The Promise and Peril of Evolutionary Theory for Explaining by vgw19124


									   The Promise and Peril of Evolutionary Theory for Explaining
                       Political Behavior

                                     Eric Oliver
                             Department of Political Science
                                 University of Chicago

                                    January 21, 2009


This paper outlines the strengths and weaknesses of an evolutionary approach to
analyzing political behavior and attitudes. Given the demands on cooperation and social
coordination in human evolutionary history, it is likely that numerous psychological traits
are adaptations to, rather than mere predecessors of, human politics and that these
evolved traits continue to shape modern political cognition. Differentiating between
adapted and learned traits, however, remains a difficult prospect. Using the example of
political values, I discuss the possible insights that an evolutionary approach may offer
and what types of questions need to be answered.

           Paper presented to the Political Theory Workshop, February 9, 2009

“Political psychology is, at the most general level, an application of what is known about
human psychology to the study of politics.”
                                       Sears, Huddy, and Jervis (2003)

        Political psychology is as old as the study of politics itself: political theorists

since Plato have utilized rudimentary psychological precepts, often described as “human

nature,” to explain political phenomena. But over the past half century, the application of

psychology to the study of politics has become formally institutionalized. Drawing from

a wide range of psychological studies, scholars have sought to explain systematically

political dynamics through the lens of the human psyche. They have established

organizations (the International Society of Political Psychology, the Political Psychology

Section of the American Political Science Association), fielded journals (Political

Psychology, Political Behavior), and, most importantly, produced a large body of

informative research. Their efforts have yielded a trove of good explanations for the

acquisition and expression of political attitudes and values (e.g., Feldman 1988, Zaller

1992), the dynamics of political cognition (e.g., Sniderman, Brody, and Tetlock 1993),

the contours of prejudice and inter-group bias (e.g., Kinder and Sanders 1996), and many

other types of political phenomena.

        Yet as a field of inquiry, political psychology remains largely unsettled. It lacks a

unifying theory or set of theories, its practices are often unscientific, its research topics

typically are episodic, and its explanations are sometimes post hoc or atheoretical (Winter

2006). Partly, these are a result of its initial orientation. Psychology itself is an

extremely broad and diverse discipline that employs a disparate range of epistemologies

and ontological perspectives. By basing itself on such a varied discipline, political

psychology inevitably manifests this intellectual fragmentation. For example, political

psychologists explain the behaviors of political leaders with psychodynamic models (e.g.,

Post 2005), the behavior of voters with cognitive models (e.g., Lau and Redlausk 2006),

and the behavior of group members with social psychological models (e.g., Bobo and

Klugel 1993), methods and orientations that have little in common and are limited to

particular domains (i.e., psychodynamic models of leaders may have little utility for

explaining the choices of voters).

       At a more fundamental level, however, political psychology is hampered by its

initial orientation. Using psychological concepts as the starting point for explaining

political phenomena is tantamount to asserting that human psychology exists prior to

politics. In other words, political psychology, as defined above, engenders the

assumption that humans are political only after the psychological processes that underlie

those behaviors are fully developed. Homo psychologicus precedes homo politicus--

humans are only political animals after they are psychological ones.

       Recent developments in the study of social neuroscience and evolutionary theory

call this assumption into question. As psychologists and biologists begin to open the

“black box” of the human mind and understand brain architecture relative to cognitive

and affective processes, it is clear that many psychological phenomena arose in

relationship to particular evolutionary demands. As early hominids began to walk

upright, fashion tools, acquire language, and deepen their intelligence, their adaptability

to a variety of environments dramatically increased. With these increased aptitudes also

came increased political pressure—as hominids became smarter and more versatile, they

also became increasingly capable of (and dependent on) cooperation with each other. In

short, they became increasingly political. For example, homo erectus, a forerunner of

homo sapiens, could migrate from Africa to Europe and Asia largely from their ability to

sustain themselves off the mega-fauna residing on savannas and steppes (Johanson and

Edgar 2006); being highly gracile and weak, early hominid hunting success depended on

large-scale teamwork. Maintaining such cooperation required the development of long-

term and sustainable agreements on the regulation of the nutritional and sexual economy

and on the limitation of intra-group violence.

       Over the past decades, scholars have begun to theorize about how such a

cooperative equilibrium can be maintained. Working largely from deductive and game

theoretic models (such as the prisoner’s dilemma, stag hunt, ultimatum, etc.) as well as

laboratory experiments, this research has demonstrated that not only can “selfless”

cooperation be successful over the long haul (Axelrod 1984) but that, in a variety of

experimental settings, subjects typically exhibit highly cooperative behaviors, such as

rewarding others and punishing defectors at significant costs to themselves (Camerer

2003, Orbell et al. 2004). Other scholars (Trivers 1971, Cosmides and Tooby 1992,

Gigerenzer and Hugg 1992) have extended this reasoning and argued that humans

evolved specific cognitive and affective adaptations for processing social exchange

information. As an example, they point out that human cognition is particularly sensitive

to issues of reciprocity and the detection of cheaters. In other words, just as humans

evolved eyebrows, opposable thumbs, or larynxes to help with specific functions, they

also evolved numerous cognitive and affective processes (i.e., a specific psychology) to

help sustain and enforce cooperation.

       Although this line of inquiry is still in its infancy and already subject to heated

debate (e.g., Fodor 2000, Buller 2005), its implications for the study of politics are

profound. Despite the elaborate cultures and sophisticated political systems of

contemporary life, the social psychological processes that evolved in the Pleistocene

should continue to delimit the ways humans address political problems. In other words,

the basic neurological mechanisms for processing political information, making political

decisions, and acting in the social sphere originate in fundamental biological processes

that were shaped over the millennia that humans and their hominid ancestors lived both in

nomadic, hunter gatherer and, possibly, agricultural societies. Identifying these

biological mechanisms could hold the potential for explaining a wide range of political

behaviors, particularly those that defy expectations of strict rationality or appear universal

across cultures.

       In this paper, I outline some of the strengths and weakness of such an approach to

explaining human political behavior. Using the example of political ideology, I describe

how evolutionary theory might help identify innate types of human values and the

mechanisms of their cultural transmission. The viability of this endeavor, however,

depends largely on the veracity of claims that humans evolved specific cognitive

mechanisms to deal with issues of cooperation and the adjudication of social resources.

Based on deductions about the central challenges to cooperation in early hominan

development, I speculate about three central domains, what I call Food, Sex, and Death,

around which some specific cognitive mechanisms might have arisen. I then discuss the

challenges to testing for these mechanisms relative to our current understandings of social


Evolution and the Study of Political Behavior

         For social scientists, evolutionary theory is both alluring and problematic. It is

alluring because it offers the promise of a unifying and more scientific account of human

economic, political and social behavior. If humans share an essential nature, that nature

is biological in origin and, to use Theodosius Dobzhansky’s oft-repeated phrase, “nothing

in biology makes sense except in light of evolution.” Evolution also offers the advantage

of being a simple, powerful, and parsimonious theory. From an evolutionary perspective,

the mind is viewed as “a set of information-processing machines that were designed by

natural selection to solve adaptive problems faced by our hunter-gatherer ancestors”

(Cosmides and Tooby 1997). In contrast to the “blank slate” model of human

psychology (in which all of the mind’s contents are derived from environmental factors),

evolutionary theory postulates that the human mind is comprised of specific cognitive

processes that were selected in the particular environment of evolutionary adaptation

(EEA) and hold predetermined content and specialized ways of processing information.1

To whatever extent humans have an “essential” human nature, it is located within the

mental architecture that was selected for over millions of years of evolutionary history.

         Evolutionary theory is problematic, however, because reconstructing the

evolution of the human mind is exceedingly difficult. This is particularly the case in

           Many naïve critics assume that by postulating evolutionary influences on human behavior, EP
dismisses “nurture” completely in favor of “nature.” The EP perspective doesn’t embody such false
dichotomies or deny that the human mind is remarkably plastic and influenced by social factors; rather, EP
supposes simply that “hardwired” cognitive structures and affective processing fundamentally constrain the
ways social environments shape human psychology. Consider the analogy of a marathon runner: even
though someone can train herself to run very fast, her top speed ultimately will be delimited by the simple
skeletal and cardio-vascular design of the human body. Similarly, even though people from different
cultures may exhibit seemingly disparate attitudes and behaviors, the cognitive and affective processes that
underlie these actions should be very similar and constrain them in predictable ways.

developing specific and testable hypotheses about psychological traits. For example, in

an ideal world, scientists would be able to link specific genes or alleles to particular

neurological processes and then measure the impact of these neurological processes on

behavioral outcomes. We are, however, still far away from the technological point of

achieving this goal—we are only beginning to decode the human genome and have a very

incomplete understanding of the linkages between genes and their corresponding

phenotypes. Moreover, despite the compelling findings from various neuro-scientific

diagnostic techniques like positron emission tomography (PET) or functional magnetic

resonance imaging (fMRI), our knowledge of brain architecture and cognitive function is

still crude. Nor do we have the capacity either to directly observe most psychological

traits or the evolutionary forces that selected them (Andrews 2007). We can look across

cultures and try to identify “universal” traits or practices (Brown 1991) but these will be

subject to questions of interpretation and comparability; we can try to piece together

inferences from genetic and archaeological records, but here the clues are inconclusive—

in fact, we still do not have a definite genealogy of human origins. Consequently, any

current assertions about the evolutionary forces that shaped the development of the

human mind must be rooted more in logical assumption than empirical fact.

       Moreover, even if we suspect that a psychological phenomenon is an evolved

trait, it is difficult to know whether it is a direct adaptation to an environment, an

exaptation (an evolved trait that comes to serve a different function), a spandrel (a side

effect accompanying another adaptive trait), a by-product of a distinct cultural process, or

some combination of these. This distinction is especially important when discussing

human psychology given the importance of cultures and social structures for shaping

human development (Richerson and Boyd 2006). Indeed, recent genetic evidence

suggests that the pace of evolution has actually increased since the advent of agriculture

in the last 20,000 years (Biello 2007). Thus researchers trying to piece the evolutionary

origins of the human political mind are faced with the dilemma of trying to determine

which psychological traits are likely to be direct adaptations, when they evolved, and

under what conditions they were selected, all with very limited tools for making direct


       Given of all of these problems, does it even make sense to explain contemporary

political behavior using an evolutionary model? The answer to this question partly

depends on the type of political problem and relevant trait in question. For some types of

political phenomena, an evolutionary explanation requires few stringent assumptions.

Take, for example, the propensity to engage in conspiratorial thinking. Throughout

recorded history and across many contemporary societies, there is wide-spread belief that

unseen, intentional, and malevolent forces, such as witches, Gods, or secret societies, are

behind catastrophic events (Marcus 1999). In a recent poll I conducted, roughly a half of

1,000 American survey respondents believed a secret cabal of “powerful forces are

behind most world events” and roughly one fifth thought the U.S. government had prior

knowledge of the September 11th attacks or that we are in the “end times” as foretold by

Biblical prophesy. Much of this thinking can be traced to a very adaptive cognitive bias

– the presumption that an unseen noise is being generated by a predator would confer a

tremendous adaptive advantage (Boyer 2001). It seems quite likely that the same

cognitive mechanisms which trigger fear and imagination in us when we walk through a

dark alley or woods would also underlie a propensity to embrace these types of

conspiratorial narratives.2

         But for many psychological traits that relate to politics, the assumptions about the

EEA must be more far reaching. Not only must the physical environment be taken into

account (which itself had wide climatic swings during much of the Pleistocene), but the

social environment must be considered as well. The evolutionary success of hominans

(i.e., modern humans and their extinct relatives), in spite of their gracility, is largely

attributable to the development of their social attributes, particularly their capacity for

cooperation.3 Cooperative and long-lasting relationships allowed hominans migrate out

of jungles and into open savannas (where a wider variety of game was available), be

highly altricial and sustain a high protein diet (both essential for disproportionate brain

growth in a bipedal species), and employ greater flexibility in response to changing

surroundings. The importance of cooperative relationship also meant that one of the

central environmental factors shaping human evolutionary development were other

humans. In short, hominans evolved in relationship to each other.

         Making assumptions about the evolutionary impact of social environments,

however, is fraught with hazards. To begin with, there are few empirical facts upon

which to make any claims. As noted earlier, the fossil record offers limited clues about

hominan social organization. The social structure of other primates, while fascinating (de

Waal 2005), must be evaluated in light of the tremendous difference between humans and

  There is also the interesting question of whether humans are hardwired to think in narrative terms. And,
of course, there is also the issue of whether these conspiracy theories are based in fact. After all, as the
saying goes, even paranoids have enemies.
  By cooperation I mean “individual behavior that incurs personal costs in order to engage in a joint
activity that confers benefits exceeding these costs to other members of one’s group.” (Bowles and Gintis

great apes in physicality, brain structure, and physical environments (e.g., chimpanzees,

bonobos, and gorillas are not bipedal, less encephalized, and confine themselves largely

to forests and jungles). Anthropological studies of the few remaining human hunter-

gatherer societies are often circumscribed by the circumstances of the ethnographic

research or the particular ecological conditions.

        Given the dearth of concrete evidence about hominid life in the Pleistocene, many

evolutionary theorists are now turning to more deductive methods, such as computer

simulations or game theoretic models, to estimate the impact of social environments on

the development of social cooperation. In such models, cooperation is typically

characterized as a strategic game, such as a prisoner’s dilemma, in which two or more

“rational” players are given a choice of cooperating or defecting. The puzzle which such

games try to solve is why cooperation occurs when the dominant strategy for any rational

actor is to always defect and never cooperate, even when the collective outcome is not a

Pareto optimum (i.e., the most advantageous for all parties).

        Over the past 40 years, a large body of literature has modeled numerous processes

in which a cooperative equilibrium can be achieved among otherwise rational and self-

interested actors. For example, in his famous simulation tournaments where various

strategies for playing the prisoner’s dilemma are pitted against each other in repeated

play, Robert Axelrod (1984) found that strategies like “Tit-for-Tat” or later what Alford

and Hibbing (2004) call “wary cooperation” have greater long-term success than a

“rational” strategy of consistent defection. In other words, strategies that seek to

cooperate but punish cheaters tend to outperform, over the long haul, strategies of narrow


       More recent work has tried to expand these models by incorporating further

realistic assumptions about human psychology. Rather than seeing cooperation develop

as the consequence of rational calculation among self-interested actors, these models

explore the impact of pro-social emotions, linguistic and physical capacities, and group-

oriented cognitive processes as evolved mechanisms for sustaining cooperation. In his

seminal analysis of reciprocal altruism, biologist Robert Trivers (1971) suggested that

affective processes such as “friendship, dislike, moralistic aggression, gratitude,

sympathy, trust, suspicion, trustworthiness, aspects of guilt, and some forms of

dishonesty and hypocrisy” could be seen as specifically adaptive for human cooperation,

what Orbell et al. (2004) call “rationality in design.” Bowles and Gintis (2003) elaborate

on the adaptation of these pro-social traits by putting them in the context of group

membership. According to their models, genetically transmitted individual behavior can

symbiotically evolve with culturally transmitted group-level practices particularly to

differentiate from proximate out-groups.

       As an example, Table 1 lists a series of affective states and their possible

functionality for sustaining cooperation. Individuals may face feelings of guilt or shame

for cheating on cooperative commitments but be rewarded with feelings of pride or self-

esteem for fulfilling them; feelings of indignation or rage may be directed toward specific

individuals who violate cooperative norms while affection or love is directed to those

who comply. These feelings may even work at the level of groups: feelings of prejudice

or animosity may be directed towards members of out-groups (simply for being part of

another group and hence a competitor) while loyalty, trust, and allegiance may arise

towards fellow group members. In other words, pro-social dispositions (such as shame,

guilt, or trust) evolved not towards all other humans but specifically towards fellow in-

group members, while more discriminatory traits evolved in relationship to out-group


           Table 1: Human Emotions and Their Roles in Enforcing Cooperation

                       Emotions Directed       Emotions Directed       Emotions Directed
                       Towards Self:           Toward Others:          Towards Group:

Punish Defection       Guilt, Shame            Indignation, Rage       Prejudice, Loathing

Reward Cooperation Pride, Self-esteem          Affection, Respect      Loyalty, Trust

       This work has been supplemented by research on brain activity during various

cooperation games, whose findings are consistent with predictions about pro-social

cognitive traits from formal models (Sanfey 2007). For example, Rilling et al (2003) find

consistent brain activation patterns among subjects playing a prisoner’s dilemma game

while under an fMRI scan. When making cooperative decisions, brain areas were

activated that are also associated with reward processing such as the nucleus accumbens,

the caudate nucleus, ventromedial frontal/orbitofrontal cortex, and rostral anterior

cingulate cortex . Similarly, de Quervain et al (2004) found, using a positron emission

tomography (PET) camera that experimental subjects who punished defectors in

cooperation games also activated similar parts of the striatum that are associated with

pleasure centers. Bechara and Damasio (2005) find that patients with lesion damage to

ventromedial prefrontal cortex have significant reductions in their capacity in performing

allocative tasks. In common parlance, human brains (and some primate brains as well)

appear “programmed” to activate pleasure sensations for both cooperating and for

punishing non-cooperators.

       Taken together, this research illustrates that much of human social cognition is

biologically rooted and thus evolutionarily determined. Even with the caveats noted

above, this suggests the potential of evolutionary theory for explaining human political

behavior. Not only does it offer a deep yet parsimonious scientific theory of human

action, it holds the potential for linking various disparate fields within social science.

Evolutionary studies of cooperation, for example, often unify insights from game theory

with observational studies from anthropology, field experiments in social psychology and

laboratory experiments in cognitive neuroscience (e.g. Gintis et al. 2005, Hammerstein

2003). In fact, evolutionary theory probably holds the greatest value in accounting for

political behaviors not amenable to rational explanations (i.e., the formation of

preferences and behaviors which incur substantial costs without tangible benefits to

participants). These include: the origins of ideology, the dynamics of group identity,

ethnic conflict, and discrimination, participation in elections and social movements, vote

choice and candidate preference, and general distortions in political cognition. But to

better appreciate the potential and limitations of an evolutionary theory, it may be useful

to focus on a particular example. For this, let us turn our attention to the case of political


The Evolution of Political Ideology

       Of the puzzles facing political scientists, perhaps none are more vexing than the

origins of people’s political preferences. Not only do citizens typically fail to meet most

basic expectations of rationality (such as having fixed, transitive preferences), it is often

difficult to discern much coherence in their belief systems (Converse 1964). Most people

exhibit highly unstable policy preferences and bear little resemblance to the informed and

ideologically oriented voters assumed by classical models of democratic politics

(Campbell 1962). But whiles sometimes incoherent, people are not entirely random in

their political behavior either. While they may be largely “innocent” of the ideology

practiced by political elites, most people do show evidence of having some core values

that constrain their thinking, guide their behavior, and orient them to the world (Feldman

1988). The difficulty for social scientists has been in identifying these fundamental

values, understanding where and how they become manifest, and pinpointing their


       Here is where evolutionary theory may provide some valuable insights. Consider

the question of where values come from. Social science offers surprisingly few clear

answers to this seemingly fundamental query. From the dominant social science

perspective, human values are typically described as social constructs learned from

particular cultures—values putatively arise from the social structure, the means and

relations of production, levels of economic development, family characteristics, age

cohort, religion, childhood experience, or ethnicity (Hitlin and Piliavin 2004, Inglehart

1977). Yet such an exhaustive list says little about why these particular forces shape

human values or why some values appear universal across cultures. In other words, each

of these explanations may account for some variation in values or for a particular cultural

practice but they say little about where the psychological inclination for the value itself


       In fact, the social science literature has largely ignored some fundamental

questions about the transmission of human values. It is generally assumed that humans

have a general-purpose mental mechanism for learning such values and, hence, any value

can be adopted by any person, but such an assumption introduces a host of questions:

what are the cognitive mechanisms that allow values to become learned or adopted? Are

these mechanisms innate or are these particular mechanisms of cultural transmission also

learned? Can all values be learned equally or are some easier to adopt than others (and if

so, why)? The social science literature has few answers to such questions, yet these

queries underlie the very core of how human value systems operate.

       If we examine values through the lens of evolutionary theory, we might actually

derive some alternative insights into their origins. From the perspective of evolutionary

theory, values are psychological mechanisms for processing information from the

external world in order to prompt effective action by the organism. This proposition

starts with the fact that all animals must manage large amounts of information about their

surroundings and, based on this input, determine an optimal response. Values are a

means of prioritizing this sensory information and instigating more advantageous

behavior. As Jaak Panksepp (2007) notes, “the central question for the behavioral

choices of humans and all other animals is the nature of the genetically provided value

system that index major survival concerns.” For example, at a most basic level, all

animals value self-preservation and reproduction. The value of self-preservation inhibits

self-destructive acts like walking off cliffs and helps with avoiding predators or

poisonous foods; the value of reproductive fitness instigates peculiar mating rituals, fights

for breeding prerogatives, and other sexual behaviors. While it is tempting to think of

values as a solely cultural, and thus human, construct, from an evolutionary perspective,

values are psychological phenomena shared by all animals.

       The biological mechanisms that enact these values, especially for larger

vertebrates, are primarily cognitive/affective. Most animals employ some rudimentary

cognitive heuristics or sensory algorithms for guiding behavior (Gigerenzer 2000) and as

animals become larger, more complicated cognitive systems evolved, particularly those

that prioritized sensory inputs and primed the organism for immediate reaction. One

example of this specialized cognitive/affective mechanism is fear. Arising from the

neural networks throughout the brain, but particularly in the amygdala, anterior and

medial hypothalamus, and certain areas of the mid-brain, the emotions of fear or anxiety

are rooted deeply to prompt quick action, such as a rise in adrenaline or heart rate.

Although the exact nature of affective brain processing is still the object of much debate,

the logic of affective systems for realizing preexisting values is clear: given the millions

of pieces of sensory input processed by animal at any given moment and most

deliberative forms of cognition are not only slow but highly taxing. Affective responses

help delimit cognition and prompt adaptive behaviors in an efficient manner. Emotional

systems of fear, rage, lust, and seeking are highly adaptive for promoting reproduction

and self-preservation and are widely observable across all vertebrates (Panksepp 2007).

       In contrast with insects and reptiles, mammals also evolved values that relate to

their functional interdependence, such as nurturance, separation distress, and inter-

subjectivity. And, as with their primordial values, mammals also evolved particular

affective processes to operationalize these values. Nurturance is facilitated through the

emotion of caring, separation distress through sadness or panic, and social bonding and

connection through playfulness. These physiological processes ensure bonding between

parent and offspring and, in some species, the maintenance of group cohesion. Once

again, these pro-social emotional systems are linked to particular brain areas (mammals

evolved a limbic core that processes social information) and distinct bio-chemical

processes to facilitate these neural processes. For example, feelings of caring and

nurturance are rooted in opioid, oxytocinergic, vasotocinergic and prolactinergic

neurochemical controls (Panksepp 1998). Other mammals that live in social groups also

demonstrate evidence of affective systems around social dominance, such as aggression,

status, and group coherence (Campbell 2006).

       And, in contrast with other mammals, humans and other primates are even more

distinctive for their even higher levels of sociability and patterns of cooperation among

non-kin. As noted above, theories about the psychological determinants of human

cooperation in the Pleistocene and Holocene are still largely speculative, but it is

plausible that humans would have evolved a number of innate values related to

cooperation with non-kin, values which are probably mediated by group life. These

would include valuing reciprocity, cooperation or assistance from in-group members or

other ways of reinforcing group ties, punishing non-cooperation among in-group

members, denigrating out-groups, and placing concern on status and reputation within a

group (Alford and Hibbing 2004). And, once again, these cooperative values, like other

genetically provided values systems, are directly tied and reinforced through particular

cognitive and affective mechanisms (as possibly sketched out in Table 1).

       Thus, from this perspective, human values are biological systems aimed at

promoting self-preservation, reproduction, kin success, and group cohesion. Rather than

arising solely from abstract principles or conscious deliberation, many human values are,

instead, intuitive and affective. The specific response of humans to most situations,

including those that represent acting on “principles” or “abstract” values, often precede

higher-order cognition (Greene 2007). Most choices are made before a person is even

away of the choice they face. Although some may offer a rationale for a particular

decision, this reason is the basis of this choice but, rather, a post hoc rationalization

(Haidt 2007). Because conscious deliberation about moral action is taxing and difficult,

it typically only occurs under conditions of uncertainty, either from a lack of information

or because values are in conflict (Tetlock 1983).

        Yet even from an evolutionary perspective, the characterization above does not

give a complete picture of human values. For in addition to being a social and

cooperative species, humans are also a distinctively symbolic species who have adapted

uniquely sophisticated mechanisms for language and higher order cognition. Other

animals may communicate crudely, but none have the generative, recursive, and open-

ended structure of human language (Hauser et al 2004). Language not only assists in the

transmission and enforcement of values (with language one can more easily establish and

communicate group norms), but can also lead to the creation of values themselves. For

example, language allows humans to assign what philosopher John Searle (2007) calls

“status functions” to certain objects or even other people. These status functions serve as

values for facilitating social exchange and delimiting individual behavior: a ten dollar

bill may have an ontological reality of being a simple piece of paper, but also hold the

status function as a mechanism of exchange and thus trigger particular behaviors such as

trading for a commodity. In fact, language enables deliberative cognition and the

formulation of abstract principles, the capability of valuation that we mostly common

ascribe to human ideology.

       The impact of symbols and language are also important for the development of

values because they enable a distinctive type of group life that, in turn, puts strong

adaptive pressure on humans to adopt even stronger social bonds. Evolutionary

biologists typically limit altruistic cooperation to either kin or small groups (Hamilton

1964). Many evolutionary psychologists, however, note that once humans are able to

symbolically differentiate particular groups or tribes, they can radically expand their

patterns of cooperation (through moralistic punishment, conformist biases, and impulse

control) to other group members in a way that selects for pro-social behavior at the group

level (Richerson and Boyd 1999, Wilson 2006). The values that help sustain cooperation

in large, non-kin groups, and may often times come in conflict with more innate,

individualistic values. For instance, groups may demand altruistic behavior on the part of

members that conflict with narrow, self-interested concerns. Thus once language and

culture allows for a strong level of group selection, it also creates a distinct type of

evolutionary environment, one based largely on symbols and group values.

       In other words, there is very strong evidence that humans developed biologically

in relationship to the distinct cultures of their particular groups—genes and culture

coevolved. One of the most blatant examples of this is lactose tolerance. Early hominans

did not possess the gene that allowed for the production of the enzyme lactase. When

Neolithic humans began domesticating and herding animals, mutations that allowed for

lactase production suddenly became highly adaptive and a distinct genetic difference thus

arose between herding and non-herding peoples. Evidence of this cultural legacy can

now be found in patterns of lactose tolerance around the globe, with high levels of lactose

intolerance in parts of east Asia and sub-Saharan Africa, areas that lacked herding

cultures (Richerson and Boyd 2002). It is possible that similar types of co-evolution also

arose with respect to particular cultural practices, particularly those related to the

restriction of aggression, sexual activity, or asocial behavior.

       The complexity of human values becomes even more vexing if we suppose that

one of the most important values that humans might have evolved was behavioral

flexibility. Compared to most mammals, humans can easily adjust their behavior to a

wide range of environmental and social constraints. Humans are particularly adept at

adopting cultural and social cues, largely because they are so dependent upon this group

information for their survival. Evolutionary theorists now emphasize how phenotypic

plasticity was highly adaptive in humans, particularly in their ability to respond to

parochial cultural information. Consequently, psychological traits, such as values, may

arise not simply from innate propensities toward that value (such as self-preservation) but

innate propensities towards learning particular types of values (such as group related

norms). This latter point will be of particular importance when trying to ascertain the

basis of human values that seem to vary so much across particular cultures.

       The challenge, therefore, when discussing the evolutionary roots of political

values comes in differentiating between those values that arise directly from biologically

adaptations (and thus universal to all humans), values that arose from the co-evolution of

genes and particular human cultures (and thus distinct to particular groups), values that

are universally latent but only expressive in certain cultural circumstances (and thus

manifest only in some groups but possible in all), and values that are solely learned. This

distinction is particularly important when referencing political preferences. Whether it is

voting, following rules, perceiving injustice, or forming opinions about policies or

candidates, many of the central issues in the study of politics are about how human values

operate within particular cultural contexts. Although some political values clearly arise

from individualistic psychological traits that evolved prior to homo sapiens, such as

concerns with safety or security, most political values are situated around collective

processes or outcomes and could have appeared at many points along the pathway of

human evolutionary or cultural development.

How to Identify Innate Human Political Values

       So how might we differentiate among political values that are innate, those that

are easily learnable, and those that are acquired only through cultural transmission?

Previous work on similar topics in evolutionary psychology provides at least one

plausible answer to this very difficult question: identify a differential cognitive capacity.

Rather than seeing the human cognitive system as a general-purpose information

processing device, evolutionary theorists are beginning to characterize the mind as a

hierarchical system of numerous, specialized adaptations selected for specific

environmental problems (Gangestad and Simpson 2006). As noted above, animals did

not evolve by simply developing solely a mental structure that generally processes

information and learns, but possibly adapted numerous specific processes that administer,

prioritize, and order information about particular sensory inputs and stimuli in

predetermined ways. Examples of such innate, specific mental mechanisms include

avoidance of vertical drops (an innate propensity in human and animal infants), spatial

orientation, facial recognition, language competency, social reasoning, and kin

recognition (Sperber 2005). Although there is a lively debate on the exact nature and

scope of mental specialization (Fodor 2000, Barkow, Cosmides, and Tooby 1992,

Carruthers 2002), most evolutionary theoriest embrace some minimal notion of particular

specificity in cognitive functions, particularly in regards to the processing of social

information (Geary 2005).

       Consequently, if humans have innate political values, then there should be some

identifiable evidence of the neural processes that enable them. For instance, if humans

innately value reciprocity, there should be some specific cognitive mechanisms that

become invoked when humans face allocating decisions—humans should process

information about concerns of equity in a specialized and predetermined way. Indeed,

Cosmides and Tooby (1992) have found that experimental subjects are far more adept at

applying Wason selection rules when couched in social terms rather than as mere

abstractions. From these experiments, Cosmides and Tooby conclude that “content free,

general purpose problem-solving mechanisms are extraordinarily weak—or even inert—

compared to specialized ones” and that the structure of human mental life, particularly in

relationship to social and political cognition and behavior, is rife with mental specificity.

       The degree of this cognitive specialization, however, is far from clear, particularly

in regards to its “content” specificity. The diversity of hominan environments and the

emergence of novel cultural constructs late in human evolutionary development would

have put enormous pressure on sustaining some flexibility in human cognition (Sperber

1994). More than most creatures, humans need a high degree of mental plasticity to

allow for learning and adapting to a variety of physical and social environments,

including patterns of social interaction.4 Complete plasticity is not only inefficient but

possibly calamitous, as the example of the Dodo bird illustrates.

        So the next challenge to an evolutionary based political theory is to identify which

neural processes are situated around a specific, pre-determined content and, in the

absence of specificity, what might be the rules or other patterns that allow for learning or

cognitive plasticity. One way to deduce these is through a catastrophe rule: if learning

about a specific stimuli would prove catastrophic, then it is likely that animals and

humans would have adapted particular mental functions in regards to that stimuli. In

hominan social environments, these specified contents might include kin-perception or

recognition of emotions. Otherwise, it seems likely that animals and humans would use

categories, heuristics, and other rules by which to process information about their

environments. Similar to the rules of grammar and syntax that underlie the “language

instinct,” these mental rules would systematically affect the way that relevant political or

social information is processed. Identifying these mental rules or heuristics may come

from looking at specific patterns or biases in information processing (Tversky and

Khanneman 198X), differential learning or memory capacities, and cross-cultural

similarities in political attitudes.

        As an example, consider how humans must deal with processing information

about dangerous animals. Poisonous spiders and snakes share nearly all the global

environments that humans do and would be a source of constant selection pressure. As

all spiders and snakes share a similar body form, it is not surprising that humans have

evolved an innate apprehension to these creatures and are particularly sensitive to

perceiving their morphology. Other dangerous animals, such as lions, bears, or water

buffalo, are limited to specific ecosystems. Cognition in response to these animals would

need to be more flexible and be based on particular morphological templates such as

body shape, eye contact, or displays of teeth and claws or on parochial learning (Barrett

2005). In this case, we would expect information processing to be situated around

heuristics and categories. If a person sees an ambiguous stimulus, such as a water

buffalo, it would need to be processed via quick heuristic categories (i.e., dangerous

animal, non-predator) but also allow for a wider range of deliberative action. In other

words, not all cognition is as reflexive as a massively modular mind would imply, and

some portion will be slower and reflective but still constrained by certain rules of thumb.

       In sum, there should be some distinctive patterns of cognition that reflect

evolutionary demands, such as the rapid transmission of certain types of information

(such as concepts that fall within particular categories or are of high salience), systematic

distortions in reasoning processes (such as with framing effects or other heuristics), and

differentiated memory processes. In regards to human politics, certain types of political

values should be evident in relationship to particular categories of conceptual primitives,

expressed in distinct patterns of cognition, and learnable in a differentiated manner. For

example, we might expect that humans have distinct mental modules about reciprocity

and thus be able to remember the names of non-reciprocators more easily or better

perform allocative tasks when framed in social terms. But we might also expect humans

to have some domain general process regarding reciprocity and thus exhibit behaviors

such as greater curiosity about violations of reciprocal norms or be more adept about

learning rules about cheating and the distribution of goods. We should expect that some

social rules will be more easily learned than others. For instance, my two-year-old

daughter has shown little problem in learning to exchange one toy for another but

demonstrates a frustratingly high level of resistance with learning how to share. This

seemingly common tendency may reflect a differentiated capacity for learning certain

types of cooperative behaviors.5

Food, Sex, and Death

        But while these evolutionary rules may tell us how to look for an “innate” or a

“learnable” value, it does not tell them where to look. The most common starting place

in this quest comes from cross-cultural surveys and anthropological research on

“universal” values. After all, if a value seems evident in every culture, then it seems a

good candidate for having a specified mental process. The difficulty in this approach,

however, is a lack of consensus about what human values are. For example, from his

synthesis of the social science research, Alan Fiske (1991) offers four elementary models

of social relations: communal sharing, authority ranking, equality matching, and market

pricing. Based on surveys from over three dozen countries, Shalom Schwartz (1992)

suggests ten other universal human values: self-direction, stimulation, hedonism,

achievement, power, security, conformity, benevolence, universalism, and tradition.

Surveying the anthropological record, Donald Brown (1991) offers even longer and more

complex list of universal attributes to human sociability such as self-conception, sexual

attraction, shelter seeking, pre and post-natal care, division of labor, and male dominance

  Another indication of mental modularity may be in the power of priming and framing in shaping political
opinion. The reason why such small changes in the presentation of an issue or question can have such large
effects on opinion or responses could lie in the types of conceptual modules being invoked (Chong and
Druckman 2007). Thus people may exhibit strong support for welfare spending when framed in terms of
the “deserving poor” because it evokes specific mental modules related to affective cues of compassion or
caring (Gilens 1999).

in the public sphere just to name a few. From this research and their own experiments,

Haidt and Joseph (2004, 2007) winnow down a list of five “moral foundations”:

Harm/Care, Fairness/Reciprocity, Authority/Respect, Ingroup/Outgroup, and


         Not only are these lists incommensurate with each other, they are largely

atheoretical. The universal lists are generated largely from surveys of social science or

contemporary cultures and are not made in reference to the environmental constraints

faced by early hominans. Consequently, they offer little distinction between which

values are innate or learned.6 This also means they have an inherently arbitrary quality.

None of the authors of these lists offers any explicit rationale for which items should be

included or excluded. One could easily critique each of these lists for what it leaves

off—Fiske’s work has little on coercion, Haidt and Joseph leave out self-esteem or the

integrity of self-concept. To better understand where to look for innate human political

values, it is important to consider the specific political conditions under which these

values would have been selected.

         At the risk of reinventing the wheel, it may be useful to start with some simple

deductions about hominan life in the Pleistocene. Cladistic analysis of primates suggests

that a common pan ancestor was a quadrupedal, arboreal ape, who lived in small, fission-

fusion social groups, subsisted largely on fruits and plants, and was capable of engaging

in forming temporary political coalitions to challenge status hierarchies (Boehm 2006,

Wrangham 1987). The evolution of a prehensile thumb, bi-pedalism, enlarged brains,

and tools use in early hominan species allowed for a wider hunting of mega-fauna on

  Haidt is currently engaged in a wide spread research project linking responses so certain moral items to
larger political tendencies.

savannas (which was also complemented by the evolution of hair loss, sweat glands, and

vascular brain cooling). These factors also introduced significant pressures on hominan

social structure. With the adoption of hunting tools, for example, status hierarchies based

on pure physical strength were no longer stable—alpha tendencies among males would

have been suppressed and sharing of game would have been more common. According

to Christopher Boehm (1999), the move to hunting large-game put strong evolutionary

pressure for humans to create moral communities with rules and enforcement

mechanisms. Indeed, it may have been precisely the distinct human adaptation of

language, higher order abstract thinking, and behavioral self-control that allowed for

these moral communities to expand and displace other types of hominan species, such as


       In these circumstances, it is plausible that distinctly political values began to

emerge (i.e., values that steered individual behavior in reference to collective concerns),

particularly in regards to three major categorical challenges to sustaining cooperation,

what I group into conceptual categories of FOOD, SEX, and DEATH. As illustrated in

Table 2, each of these political values would have reflected a tension between preexisting

value systems that promoted individual or kin interests and those that facilitated group

level success. Each of these would have also required particular (although not mutually

exclusive) types of cognitive adaptations to overcome the individualistic or nepotistic

values and, perhaps, to adjudicate the value ambivalence.

Table 2 – Value Conflicts in Human Evolution

                       Value Conflict         Mental Module         Political
 FOOD                  Sharing vs. self-      Equality,             Egalitarianism
                       preservation           Reciprocity, Status
 SEX                   Sexual restraint vs.   Taboo Activity,       Moralism
                       reproductive           Status, Purity
 DEATH                 Limiting violence      Group Affiliation     Ethnocentrism
                       vs. status

       Consider, for example, the category of FOOD. Food provision in the time of the

shared ancestor was probably similar to that of other primates—most food is acquired

individually with some sporadic collective hunting activities for small game (Wrangham

1987). Consequently, the nutritional economy required little adjudication. Once

hominans started hunting large fauna on the savanna, they needed to better coordinate

their activities and to develop means of appropriating food stuffs. Although this may

have generated numerous cognitive adaptations, it seems likely that hominans would

have become extremely sensitive to issues of sharing, social status, and reciprocity. On

the one hand, exclusionary consumption of game would have posed a significant

challenge to sustaining group cohesion (and thus put a pressure on developing egalitarian

values within the group); on the other hand, purely altruistic behavior would be

collectively untenable and individuals would have valued some differential compensation

for hunting skill. Most likely this would have resulted in some expression of social status

and reciprocity around hunting success. This pattern is evident in many hunter-gatherer

societies, where successful male hunters typically exhibit greater sexual activity (Ridley

1996), but this might have also varied among groups. Nevertheless, it seems plausible

that hominans would have evolved a set of what could be loosely labeled egalitarian

cognitive processes. These would be mechanisms that would prioritize the identifying

the status and distributive behavior of other group members, would utilize certain rules or

heuristics for interpreting social information, and would provide a keen interest in

learning rules or norms about the distribution of resources.

       Similarly, the adoption of hunting weapons would have also demanded new

cognitive adaptations with respect to SEX. The shared ancestors, like most primates,

coercively organized themselves in a status hierarchy to adjudicate sexual prerogatives

(albeit to varying degrees of stringency). They also had innate tendencies towards incest

avoidance, probably adapted through affective mechanisms registering disgust and

cleanliness (Douglas 1966). Once again, the development of projectile weapons

dramatically altered this social dynamic. As raw physical coercion would be less tenable,

sexual prerogatives would have been more fiercely contested, as when any primate

society is in a period of status disequilibrium (Mastripieri 2007). Hominans would have

faced very strong adaptive pressures to evolve new mechanisms for defining and

sustaining sexual prerogatives. Status could be established relative to hunting prowess,

leadership, or social skill (Cummins 2005). Sexual restraint could be enforced through

group sanctions regarding appropriate sexual behavior and linking status position to

perceived moral violations, particularly among females. The confluence of these forces

would result in a series of cognitive processes particularly attuned to learning moral

norms, perceptions of norm violations, and different perceptions of acceptable behavior

by age, sex, and family position. This could be manifest in specific inference systems

concerning individual sexual behavior and affective responses to non-sanctioned sexual

behavior. Such developments also would have created a tension between innate values of

promoting reproductive advantage (which would encourage sexual licentiousness) and

group cohesion around maintaining an equitably sexual economy. Together, these would

underscore a significant set of moralistic mental processes concerned primarily with

identifying and regulating individualized sexual behavior in relationship to group norms.

If this is the case, then there should be different patterns of information seeking and

processing, memory, self-regulation, and learning around sexual behaviors.

        Another pressure on hominan cooperation would have been in relationship to

channeling aggression within and among human groups, what I put under the heading of

DEATH. Lethal aggression among primates rarely occurs among group members

contesting status hierarchies; rather, it is limited largely to members of out-groups or

during periods of hierarchical instability (Cummins 2005). The advent of hunting

weapons, however greatly increased the power and danger of intra-group violence.

Boehm (1999) suggests that the reduction of alpha bullying behavior was the first moral

imperative that hominans faced and that strong group sanctions would have evolved in

relationship to restraining aggression against in-group members. The creation of such a

moral community, however, would have created pressure for a greater group identity, as

extending non-aggression to out-group members would have been maladaptive.

Presumably, cognitive mechanisms evolved focusing attention on group related norms,

in-group versus out-group membership, and wariness in regards to expressions of

aggression depending on the target. These collective processes could be grouped under

the heading of ethnocentrism.

       Together, these three categories represent the major adaptive pressures by

hominans to develop specific types of cognitive competences in regards to issues of

group life. Several characteristics of these mental mechanisms are worth noting. First,

nearly all of the values concerning group life conflict in some way with narrow

individualistic values of self-preservation or maximizing reproductive fitness. Although

it does not mean that such individualistic or nepotistic values will disappear (as they

would remain highly functional across many domains), but rather new cognitive

mechanisms would have to develop that could either suppress or override them. Most

importantly, this also means that some level of ambivalence is inherent to any system of

political values, an ambivalence that would need to be reconciled through some type of

cognitive or affective mechanisms (and reinforced through nascent cultural practices).

This would imply that any types of mental modules relating to political values would

have to retain some type of flexibility or be in conflict with one another.

       Second, although these categories of mental modules would have arisen in

response to distinct adaptive pressures, these categories were, by no means, mutually

exclusive. Status differences would have impacted the distribution both of food and

sexual prerogatives, concerns of reciprocity would have occurred for sex and resource

distribution, and group cohesion would have been important for identifying appropriate

targets for sharing and reproduction. Although issues of FOOD, SEX, and DEATH each

present some distinct challenges to cooperation, there is also a high degree of overlap and

interaction among them as well.

       Third, many of the values would invoke a combination of content-specific

cognitive processes and general heuristics. The parochial nature of human social life,

particularly in regard to specific ecological niches, would create a high degree of

variation in the patterns described above. For instance, humans living in resource poor

environments may be far more attuned to cheating or non-reciprocity (Boehm 1999).

Although some types of political knowledge will be semantic (i.e., organized along

specific principles, cued by particular stimuli, and based on specific neural structures),

other will have to be flexible to reflect war and inter-breeding among groups. The

difference between these two would be most apparent where learning would be both

costly and inaccessible. For instance, learning that snakes bite or that walking off cliffs

leads to falls can be catastrophic. In political terms, such semantic knowledge could

include a general propensity for differentiating between in-groups and out-groups,

monitoring sexual or violent behavior in others, attention to patterns of reciprocity,

punishing the violation of social norms, or engaging in altruistic behavior with fellow

group members. Even as the definition of each of these actions could be parochial and

thus learnable, the cues of such actions (and thus the ability to differentiate them) would

presumably be innate. In short, the ability to learn most political values is based partly on

the presence of innate value systems that relate to the functioning of group life.

       In sum, this type of evolutionary explanation could hold great promise for

opening up a new way of understanding political values. From an evolutionary

perspective, we could say that most people are “innocent of ideology” not because they

are valueless but because common political discourse does not neatly align with the way

they cognitively organize their values. As noted above, all political values entail some

inherent ambivalence for their holders and people may also experience conflicts along

many of the same issues. For example, Americans may seem inconsistent or hypocritical

in their racially motivated support for welfare spending (Gilens 1999), but this is because

welfare invokes so many contradictory value dimensions. Most political issues are

largely abstract and do not concern issues that most people encounter in their daily lives

and few people are offered the opportunity to develop the cognitive proficiencies that

generate highly predictable behaviors (Lieberman 2006).


        Like any nascent scientific endeavor, an evolutionary approach to political

psychology raises as many questions as it answers. Perhaps the biggest is differentiating

between what types of political behaviors arise from innate psychological traits and what

types are learned or contingent on the specific environment. The answer to this question

depends upon how we understand the evolution of the human mind. There is a

considerable debate about how much of our cognition is delimited by thousands of

distinct and encapsulated modules adapted for specific problems and how much of it is a

general-purpose, information-processing mechanism. If it is the former, then how

specific are these modules? How do they interact? How do they relate to learning

processes? These are essential questions for which there are no clear answers at this


        Even if these issues get resolved, there will still be numerous other challenges for

understanding the linkages between innate psychological traits and exhibited political

behaviors. As an example, consider the question of why some sets of political values

predominate in some people over others. In other words, why do some people stress

egalitarianism but not moralism (i.e., a contemporary political liberal), while others

deemphasize egalitarian concerns and concentrate on moralistic issues (i.e., contemporary

conservatism)? From an evolutionary perspective, one can imagine at least three

important considerations that would have to be measured. First, given the genetic

diversity in the population, the traits that promote egalitarianism or moralism may be

more pronounced in some people than others: some people may be “hardwired” to be

more conservative or liberal. To test for this, we would need to identify the genetic basis

of egalitarian or moralistic tendencies and see if mutations occurred along certain alleles.

       Second, socialization and childhood experience may evoke some kinds of traits

and suppress others, particularly if we think that many modules are constructed as

learning devices. Political psychologists have long theorized that childhood experiences

could shape adult ideology. Altemeyer (1988), for example, reports that childhood

spanking is a significant predictor of later authoritarian attitudes. Presumably some types

of political values may become manifest during childhood. To test for this, we would

need information not only on life history, but to measure some kind of effect between

individual experience and the expression of a psychological trait.

       Third, we would need to consider the current social and political circumstances of

the subject. In light of the oft-repeated adage that a conservative is a “liberal who has

been mugged,” and one might consider whether certain traits become expressed

depending on one’s political situation. Such contextual variability is evident with other

types of bio-psychological processes. For instance, testosterone levels in males vary

considerably depending on one’s status position (Cummings 2006) – men who lose

contests or gambling bets often demonstrate significant drops in testosterone and

aggression. The importance of context maybe particularly high in regards to the political

values that evolved largely around concerns of regulating group life. The salience of any

political value will fundamentally be related to the relative position of the individual vis-

à-vis the group. Thus, someone with disproportionate low resources should be more

concerned with egalitarian cues, while another person inexorably aligned with particular

groups may be more oriented around ethnocentric considerations. These individualized

dynamics will also be reinforced or suppressed by larger social and institutional forces.

The likelihood of acting on egalitarian political values will be a lot higher if there are

other group members in a similar position. To make matters even more complicated, we

should also expect that each of these three considerations above (innate tendencies,

learned experiences, and contemporary social circumstances) would interact with each

other as well.

       Given the daunting nature of these types of issues and the sheer complexity of the

human brain, we may very well question the viability of pursuing research along

evolutionary lines. After all, the brief example above only scratches the surface of all the

possible linkages between political values and cognition. Other issues that would need to

be considered include self-concept, memory, learned competencies, abstract reasoning

processes, and the dynamics of language. Despite these facts, the potential of

evolutionary theory remains a powerful lure for understanding contemporary political

psychology. As long as the limitations of our inferences are kept clearly in sight,

evolution could provide some fruitful new perspectives on solving some of the most

puzzling questions about human political behavior.

Figure 1: Three Dimensions of Political




To top