Modernization, self-control and lethal violence. The long-term
dynamics of European homicide rates in theoretical perspective. British Journal of
Criminology, 41(4), 618–638.
Fisher, R. A. (1918). The correlation between relatives on the supposition of
Mendelian inheritance. Transactions of the Royal Society of Edinburgh, 52(2), 399–433.
Girard, S. L., Gauthier, J., Noreau, A., Xiong, L., Zhou, S., Jouan, L., … Diallo, O.
(2011). Increased exonic de novo mutation rate in individuals with schizophrenia.
Nature Genetics 43, 860–863.
Kahn, H., Aligic˘a, P. D., & Weinstein, K. R. (2009). The essential Herman Kahn: In
Defense of thinking. Lanham, MD: Lexington Books.

Genetics and Social Behavior

13

Kendler, M. D., & Greenspan, P. D. (2006). The nature of genetic influences on
behavior: Lessons from “Simpler” organisms. American Journal of Psychiatry, 163,
1683–1694.
Kondrashov, A. (2012). Genetics: The rate of human mutation. Nature, 488, 467–468.
Mayr, E. (1963). Animal species and evolution. Cambridge, England: Harvard University Press.
Morawski, J. G. (1986). Organizing knowledge and behavior at Yale’s Institute of
human relations. Isis, 77(2), 219–242.
Purcell, S. M., Wray, N. R., Stone, J. L., Visscher, P. M., O’Donovan, M. C., Sullivan,
P. F., … Morris, D. W. (2009). Common polygenic variation contributes to risk of
schizophrenia and bipolar disorder. Nature, 460(7256), 748–752.
Reich, R., Cloninger, C. R., & Guze, S. B. (1975). The multifactorial model of disease
transmission: I. Description of the model and its use in psychiatry. British Journal
of Psychiatry, 127, 1–10.
Saha, S., Chant, D., Welham, J., & McGrath, J. (2005). A systematic review of the
prevalence of schizophrenia. PLoS Medicine, 2, e141.
Scott, A. L., Bortolato, M., Chen, K., & Shih, J. C. (2008). Novel monoamine oxidase A
knock out mice with human-like spontaneous mutation. Neuroreport, 19, 739–743.
Sinervo, B., & Lively, C. M. (1996). The rock-paper-scissors game and the evolution
of alternative male strategies. Nature, 380, 240–243.
Smith, J. M., & Price, G. R. (1973). The logic of animal conflict. Nature, 246, 15.
St Clair, D., Xu, M., Wang, P., Yu, Y., Fang, Y., Zhang, F., … Sham, P. (2005). Rates
of adult schizophrenia following prenatal exposure to the Chinese famine of
1959–1961. JAMA, the Journal of the American Medical Association 294, 557–562.
Susser, E. S., & Lin, S. P. (1992). Schizophrenia after prenatal exposure to the Dutch
Hunger Winter of 1944–1945. Archives of General Psychiatry, 49, 983.
Tal, O., Kisdi, E., & Jablonka, E. (2010). Epigenetic contribution to covariance between
relatives. Genetics, 184(4), 1037–1050.
Tooby, J., & Cosmides, L. (1990). On the universality of human nature and the uniqueness of the individual: The role of genetics and adaptation. Journal of Personality,
58, 17–67.
Tsuang, M. (2000). Schizophrenia: Genes and environment. Biological Psychiatry, 47,
210–220.
Turkheimer, E. (2011). Still missing. Research in Human Development, 8(3–4), 227–241.
Visscher, P. M. (2010). A commentary on ‘common SNPs explain a large proportion
of the heritability for human height’ by Yang et al. (2010). Twin Research and Human
Genetics, 13(6), 517.
Wang, J., Wurm, Y., Nipitwattanaphon, M., Riba-Grognuz, O., Huang, Y.-C., Shoemaker, D., & Keller, L. (2013). A Y-like social chromosome causes alternative
colony organization in fire ants. Nature, 493, 664–668.
Xu, B., Roos, J. L., Dexheimer, P., Boone, B., Plummer, B., Levy, S., ... Karayiorgou, M. (2011). Exome sequencing supports a de novo mutational paradigm for
schizophrenia. Nature Genetics, 43, 864–868.

14

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Yang, J., Benyamin, B., McEvoy, B. P., Gordon, S., Henders, A. K., Nyholt, D. R., …
Visscher, P. M. (2010). Common SNPs explain a large proportion of the heritability
for human height. Nature Genetics, 42(7), 565–569.

FURTHER READING
Reich, R., Cloninger, C. R., & Guze, S. B. (1975). The multifactorial model of disease
transmission: I. Description of the model and its use in psychiatry. British Journal
of Psychiatry, 127, 1–10.
Turkheimer, E. (2011). Still missing. Research in Human Development, 8(3–4), 227–241.
Visscher, P. M. (2010). A commentary on ‘common SNPs explain a large proportion
of the heritability for human height’ by Yang et al. (2010). Twin Research and Human
Genetics, 13(6), 517.

HENRY HARPENDING SHORT BIOGRAPHY
Henry Harpending is Professor of Anthropology at the University of Utah.
His academic webpage is http://harpending.humanevo.utah.edu. He has
spent several long fieldtrips in the northern Kalahari studying genetics and
family organization of foraging groups there as well as ranchers. He has also
had a parallel career working on human biological and cultural diversity,
human genetics, and modern human origins. Cochran and Harpending published The Ten Thousand Year Explosion (Basic Books) in 2010.
GREGORY COCHRAN SHORT BIOGRAPHY
Gregory Cochran is a physicist and an adjunct Professor of Anthropology
at the University of Utah. His early career was in optical physics working
in our aerospace industry. A decade or so ago, his interests turned to theoretical biology and anthropology. He has written about an infectious model
of human male homosexuality, immune system diversity, and immune system evolution among human continental populations, and acceleration of the
rate of evolution in humans, especially since the end of the ice ages and the
origins of agriculture.
RELATED ESSAYS
Telomeres (Psychology), Nancy Adler and Aoife O’Donovan
Social Epigenetics: Incorporating Epigenetic Effects as Social Cause and
Consequence (Sociology), Douglas L. Anderton and Kathleen F. Arcaro
Kin-Directed Behavior in Primates (Anthropology), Carol M. Berman
The Sexual Division of Labor (Anthropology), Rebecca Bliege Bird and Brian
F. Codding

Genetics and Social Behavior

15

Genetics and the Life Course (Sociology), Evan Charney
Genetic and Environmental Approaches to Political Science (Political Science),
Zoltán Fazekas and Peter K. Hatemi
Food Sharing (Anthropology), Michael Gurven and Adrian V. Jaeggi
An Evolutionary Perspective on Developmental Plasticity (Psychology),
Sarah Hartman and Jay Belsky
Grandmothers and the Evolution of Human Sociality (Anthropology), Kristen
Hawkes and James Coxworth
The Neurobiology and Physiology of Emotions: A Developmental Perspective (Psychology), Sarah S. Kahle and Paul D. Hastings
Herd Behavior (Psychology), Tatsuya Kameda and Reid Hastie
Genetic Foundations of Attitude Formation (Political Science), Christian
Kandler et al.
Cultural Neuroscience: Connecting Culture, Brain, and Genes (Psychology),
Shinobu Kitayama and Sarah Huff
Reconciliation and Peace-Making: Insights from Studies on Nonhuman
Animals (Anthropology), Sonja E. Koski
Cooperative Breeding and Human Evolution (Anthropology), Karen L.
Kramer
Niche Construction: Implications for Human Sciences (Anthropology), Kevin
N. Laland and Michael O’Brien
Culture, Diffusion, and Networks in Social Animals (Anthropology), Janet
Mann and Lisa Singh
Evolutionary Perspectives on Animal and Human Personality (Anthropology), Joseph H. Manson and Lynn A. Fairbanks
Neural and Cognitive Plasticity (Psychology), Eduardo Mercado III
Evolutionary Theory and Political Behavior (Political Science), Michael Bang
Petersen and Lene Aarøe
Gestural Communication in Nonhuman Species (Anthropology), Simone Pika
Darwinism as a Decryption Key for the Human Mind (Psychology), Csaba
Pléh and Ottilia Boross
DNA Revolution and the Social and Behavioral Sciences (Psychology), Maciej
Trzaskowski and Robert Plomin
Born This Way: Thinking Sociologically about Essentialism (Sociology),
Kristen Schilt
Vocal Communication in Primates (Anthropology), Katie E. Slocombe
Primate Allomaternal Care (Anthropology), Stacey Tecot and Andrea Baden
How Form Constrains Function in the Human Brain (Psychology), Timothy
D. Verstynen
A Gene-Environment Approach to Understanding Youth Antisocial
Behavior (Psychology), Rebecca Waller et al.


Genetics and Social Behavior
HENRY HARPENDING and GREGORY COCHRAN

Abstract
We focus on the effects of gene differences on social and behavioral differences
among individuals and among larger groups of individuals. Many specific genetic
markers are known that influence aspects of personality and behavior. The focus
on single genes and groups of genes is giving way to quantitative genetics, the
statistical study of transmission of characteristics viewed as the outcome of the
effects of very large numbers of genes. While traditional social science largely
ignores the effects of genetically transmitted influences, the subject persists and
grows in importance. Classical quantitative genetic methods may give much insight
into human behavioral diversity and they provide the “right” way to measure
and assess variation in rates of threshold traits. We discuss examples, trends, and
possibilities for the incorporation of genetic data and models in the social and
behavioral sciences without advocating major changes in practice.

INTRODUCTION
There are two distinct ways we could approach the task of reviewing genetics
and human social behavior. One is to review aspects that are genetic in the
sense of hard-wired in the human genome and universal. For example, we
all learned spoken language early in our lives and most agree that the trait
“learn to speak the language(s) in your social environment” is such a universal, while the specific language(s) we learned were entirely dependent on our
rearing environment. In spite of the existence of people who never learned to
speak, we do not hesitate to regard learning speech as “genetic.” However,
things are not so simple: speech does not have to be taught while literacy
does. We are all literate, but many of us struggled to learn, say, calculus, in a
way we never struggled with literacy. This would lead to a consideration of
the nature of cognitive modules, a lively area of current research.
Another way to proceed is to consider how genetic differences among people or among groups of people lead to or have been generated by individual
or group differences in social behavior. Are individual differences in social
behavior determined by gene differences? Do group level gene differences
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.

1

2

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

modulate group differences in social behavior? A century ago, the response
of most educated people would have been “of course” to questions such as
these, whereas today, the response of many educated people and scientists is
“of course not.” We suggest that this change has gone too far, motivated in
many cases by political undercurrents, and that an important task for social
science it to understand and investigate gene differences.
In the past century, many anthropologists were interested in temperament,
for example, and culture and personality was a prominent subfield. It is
striking that in all that literature, the possibility of such things reflecting gene
differences was never on the table. Roughly between 1900 and 1950, there
had emerged a new cluster of academic disciplines, social science. The vision
was that this would be a third kind of knowledge, somewhere between the
humanities and the sciences, a branch that would stand between and perhaps unite human knowledge. The optimism and ambition of practitioners
of the new disciplines were great and they were contagious so, for example,
Yale University created the Institute of Human Relations in the 1930,
“directly concerned with the problems of man’s individual and group conduct” (Morawski, 1986). Harvard University created what was practically a
whole new division of the university with its Department of Social Relations
in 1946, a hybrid of sociology, anthropology, and psychology.
A simple, indeed simplistic, statement of the origin of social science in its
early years was that it insisted on the separation of the body and soul, the soul
being the locus of social behavior and personality. If social science was to be
the scientific study of the human soul, it was necessary that there must be a
universal human soul separate from the human body. The result was a denial
of human biological diversity’s relevance to social behavior and, in its strong
form in Anthropology due to Boas and his followers, a denial of ongoing
evolution in humans of that soul. There was in universities something like a
truce between the biological sciences and the social sciences so that even the
great biologist Ernst Mayr could say that “We cannot escape the conclusion
that man’s evolution toward manness suddenly came to a halt” (Mayr, 1963).
FOUNDATIONAL RESEARCH
There have always been cracks in the strong social science stance but in the
past decade or so it has been noticeably weakening. A prominent example
has been psychiatry converting essentially to biological psychiatry. Advances
in biochemistry such as radioimmunoassay, allowing assays of hormones
and other chemicals at reasonable cost, led to a lively literature on the effects
of hormones and other chemicals on behavior.
Much of the study of genetic influences on behavior and personality relied
for decades on the similarity between genetic relatives, especially identical

Genetics and Social Behavior

3

and fraternal twins. Twins separated at birth were of special interest because
it was presumed that they did not share a common environment and their
resemblance reflected shared genes. A not unreasonable criticism of such
studies was that (i) they shared a common prenatal uterine environment and
(ii) foster families were not random families but were often chosen to match
the adoptees in terms of race, social class, and so on.
In the past several decades, new technology from genomics has allowed
genotyping of hundreds of thousands to millions of genetic markers in
humans, directly measuring similarity (i.e., kinship) between individuals.
This means, for example, we can look at similarities between third cousins
and compare those similarities to those between fifth cousins, cryptic cousins
as the individuals are complete strangers to each other. The possible errors
from looking at close genealogical relatives in quantitative genetic studies
are now bypassed (Purcell et al., 2009; Visscher, 2010), as the new methods
for estimating heritability have replicated the old estimates. An old nagging
source of doubt and controversy in behavior genetic studies is gone.
CUTTING EDGE RESEARCH
SINGLE LOCUS TRAITS
Much of human behavior is universal and adaptive, or used to be in past environments. We eat when hungry, fear spiders and snakes, love our children.
Individuals without these behaviors failed to reproduce, or to reproduce at
rates comparable to those with the behaviors. Such patterns can be thought
of as strategies, patterns or rules of conduct that, on average, led to reproductive success in the human past. However, many aspects of our behavior vary.
The differences show up early in life and do not seem to be affected much by
parental rearing style.
What are the sources of these differences between individuals? Genetic differences must contribute, as personality is heritable, in humans as in other
species (Kendler & Greenspan, 2006). To say that some trait is heritable is to
say that it is genetically transmissable to offspring. More precisely, heritability is a fraction between 0 and 1 that specifies the amount of trait diversity in
a population that reflects gene differences.
Environmental insults contribute as well, such as neurotropic viruses,
toxins, and other influences. However, here we are going to talk genetics.
Genetic influences on behavior fall into two fundamental categories: those
that are the product of natural selection, and those that are not.
To what extent is existing variation in human behavior adaptive? For
example, some people are scrupulous and hyper-moral while, at the other
extreme, some people are without apparent conscience. (We call them

4

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

sociopaths.) There is apparently a continuum from one extreme to the other,
with most people falling somewhere between the two extremes. Why should
not there be a single best strategy, that is, why is not everyone the same?
Often there is a single optimum, but not always. It depends on
whether the payoff in terms of fitness of a particular course of action
is frequency-dependent—in other words, whether it depends on the actions
of other individuals. Running from a forest fire pays off whether anyone
else does or not, but the payoff of running for tribal chief decreases as the
number of candidates increases. If everyone is passive (a “dove”), aggressive
individuals (hawks) prosper, but as hawks becomes more common, they
increasingly run into and fight other hawks, so their payoff decreases. If the
cost of fighting among hawks is high enough, the equilibrium solution is a
mixed state consisting of both hawks and doves (Smith & Price, 1973). Thus
there is no single best solution, no single optimal behavior. This is often the
case with social interactions. It is worth remembering that hawks do not
benefit the species as a whole: The species would do better if everyone just
got along. However, when the strategy benefits individual hawks and the
genes they carry, they increase in frequency.
A related issue is whether adaptive variation in behavior and personality
is fixed or flexible, heritable or not. Sometimes an individual can pick one of
several possible life strategies based on exterior clues. If a female bee larva is
fed royal jelly, she becomes a queen; otherwise, she becomes a sterile worker.
The capability to assume those different roles is adaptive and presumably a
result of natural selection, but it is not noticeably heritable, because all female
bees have it. In other cases, like those pesky fire ants that exhibit single-queen
and Los Angeles-style multiqueen colonies, the two morphs are determined
by the two alleles of a single supergene (Wang et al., 2013), and the behavior
difference is completely heritable. In the former case of queens and workers
the two outcomes reflect environment effects with no contribution from gene
differences while in the latter case of colony structure the outcomes reflect
simple gene differences with no contribution from the environment.
As we are arguably a lot smarter than ants or bees, one might think that
most adaptive personality variation in humans would be learned (a response
to exterior cues) rather than heritable. Certainly some is, but much variation
looks heritable. People do not seem to learn to be aggressive or shy—they just
are, and in those tendencies resemble their biological parents. There are models that may explain this. One is that jacks of all trades are masters of none:
If you play the same role all the time, you will be better at it than someone
who keeps switching personalities. It could be the case that such switching is
physiologically difficult and/or expensive. Moreover, in at least some cases,
being predictable has social value. Someone who is known to be implacably aggressive will always win at chicken (Kahn, Aligică, & Weinstein, 2009).

Genetics and Social Behavior

5

Being known as the sort of guy who would rush into a burning building to
save ugly strangers may pay off, even though actually running into that blaze
does not.
In addition, if a particular role or personality type only became viable relatively recently—say, 10,000 years ago, in the early Neolthic—a mutation that
induces that personality may have become fairly common, but would not
yet be part of a precise and flexible system. The required modifier genes that
would turn those tendencies on when they pay and off when they do not,
would take longer to evolve. Evolution is ongoing, so many new adaptations
are likely to be imperfect and incomplete.
John Tooby and Leda Cosmides, leading founders of evolutionary psychology, have argued strongly against the possibility of heritable, adaptive
behavioral variation in humans (Tooby & Cosmides, 1990). Their arguments
imply that no such thing should exist in any species, but we have found
genetic morphs in lizards, birds, crustaceans, ants, and butterflies. They
are widespread in nature. However, one of their arguments is especially
interesting: they argue that adaptations, including behavioral adaptations,
are usually generated by complex, coadapted sets of genes (true), and that
such gene complexes would be broken up by sexual reproduction (also
true): Even if the dad had a set of genes that made him a natural blacksmith
or tap dancer, his kids would never inherit that whole set and the talent
should disappear.
They are on to something with this argument, but nature seems to have
found solutions. In some cases, a set of genes that determine the alternative
phenotype are arranged as a supergene, a group of genes that are physically
close and strongly linked. That keeps the gene complex from being broken
up. In other cases, the alternate forms are determined by a single gene that
acts as a switch. Different alleles of that gene specify different morphs. This
kind of mechanism is also largely unaffected by recombination.
Perhaps the heritable variant is in some way a simple trait. For example,
what if simply losing a particular complex adaptation was adaptive when
rare? It is easy to see that a complex behavioral adaptation could be stopped
cold by a single mutation. Alternatively, for that matter, what if greatly intensifying a particular behavior or drive—turning up the volume knob—was
adaptive when rare? It is possible to imagine a simple mutation that turns up
the volume in some way? Surely. In addition, of course, such simple initial
changes can be gradually refined by natural selection.
We know of many clear examples of distinct behavioral morphs or strategies in other species. We mentioned the two different kinds of fire-ant society.
Many of the examples are of male morphs with different reproductive strategies. One of the most interesting and well-known example is Uta stansburiana,
a species of lizard studied by Barry Sinervo (Sinervo & Lively, 1996). The

6

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

species has three male morphs with distinct color patterns: orange, yellow,
and blue. Orange males are large, aggressive, and control a territory with
several females. They dominate blue males, but yellow males, which mimic
females, often cuckold them. Blue males have a smaller territory with one
female, which they can effectively defend against yellow males. Orange beats
blue, blue beats yellow, and yellow beats orange. It is scissors-paper-rock.
Within-group adaptive behavioral variation is possible, as we know of
many examples in other species. Behavioral differences between geographically separated populations of the same species are also possible, as selection
pressures often vary with location. That means that genetically induced
adaptive behavioral variation could exist in humans—but does it?
The usual measures of personality all show significant heritability
(Turkheimer, 2011), which is consistent with adaptive personality variation
but also with nonadaptive genetic influences, such as mutational load. For
example, aggressive individuals could be present in a population as evolved
players of a “hawk” strategy in their interpersonal relations, or they could
simply be unfortunate bearers of deleterious genetic mutations leading to
their aggressive behavior. The case that would be clearest, easiest to prove,
would be one in which different variants of a single gene have a significant
influence on behavior. If such variants existed, they would almost certainly
be the product of adaptive evolution. Mutational pressure would create a
very different picture, with rare deleterious mutations of many different
genes, rather than two or three common variants of a single locus.
The most promising possible case of adaptive genetic variation in humans
is in MAOA, an enzyme that degrades several neutrotransmitters. The gene,
located on the X chromosome, contains a 30-base sequence that is repeated
different numbers of times in different MAOA variants—2R, 3R, 3.5R, 4R,
and 5R. These repeats are in a regulatory region and affect gene activity.
We know that complete loss of MAOA activity has a strong impact on
human behavior. A null mutation of MAOA has been seen in large Dutch
kindred (Brunner, Nelen, Breakefield, Ropers, & Van Oost, 1993). Males with
this mutation show impulsive aggressiveness and mild mental retardation.
Low levels of the gene products (rather than any particular variant) are
associated with trait aggression in men (Alia-Klein et al., 2008).
In mouse models, inactivation of MAOA causes increased aggressiveness
(Cases et al., 1995; Scott, Bortolato, Chen, & Shih, 2008). In rhesus monkeys,
there is an association between low activity MAOA alleles that is dependent
on early environment. There seems to a similar interaction effects in humans:
low activity MAOA alleles (the 3R and 2R variants) seem to increase vulnerability to environmental stresses such as abuse in childhood. In this case, the
environmental effect of being abused as a child, like the jelly to the female bee
larva, leads to behavior change in adulthood only if a specific genetic variant

Genetics and Social Behavior

7

is present. With this level of complexity, spreading of a trait as learned or not
loses its meaning.
Despite the special difficulties of performing these investigations in
humans, it appears there really are genetically induced behavioral variants
that are the product of selection. MAOA is not the only locus for which we
have evidence of this, but it is probably the best-understood, and makes the
point.
There are important aspects of human behavioral variation that are almost
certainly maladaptive, such as mental retardation and mental illness. They
too are heritable, and to a significant extent genetic in origin. That is not to
say that environmental insults and pathogens do not play a role. Prenatal
starvation doubled the incidence of schizophrenia in the cohorts affected by
the Dutch famine of 1944 and the Chinese famine associated with the Great
Leap Forward (St Clair et al., 2005; Susser & Lin, 1992). Neurosyphilis once
accounted for about half of the patients in psychiatric hospitals (Barondes,
1990). However, we have reduced the impact of syphilis and other pathogens
over the past century, as well as famine. Genetic causes have not diminished,
and therefore account for a larger fraction of cases than they once did.
Personality variations generally do not drastically reduce reproductive
fitness, but mental illness and mental retardation do. For example, in
schizophrenia, fitness is drastically lowered in men (about 25% of normal)
and significantly lowered in women as well, about 50% of normal (Bassett,
Bury, Hodgkinson, & Honer, 1996). As schizophrenia is quite heritable
(Tsuang, 2000) there is an apparent paradox: how can schizophrenia (and
other forms of mental illness) persist over time at moderately high frequency,
0.5–1%? (Saha, Chant, Welham, & McGrath, 2005).
Increasingly, it looks as if the cause is ongoing mutation. Our understanding of mutation in humans and its negative consequences for human health is
developing rapidly, because of advances in sequencing technology. It is now
possible to identify mutations by high accuracy whole-genome sequencing
of family triads: if you see a sequence in the child that does not exist in his
parents, it is a new mutation. (Of course it could also be parental assignment
error, but that is easily detected because there would be thousands of mismatches between the parental genes and those in the child.)
There are a number of recent reports that have found evidence for de novo
mutations as an important cause of schizophrenia and autism (Awadalla
et al., 2010; Girard et al., 2011). A high rate of ongoing mutations also explains
why a significant fraction of schizophrenia cases are sporadic: they appear
in families that do not have previous cases of schizophrenia (Xu et al., 2011).
These techniques have also confirmed that most mutations originate in
the father, and that the mutation rate increases linearly with post-pubertal
father’s age.

8

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Many problems are more common in the children of older fathers, but disorders that result from impaired brain function, such as autism, schizophrenia,
and reduced intelligence, are particularly common (Kondrashov, 2012). This
seems to be a consequence of the great complexity of brain development and
function, making use of the majority of all genes.
Schizophrenia must be made up of many different mutational disorders.
At the same time, a given mutation often seems to increase risk for several different mental disorders. (Cross-Disorder Group of the Psychiatric
Genomes Consortium, 2013). Clearly, progress in this area has been hampered by simultaneously lumping together causally unrelated syndromes
and splitting a syndrome with a single cause into several different categories.
On the other hand, if this new picture is correct, it may suggest new and
effective approaches for fighting mental illness. In the short run, finding ways
to reduce the mutation rate might pay off, especially if any factors other than
paternal age are found to materially influence that rate. In the longer run,
gene therapy may offer hope.
QUANTITATIVE TRAITS
Many regard the origin of evolutionary biology as a science to be the
publication by R.A. Fisher (1918) of a paper titled “The correlation between
relatives under the supposition of Mendelian inheritance.” In it he showed
that if a quantitative trait, such as height or blood pressure that is measured
on a continuous scale, was determined by an environmental effect and
a large number of genes, each of small effect, the trait should follow a
Normal or Gaussian distribution in the population. The model predicts
correlations between relatives of different degrees. Before this publication
there was widespread doubt about Darwin’s theory and conflict between
two traditions—Mendelians, who sought to understand variation using
Mendel’s principles, and the Biometricians such as Galton and Pearson, who
doubted that Mendel’s insights were relevant to quantitative traits. With the
two schools reconciled, biologists stopped throwing stones at each other,
and by the 1930s modern evolutionary theory had essentially taken shape.
Fisher defined the heritability of a quantitative trait as the fraction of the
population variance of the trait was due to genetic variation, in particular
to additive genetic variation, additive variation being variation that is transmitted from parent to offspring. By convention, heritability is written as h2 .
The model led to the “breeder’s equation” giving the predicted response of a
trait to selection as r = h2 s where r is the response to selection, the difference
between the average value in the population before selection and the average value of the offspring of selected parents. This equation is the workhorse
of quantitative genetics as it relates the intensity of selection on a trait to the

Genetics and Social Behavior

9

rate of change in the character that would result. It seems too simple, yet in
general it works very well.
A general finding is that nearly everything one can measure in humans is
partially heritable, from stature and IQ on the high end (∼0.80) to personality
traits from questionnaires on the low (∼0.25) (Turkheimer, 2011). Such estimates are obtained from correlations between relatives, most famously from
the extent of similarity between pairs of identical twins. The implication from
the generally positive findings is that many of the differences among us that
have generally been assumed to be entirely learned, such as how religious
we are or whether we are politically liberal or conservative, reflect in part
gene differences. This suggests that group differences without our species
also reflect in part gene differences, but there is little direct exploration of
that possibility yet in the literature.
Heritability studies in humans have always been subject to doubts and criticisms. For example, similarities between identical twins separated early in
life are a staple of the literature, but there remained doubt about whether their
environments were really different. Further, even if separated at birth they
had shared a uterine environment during gestation and this shared environment might account for similarities. These shadows of doubt, together with
the old idea that gene differences are not the proper domain for the social sciences, have kept the impact of heritability studies low, or at least lower than
they ought to be.
Most of these doubts and reservations are now resolved with the use of SNP
(single nucleotide polymorphism) chips that provide for any individual his
or her genotype at several hundred thousand or more genetic loci. These loci
are not necessarily or even usually functionally significant, but these chips
allow us to measure overall genetic similarity, called kinship, between any
two individuals. “Kinship” in common usage means degree of relationship
computed from a pedigree, while in genetics, kinship often means simply
overall genetic similarity. The meanings are equivalent in practice, but kinship from chips is more accurate. For example, from a pedigree my kinship
with my child is 1/4 or 0.25, but in fact my kinship with my child might be
as low as 0.2 or so or as high as 0.30 or more. (Social scientists may be more
familiar with what is called the coefficient of relationship rather than kinship:
roughly relationship is twice kinship.)
In a series of papers, Peter Visscher and colleagues have computed kinship
between all pairs of individuals from large longitudinal studies and looked
at distant kinship and interesting quantitative traits such as stature (Yang
et al., 2010), and intelligence (Davies et al., 2011) and schizophrenia and
bipolar disorder (Purcell et al., 2009). In order to avoid the possibility of
shared environments any pair more closely related than third cousins were
not used. They showed that the new method using computed genetic

10

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

similarity yielded that same heritability estimates that had been known for
years, essentially falsifying the doubts and quibbles that had plagued human
quantitative genetics for decades. These papers also showed quite clearly
that Fisher’s original model of 1918 was in fact the right description of the
inheritance mechanism: for both stature and intelligence our differences are
caused by the small effects of a very large number of genes scattered over
the genome, ending hopefully the futile search for “genes for” one thing and
another. The details are somewhat technical: a lucid explanation is given by
Visscher (2010) in a commentary on his earlier paper.
Epigenetic transmission has received much attention in the past several
years. In the simplest case, some environmental effect causes DNA to be
modified so that a gene is turned off or turned on, and this modified gene
is transmitted to later generations. The modification is not permanent and
the modification can be reversed (reset) in the future. We have so far no clear
picture of its importance in human social behavior. The new SNP technology will certainly be an important tool in following this up, as a correlation
between, say, parent and child would be inflated by shared epigenetics while
correlations between more distant relatives would not. The SNP approach
informs about correlations between, for example, sixth cousins, who have
no idea of their relationship and who would not share epigenetic effects as
parent and child or siblings would (Tal, Kisdi, & Jablonka, 2010).
An important category of model is the threshold model in which there is
an underlying Normal distribution of some unobserved trait. Individuals
with more than the threshold value of the underlying trait exhibit the phenotype under study. In an important series of papers in the 1970s Robert
Cloninger and others proposed that sociopathy, defined by rigid explicit criteria, followed such a model in which the thresholds were different in males
and females (Cloninger, Reich, & Guze, 1975a, 1975b; Reich, Cloninger, &
Guze, 1975): the threshold for sociopathic behavior was higher in females.
It required a higher concentration of the underlying trait to cause outright
sociopathy in women.
Threshold models provide useful insight even without assuming anything
about a transmission mechanism. For example, Eisner (2001) discusses the
decline in homicide victimization rates in Europe from the Middle Ages to
the present. A typical comparison in his data is from 50 homicides per 100,000
population in the year 1200 to 2 per 100,000 today, a 25-fold decline. This
seems to be a big decline, but is it really? How does it compare to a decline
from 100 to 4 per 100,000?
A model, perhaps a poor model but better than no model initially, is
to imagine that homicide is a reflection of some underlying normally
distributed genetic trait that has varied over time, perhaps because of
selection against violence. A rate of 50/100,000 implies that this fraction of

Genetics and Social Behavior

11

the population, 0.05%, is homicidal. If the underlying trait has a standard
normal distribution, 0.05% corresponds to the fraction of the population
that is >3.3 standard deviations from the mean. A decline to 2/100,000,
0.002%, corresponds to 4.1 standard deviations from the mean. Assuming
that selection has simply shifted the underlying distribution of violence
proneness to the left without changing the variance, which is a consequence
of Fisher’s model and often seen in experiments with animals, the amount
of shift is 4.1–3.3 or 0.8 standard deviations. In more familiar terms this 0.8
standard deviations would correspond to about 2 inches in height or 12 IQ
points, not a very impressive change at all over 900 years.
What if were studying another social indicator, 100 times as common, that
changed from in the same time from 50 to 2 per 1000 rather than 100,000. With
the same assumptions about an unchanging variance the implied change in
the underlying distribution is 1.3 standard deviations over 900 years, nearly
double the change in proneness to homicide, corresponding to a change in
height of 3.5 inches or a change in IQ of nearly 20 points. If this were due
to selection, the implied selection intensity against the latter trait is much
greater.
None of this is to claim that homicide or drunk driving or anything else is
“genetic.” If we had more information we could investigate the heritability of
the trait(s) and build more accurate and more useful models. Even so, thinking in terms of a threshold and an underlying Normal distribution provides
an important and useful way to understand changes in discrete traits over
time.
ISSUES FOR FUTURE RESEARCH
Genetics and genetic approaches will become more and more important in
social science research and understanding. Specific loci such as MAOA that
we discussed earlier, and numerous others, are of limited interest as few of
these loci are very significant in determining everyday human differences.
This means that the practice of looking for “genes for” one thing and another
will lose salience in social science. While the occasional locus may remain
salient and important, it is increasingly apparent that most of our diversity
reflects our small differences at large and very larger numbers of genetic
loci, the effect of each locus so small as to be essentially undetectable. While
the old paradigm of looking for “genes for” one thing and another will
persist it will surely fade in the face of the even older quantitative genetic
models as they come back for their place in the social sciences. The recent
use of SNP chips has essentially resolved old doubts and controversies
about biases and inaccuracies in heritability estimates. The standard model

12

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

of genetic influence being the combined small effects of a large number of
genes scattered throughout the genome, fits most data quite well.

REFERENCES
Alia-Klein, N., Goldstein, R. Z., Kriplani, A., Logan, J., Tomasi, D., Williams, B., … ,
Fowler, J. S. (2008). Brain monoamine oxidase-A activity predicts trait aggression.
Journal of Neuroscience 28, 5099–5104.
Awadalla, P., Gauthier, J., Myers, R. A., Casals, F., Hamdan, F. F., Griffing, A. R., …
Tarabeux, J. (2010). Direct measure of the de novo mutation rate in autism and
schizophrenia cohorts. The American Journal of Human Genetics 87, 316–324.
Barondes, S. H. (1990). The biological approach to psychiatry: History and prospects.
The Journal of Neuroscience, 10, 1707–1710.
Bassett, A. S., Bury, A., Hodgkinson, K. A., & Honer, W. G. (1996). Reproductive
fitness in familial schizophrenia. Schizophrenia Research, 21, 151–160.
Brunner, H. G., Nelen, M., Breakefield, X. O., Ropers, H. H., & Van Oost, B. A. (1993).
Abnormal behavior associated with a point mutation in the structural gene for
monoamine oxidase A. Science, 262(5133), 578–580.
Cases, O., Seif, I., Grimsby, J., Gaspar, P., Chen, K., Pournin, S., … Shih, J. C. (1995).
Aggressive behavior and altered amounts of brain serotonin and norepinephrine
in mice lacking MAOA. Science 268, 1763.
Cloninger, C. R., Reich, T., & Guze, S. B. (1975a). The multifactorial model of disease
transmission: III. Familial relationship between sociopathy and hysteria (Briquet’s
syndrome). British Journal of Psychiatry, 127, 23–32.
Cloninger, R. C., Reich, T., & Guze, S. B. (1975b). The multifactorial model of disease transmission: II. Sex differences in the familial transmission of sociopathy
(antisocial personality). British Journal of Psychiatry, 127, 11–22.
Cross-Disorder Group of the Psychiatric Genomes Consortium (2013). Identification
of risk loci with shared effects on five major psychiatric disorders: A genome-wide
analysis. The Lancet, 381(9875), 1371–1379.
Davies, G., Tenesa, A., Payton, A., Yang, J., Harris, S. E., Liewald, D., … Deary, I.
J. (2011). Genome-wide association studies establish that human intelligence is
highly heritable and polygenic. Molecular Psychiatry, 16(10), 996–1005.
Eisner, M. (2001). Modernization, self-control and lethal violence. The long-term
dynamics of European homicide rates in theoretical perspective. British Journal of
Criminology, 41(4), 618–638.
Fisher, R. A. (1918). The correlation between relatives on the supposition of
Mendelian inheritance. Transactions of the Royal Society of Edinburgh, 52(2), 399–433.
Girard, S. L., Gauthier, J., Noreau, A., Xiong, L., Zhou, S., Jouan, L., … Diallo, O.
(2011). Increased exonic de novo mutation rate in individuals with schizophrenia.
Nature Genetics 43, 860–863.
Kahn, H., Aligică, P. D., & Weinstein, K. R. (2009). The essential Herman Kahn: In
Defense of thinking. Lanham, MD: Lexington Books.

Genetics and Social Behavior

13

Kendler, M. D., & Greenspan, P. D. (2006). The nature of genetic influences on
behavior: Lessons from “Simpler” organisms. American Journal of Psychiatry, 163,
1683–1694.
Kondrashov, A. (2012). Genetics: The rate of human mutation. Nature, 488, 467–468.
Mayr, E. (1963). Animal species and evolution. Cambridge, England: Harvard University Press.
Morawski, J. G. (1986). Organizing knowledge and behavior at Yale’s Institute of
human relations. Isis, 77(2), 219–242.
Purcell, S. M., Wray, N. R., Stone, J. L., Visscher, P. M., O’Donovan, M. C., Sullivan,
P. F., … Morris, D. W. (2009). Common polygenic variation contributes to risk of
schizophrenia and bipolar disorder. Nature, 460(7256), 748–752.
Reich, R., Cloninger, C. R., & Guze, S. B. (1975). The multifactorial model of disease
transmission: I. Description of the model and its use in psychiatry. British Journal
of Psychiatry, 127, 1–10.
Saha, S., Chant, D., Welham, J., & McGrath, J. (2005). A systematic review of the
prevalence of schizophrenia. PLoS Medicine, 2, e141.
Scott, A. L., Bortolato, M., Chen, K., & Shih, J. C. (2008). Novel monoamine oxidase A
knock out mice with human-like spontaneous mutation. Neuroreport, 19, 739–743.
Sinervo, B., & Lively, C. M. (1996). The rock-paper-scissors game and the evolution
of alternative male strategies. Nature, 380, 240–243.
Smith, J. M., & Price, G. R. (1973). The logic of animal conflict. Nature, 246, 15.
St Clair, D., Xu, M., Wang, P., Yu, Y., Fang, Y., Zhang, F., … Sham, P. (2005). Rates
of adult schizophrenia following prenatal exposure to the Chinese famine of
1959–1961. JAMA, the Journal of the American Medical Association 294, 557–562.
Susser, E. S., & Lin, S. P. (1992). Schizophrenia after prenatal exposure to the Dutch
Hunger Winter of 1944–1945. Archives of General Psychiatry, 49, 983.
Tal, O., Kisdi, E., & Jablonka, E. (2010). Epigenetic contribution to covariance between
relatives. Genetics, 184(4), 1037–1050.
Tooby, J., & Cosmides, L. (1990). On the universality of human nature and the uniqueness of the individual: The role of genetics and adaptation. Journal of Personality,
58, 17–67.
Tsuang, M. (2000). Schizophrenia: Genes and environment. Biological Psychiatry, 47,
210–220.
Turkheimer, E. (2011). Still missing. Research in Human Development, 8(3–4), 227–241.
Visscher, P. M. (2010). A commentary on ‘common SNPs explain a large proportion
of the heritability for human height’ by Yang et al. (2010). Twin Research and Human
Genetics, 13(6), 517.
Wang, J., Wurm, Y., Nipitwattanaphon, M., Riba-Grognuz, O., Huang, Y.-C., Shoemaker, D., & Keller, L. (2013). A Y-like social chromosome causes alternative
colony organization in fire ants. Nature, 493, 664–668.
Xu, B., Roos, J. L., Dexheimer, P., Boone, B., Plummer, B., Levy, S., ... Karayiorgou, M. (2011). Exome sequencing supports a de novo mutational paradigm for
schizophrenia. Nature Genetics, 43, 864–868.

14

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Yang, J., Benyamin, B., McEvoy, B. P., Gordon, S., Henders, A. K., Nyholt, D. R., …
Visscher, P. M. (2010). Common SNPs explain a large proportion of the heritability
for human height. Nature Genetics, 42(7), 565–569.

FURTHER READING
Reich, R., Cloninger, C. R., & Guze, S. B. (1975). The multifactorial model of disease
transmission: I. Description of the model and its use in psychiatry. British Journal
of Psychiatry, 127, 1–10.
Turkheimer, E. (2011). Still missing. Research in Human Development, 8(3–4), 227–241.
Visscher, P. M. (2010). A commentary on ‘common SNPs explain a large proportion
of the heritability for human height’ by Yang et al. (2010). Twin Research and Human
Genetics, 13(6), 517.

HENRY HARPENDING SHORT BIOGRAPHY
Henry Harpending is Professor of Anthropology at the University of Utah.
His academic webpage is http://harpending.humanevo.utah.edu. He has
spent several long fieldtrips in the northern Kalahari studying genetics and
family organization of foraging groups there as well as ranchers. He has also
had a parallel career working on human biological and cultural diversity,
human genetics, and modern human origins. Cochran and Harpending published The Ten Thousand Year Explosion (Basic Books) in 2010.
GREGORY COCHRAN SHORT BIOGRAPHY
Gregory Cochran is a physicist and an adjunct Professor of Anthropology
at the University of Utah. His early career was in optical physics working
in our aerospace industry. A decade or so ago, his interests turned to theoretical biology and anthropology. He has written about an infectious model
of human male homosexuality, immune system diversity, and immune system evolution among human continental populations, and acceleration of the
rate of evolution in humans, especially since the end of the ice ages and the
origins of agriculture.
RELATED ESSAYS
Telomeres (Psychology), Nancy Adler and Aoife O’Donovan
Social Epigenetics: Incorporating Epigenetic Effects as Social Cause and
Consequence (Sociology), Douglas L. Anderton and Kathleen F. Arcaro
Kin-Directed Behavior in Primates (Anthropology), Carol M. Berman
The Sexual Division of Labor (Anthropology), Rebecca Bliege Bird and Brian
F. Codding

Genetics and Social Behavior

15

Genetics and the Life Course (Sociology), Evan Charney
Genetic and Environmental Approaches to Political Science (Political Science),
Zoltán Fazekas and Peter K. Hatemi
Food Sharing (Anthropology), Michael Gurven and Adrian V. Jaeggi
An Evolutionary Perspective on Developmental Plasticity (Psychology),
Sarah Hartman and Jay Belsky
Grandmothers and the Evolution of Human Sociality (Anthropology), Kristen
Hawkes and James Coxworth
The Neurobiology and Physiology of Emotions: A Developmental Perspective (Psychology), Sarah S. Kahle and Paul D. Hastings
Herd Behavior (Psychology), Tatsuya Kameda and Reid Hastie
Genetic Foundations of Attitude Formation (Political Science), Christian
Kandler et al.
Cultural Neuroscience: Connecting Culture, Brain, and Genes (Psychology),
Shinobu Kitayama and Sarah Huff
Reconciliation and Peace-Making: Insights from Studies on Nonhuman
Animals (Anthropology), Sonja E. Koski
Cooperative Breeding and Human Evolution (Anthropology), Karen L.
Kramer
Niche Construction: Implications for Human Sciences (Anthropology), Kevin
N. Laland and Michael O’Brien
Culture, Diffusion, and Networks in Social Animals (Anthropology), Janet
Mann and Lisa Singh
Evolutionary Perspectives on Animal and Human Personality (Anthropology), Joseph H. Manson and Lynn A. Fairbanks
Neural and Cognitive Plasticity (Psychology), Eduardo Mercado III
Evolutionary Theory and Political Behavior (Political Science), Michael Bang
Petersen and Lene Aarøe
Gestural Communication in Nonhuman Species (Anthropology), Simone Pika
Darwinism as a Decryption Key for the Human Mind (Psychology), Csaba
Pléh and Ottilia Boross
DNA Revolution and the Social and Behavioral Sciences (Psychology), Maciej
Trzaskowski and Robert Plomin
Born This Way: Thinking Sociologically about Essentialism (Sociology),
Kristen Schilt
Vocal Communication in Primates (Anthropology), Katie E. Slocombe
Primate Allomaternal Care (Anthropology), Stacey Tecot and Andrea Baden
How Form Constrains Function in the Human Brain (Psychology), Timothy
D. Verstynen
A Gene-Environment Approach to Understanding Youth Antisocial
Behavior (Psychology), Rebecca Waller et al.



Media

Genetics and Social Behavior