Ontogeny and Social Dominance: A Developmental View of Human Power Patterns

Abstract

Developmental science has long evolutionary roots and has historically focused on individual differences. Accordingly, developmental models can inform conversations about phylogeny and personality. The present paper evokes life history theory to describe a theoretical model of competitive behavior that applies to both children and adults (resource control theory: RCT). The model suggests that prosocial and coercive behavior, though different in manifest form, serve similar evolutionary functions. Accordingly, RCT presents a view on social dominance that gives primacy to function over form that contrasts sharply from traditional views. This reformulation gives rise to novel questions (both developmental and non-developmental) and challenges long accepted views on prosociality (e.g., that it is altruistic) and aggression (e.g., that it is maladaptive). Similarly, RCT gives rise to a minority perspective that aligns aggression with social competence.

Evolution, Development, and Individual Differences

In 2011, David Buss and I published the first edited volume on the evolution of individual differences in humans (Buss and Hawley, 2011). Our claims were quite lofty, the volume “heralding a sea-change in thinking within evolutionary psychology” (Hawley and Buss, 2011, p. xvi). We claimed this because variation in the population (i.e., individual differences) is known to be fundamental to the evolutionary process. Without it, natural selection could not occur. However, like our contributors, we turned the logic: Individual differences are not only the input for selective forces, but also their outcome.

This minority view made the volume unique. To many, the “evolution of personality” is nearly an oxymoron. The logic for dismissing individual differences is fairly straightforward: Adaptations, though admittedly context specific, are “universals,” and variability surrounding some modal response pattern can be reasonably seen as error variance or noise (Tooby and Cosmides, 1990). Accordingly, most scholars in evolutionary psychology heretofore have focused on said universals, except for variants associated with very broad categories (e.g., sex). This state of affairs may seem perplexing given the early success of the behavioral genetics paradigm and the fact that personality factors have long been known to play a key role in adaptive outcomes (Buss and Greiling, 1999; Jokela, 2012; Jokela, Kivimäki, Elovainio, and Keltikangas-Järvinen, 2009; Ozer and Benet-Martinez, 2005). In fact, for several decades, quantitative (Figueredo, Petrinovich, and Ross, 1992; Hawley and Little, 2003) and theoretical (Gosling, 2001; Wilson, Coleman, Clark, and Biederman, 1993) personality work was being conducted with animals (Carere and Maestripieri, 2013).

Though at least two broad classes of evolutionary theoretical frameworks give rise to empirical work on the evolution of personality ¹ , one stands out as particularly relevant for our present purposes. Namely, until fairly recently, a long standing puzzle (ontogenetically, if not phylogenetically) was the fact that multiple morphs (some visible and striking, others subtly quantitative) simultaneously exist in populations. The most widespread across taxa is reproductive (male, female), recognized as such even by Darwin (1871). In many fish and insects, these morphs can be so strikingly different that specimens were long mistaken for different species, making taxonomy work especially challenging (Gadgil, 1972; Gross, 1991). Take Schistocerca gregaria, for example. It will become a harmless, solitary grasshopper if the nymph is hatched under a low population density where it is unperturbed by the presence of others. However, if hatched under densely populated conditions (coinciding with resource scarcity), it is more likely to become a social, swarming locust. In response to physical stimulation in the nymph phase, neurotransmitters (e.g., serotonin) are released. In turn, these proximal mechanisms restructure its behavioral and physical phenotype (Homberg, 1991). Its manifest form is dependent on its local conditions, hence “condition dependent adaptations” (i.e., “alternative tactics”; Caro and Bateson, 1986). That is, key environmental inputs calibrate the system toward a favorably competitive strategy for the prevailing local environmental conditions. Yet, the genome results in no permanent changes.

“Plasticity… is a universal property of living things” (West-Eberhard, 2003; p. 34). Responses to environmental change can be discrete or graded, reversible or irreversible, behavioral or morphological, adaptive or non-adaptive (Caro and Bateson, 1986; Pigliucci, 2001; Whitman and Agrawal, 2009). Plasticity, like individual differences, was largely dismissed by biologists as “noise” (Whitman and Agrawal, 2009) until roughly two decades ago, when biologists began focusing on proximate causation and ontogenetic histories (Caro and Bateson, 1986; Whitman and Agrawal, 2009).

The above described evolutionary model (cf. Footnote 1) is inherently developmental. Developmental psychologists ² have long explored proximate mechanisms for abrupt, ontogenetic behavioral changes, though the present language used to describe said processes was adopted not from biology, but rather medicine (e.g., “risk factors”; Jessor, 1991). In sharp contrast, one research tradition that famously borrowed from evolutionary biology and ethology was attachment theory. John Bowlby and Mary Salter Ainsworth addressed similar calibrations, though in language appropriate for their time. Drawing on biological models, Bowlby proposed that the mother-infant bond was the most important first context for humans that provides the basis of personality development, one in which automatic and ancient primate behaviors (e.g., reflexes) function to maintain proximity to the primary caregiver (Bowlby, 1969/1982). These adaptations, Bowlby explained, take different forms, including grasps, separation anxiety, vocalizations, smiles (positive affect), and cries (negative affect). In other words, positive and negative social cues in the infant, though very different in form (phenomenal manifestation), serve a similar function (purpose); in this case, maintaining proximity to the primary caregiver. On this latter point evolutionists are well versed; in the natural world there are several ways (forms) to solve similar problems (function).

Bowlby's perspective was refreshingly modern, and, as a clinician, he appreciated the importance of differential response patterns resulting from varying ecologies:

Instinctive behavior is not inherited; what is inherited is a potential to develop certain sorts of system, termed here behavioral systems, both the nature and the forms of which differ in some measure according to the particular environment in which development takes place. (Bowlby, 1969/1982; p. 45)

Mary Ainsworth elaborated the evolutionary foundations of attachment theory to include a more articulated perspective on the calibration of personality, cognitive schema about the self, others, and one's relationship to the material world (e.g., Ainsworth, Blehar, Waters, and Wall, 1978). In other words, the nature of the attachment relationship calibrates one to the social environment in which one finds oneself (e.g., secure, avoidant, anxious). More recent instantiations have extended attachment theory to adult reproductive strategies (Belsky, Steinberg, and Draper, 1991; Mikulincer and Shaver, 2007). Humans, not unlike Schistocerca gregaria, are in some sense polymorphic, dependent on one's local ecological conditions.

Given developmental psychology's deep evolutionary roots, concern with individual responses to environments, and experience with early calibration processes, developmental scholars are uniquely prepared to weigh in on evolutionary processes and individual differences.

Life History Theory (LHT)

All species have a phylogenetic history that shaped the modal course of their life trajectories. Life history theory (e.g., Roff, 1992) is the metatheoretical framework that describes and explains these processes and outcomes as functions of ecological variables, such as environmental instability and unpredictability (Pianka, 1970) and concomitant rate of population loss. A species “life history” is the modal pattern of a lifelong stream of adaptive trade-offs for allocating finite resources toward competing life functions, including growth, reproduction, and survival (West-Eberhard, 2003). These patterns have been described in terms of being fast (rapid development, early reproduction, brief gestation, short life) or slow (protracted development, later reproduction, longer gestation, longer life) on a continuum. Fast strategists (formerly known as “r” strategists) are well represented in the parasite world, the constituents of which generally live briefly, come to reproductive age quickly, and bear hundreds of offspring that require no parental care but at the same time suffer high mortality rates. In contrast, humans (like apes in general) tend to be positioned on the “slow” end of the continuum (formerly referred to as “K” strategists) as indicated by our slow maturation, late reproduction, and bias toward bearing few, large-brained, slow-developing offspring that require a good deal of parental care.

Generally, life history theory has been applied to cross-species variation (interspecies questions) in the field of biology. Important for our present purposes, it has also been successfully applied to account for variation within species, as demonstrated by S.g. above, which takes different forms to serve similar functions (reproductive competition) in the different environments in which the organism finds itself. Moreover, the framework has revealed the variability in life history courses across humans (Ellis, 2004; Figueredo et al., 2006; Rushton, 1985). LHT's explanatory power lies in its explicit recognition that organisms are responsive to complex inputs (the modality of detection varies across species) from their physical and social environments and, accordingly, are predictably flexible in their developmental course This intraspecific variation includes not only physical characteristics (e.g., environmental sex determination in some turtles, snakes, and reptiles; Bull, 1983) but also continuous behavioral traits (e.g., aggressiveness, agreeableness).

Attachment profiles are a case in point. With LHT lenses (e.g., Belsky et al., 1991), avoidant attachment is not seen as a maladaptive response to suboptimal environments simply because it leads to unappealing behavior (e.g., aggressive self-assertion). Rather, avoidance is seen as a persistent adaptive psychological response to a certain psychological climate within the family ecology (Frankenhuis and del Giudice, 2012). In other words, attachment styles are different solutions to the problems faced by the child after birth (an individual's adaptation to local circumstances), or, like S.g., different forms to serve similar functions in the different environments in which the organism finds itself. The group's response to those solutions is a secondary consequence to that adaptive process. Three primary points from LHT are important to set up the argument that follows. First, humans, like many organisms, have evolved a degree of structured responsivity to key social and material cues in the environment (i.e., phenotypic plasticity). This first point is well-documented by developmental psychologists, and we know what these social and material cues are (e.g., poverty, harsh parenting). Where developmental psychology and LHT part ways, however, is on whether this calibration leads to maladaptation or something that can be seen as beneficial to the organism (e.g., Ellis et al, 2012; Hawley, 2011). “Human nature,” then, is in part decided by the context within which we find ourselves. Bickering over which is the true human nature will generate more heat than light.

Second, we must distinguish form from function in all we do as evolutionists. To those in the field, this may seem obvious. But there has been at least one domain, I will argue, where this canon has fallen through the cracks.

Third, individual differences in behavior (including sociality) can be seen as inherently competitive or cooperative insofar as individuals are attempting to maximize gains and minimize costs in the presence of others doing the same. In this sense, environmental cues carry information about important aspects of the competitive landscape that organisms detect and respond to. These enduring individual differences (traits) can be considered “strategies” in LHT and behavioral ecology parlance (Ellis, Figueredo, Brumbach, and Schlomer, 2009; Hawley, 1999; West-Eberhard, 2003), and they can sometimes take surprising forms.

Natural selection and the competitive ethos

Natural selection at some level implies competition insofar as asymmetrical outcomes are a key driver of evolution (Dawkins, 1976; Hamilton, 1964). That is, adaptations, or those alleles or phenotypic traits that come to be well-represented in the population, have done so at the expense of others that become less well-represented. This point of course is non-controversial and has been demonstrated in laboratories, via mathematical modeling, and shown to occur in actual populations.

Conflating evolutionary and psychological egoism

Conventionally – and mistakenly – this underlying competitive foundation is taken to mean that animals behaviorally compete, as if in some naturally occurring Olympics. Of course some do this, but these types of contests are only a small minority of the types of competition the theory of evolution by natural selection predicts.

For example, from very early on it was understood that prosocial behavior (e.g., helping, sharing, cooperating) and emotions (e.g., sympathy, empathy) are just as much a part of nature as are aggression and agonism (e.g., Kropotkin, 1902). In the meantime, these perspectives have been well-elaborated (e.g., Axelrod, 1984; Bowles and Gintis, 2011; de Waal, 1996). Humans (and wolves, and elephants, and ants, and slime mold spores) have evolved to be nice to one another.

Parental love, for example, is inherently biologically selfish, and attentive and protective parents can out-reproduce inattentive ones. Moreover, attentive parents rear children who are themselves more successful parents than are children of inattentive parents. Yet, this scenario is ill-described with appeals to psychological egoism. In fact, neither psychological egoism nor “competition” in the conventional sense are required of superior strategies. On the other hand, it is preposterous to think of a cellular slime mold spore that adopts the role of the supportive stalk and consequently foregoes its own reproduction to support that of another cell as psychologically altruistic (see Hawley, 2014 for extended handling of psychological and evolutionary altruism and egoism).

Biological altruism (or selfishness) applies to every living organism, regardless of where you stand in the eusociality-inclusive fitness debate. Psychological altruism/selfishness applies to very few. Having a brain of some kind is a bare minimal condition. If we're going to create a general model of behavior, we should consider the whole continuum of life and eschew human exceptionalism as a philosophical starting point. Yet, one still finds confusion even among academics over whether evolutionary self-interest (or “selfishness”; Dawkins, 1976) means behavioral or psychological self-interest (e.g., Dodge and Albert, 2012; Gintis, 2013; Nowak, 2006). ³ Darwin (1871) conflated the two as well. ⁴

Consequences of conflating evolutionary and psychological egoism

This confounding of psychological and biological egoism has led to a number of circuitous detours in evolutionary circles, and has seriously confounded non-evolutionary writers' attempts to incorporate evolutionary logic. One sees this confounding emerge regularly in debates over group- versus individual-level selection, for example, and has led some very influential, venerated writers to claim that group selection accounts for the origin of “virtue” whereas individual level selection gives rise to “sin” (Wilson, 2012). These discussions are highly visible because teams of outspoken eminent writers are sparring in open forums (see, for example, Edge.org; Abbot et al., 2011). These discussions are interesting to the lay public because one side seems to provide the biological underpinnings for morality and religion while the other does not.

But in one domain of inquiry, this influence has crept in invisibly to great effect; namely, in discussing hierarchical political systems, such as leadership, egalitarianism, and social dominance. If you want humans to be “good,” you will favor leadership and egalitarianism, and eschew social dominance. In my view (and as I have argued at length elsewhere in the developmental literature over a decade ago; Hawley, 1999), this is a false and misleading debate based on overly narrow conceptions of dominance and a conflation of form and function.

Dominance: What's in a Word?

Dominant: Exercising chief authority or rule; Occupying a commanding position.

Etymology: Latin domināt- participial stem of dominārī to bear rule, govern.

The above definition of dominant was drawn from the Oxford English Dictionary, which indicates this is accepted use since the mid-16th century. Important for our purposes, notice that nowhere in this definition is implied how dominance is accomplished. In terms of the form-function discussion above, this is a functional definition. ⁵

Dominance and dominance hierarchies entered the modern biological vernacular in the early 19^th century. At that time, narrow views of the competitive essence of natural selection predominated. Accordingly, this view led early animal behavior researchers to look for hierarchical behavior in asymmetries in agonism and aggression. The very first dominance hierarchy to be described was the peck-order of chickens (Schjelderup- Ebbe, 1922). Quite literally, it was a description of how hens compete for desired resources by pecking each other. This work was highly influential and gave rise to the aggression-based view of social dominance that persists to this day across species and taxa (e.g., Alonso, Honji, Moreira, and Pandolfi, 2012; Gintis, 2013; Hofer and East, 2003; Ishikawa, Yamada, Matsuura, Tsukada and Tsuchida, 2011; Krebs and Davies, 1997; Pelletier and Festa-Bianchet, 2006). Textbooks and journal articles define dominance as a “sexually selected manifestation of conflict” (Alcock, 2005; p. 332), a “pure exercise of power by socially dominant males” (Gintis, in press), where “alpha males control group activities and others are intimidated or forced to acquiesce” (van Vugt, Hogan, and Kaiser, 2008, p 191; see also Boehm, 2000).

Moreover, the peck order construal quickly impacted work in child development (e.g., Bühler, 1927), and similarly has left a legacy of continued focus on the form of behavior (agonistic contests, fighting ability, submissive gestures), even though the function of such contests is always understood to be resource control (the theoretical model; see Hawley, 1999 for extended discussion; Carpenter, 1942 for conflation of form and function in a single definition ⁶ ) For example, Abramovitch (1976) defined dominance as property fights, involving “struggles” or “tussles.”

Consequences to favoring agonism as a strategy of resource control

There are at least two consequences of this form-focused view that equates dominance with agonism. First, it is (adult) male centered, which leads to several consequences of its own. Second, it ignores the instrumentality of being nice.

Dominance and the masculine political structure. Males are more physically aggressive than females as early as the age of 3, on average (Maccoby and Jacklin, 1974). This is not controversial. However, it is then just a small step to conclude that social dominance is the purview of males, and a smaller step further to conclude, therefore, the predominant political structure in group living social species (i.e., hierarchical structure) is essentially masculine. This conclusion is reified into an assumption. Not surprisingly then, it is often claimed that socially dominant males hold the central positions in their social groups (Gintis, in press) and consequently enjoy differential attention and grooming, as well as priority in ally selection and access to mates (de Waal, 1982a; Dunbar, 1988).

There is nothing wrong with this focus as it led to discovering the signaling value of dominance and submission displays, and uncovering even in “simple” species (e.g., fish; Hsu, Early, and Wolf, 2006) the ability to learn from one's previous interactions with a conspecific. I would argue, however, that the agonism view is incomplete, or certainly so say primatologists who have beautifully described and explained the complex strategies – some nice and some nasty – that primates use to “occupy a commanding position” (Boehm, 1993; de Waal, 1982a). Indeed, neither chimpanzees (Boehm and Flack, 2010) nor human infants (Mascaro and Csibra, 2012) need see any act of aggression to draw important and enduring conclusions about relative ranks among others (see also early work on leadership in humans that similarly demonstrated that humans can “exercise chief authority or rule” in variegated ways; the most eloquent and thorough, in my view, was Machiavelli's [1513/1966] treatise).

An unintended consequence of this focus on male aggression is that the political structures of females were (and continue to be) downplayed relative to males' (as were patterns of aggression in females). Indeed, some prominent primatologists (e.g., Hrdy, 1981/1999; Drea, 2005; Gowaty, 1997) have argued that female political structures, though less visible to field workers, are in fact longer lasting. And presumably because behaviors used for hierarchy ascent take a different form (but not a different function), they have been much harder to document (e.g., reproductive suppression). In contrast, the function-focus approach presented here shines light on variegated ascension strategies, and as a side-effect (not a central goal), female status-striving.

Form-focus ignores instrumentality of being a good group member and a good cooperator. Second, being a good group member is highly instrumental. I use the term “instrumental” because it is more neutral than its linguistic cousins, “egoistic” or “selfish,” which are highly pejorative and a source of confusion in psychological circles. “Instrumentality” neutrally connotes that some goal is attained by what an organism does, and yet it says nothing about psychological egoism or altruism, which are, for the most part, irrelevant for the present work.

Cooperation, for example, is a very effective material goal pursuit strategy (Charlesworth, 1996). Together we can pursue goals (e.g., grants) that neither can successfully attain alone. Admittedly, the word “cooperation” is tricky because it has been used in different ways. For example, evolutionary perspectives from economics use the term interchangeably with altruism, which in biological circles (not psychological circles) generally means engaging in behavior that enhances the fitness benefits of another while bearing a fitness cost to the self (−/+). Thus, in this literature, cooperation is assumed to be cost-bearing (e.g., Boyd, Gintis, Bowles, and Richerson, 2003; Dreber, Rand, Fudenberg, and Nowak., 2008; Nowak, 2006: “…the self-interest model predicts no cooperation”; Gintis, in press, p. 3), its extreme form manifesting in eusociality where most of the group members are of castes that forego reproduction altogether to support the reproduction of a few (Nowak, Tarnita, and Wilson, 2010). Cooperation, then, is viewed in these circles to demonstrate some degree of sacrifice for the welfare of others.

The above view may strike some as narrow and semantically confusing because of the cooperation-altruism conflation. Say “altruism” to a psychologist and they think motivationally “selfless.” Use cooperation and altruism interchangeably, the psychologist assumes cooperation is motivationally selfless. The surface logic seems sound. Therein lies the danger of using metaphor to explain biological concepts (see Hawley, 2014, for extended treatment).

Instead, the present view is more explicitly in line with Adam Smith's proposal that cooperation stems not out of self-sacrifice, but rather self-interest as a primary concern. Smith, unaware of biological evolution, could only then be talking about psychological self-interest (e.g., I sell you bread so that I can make money. You having bread is secondary to me). Like other evolutionary thinkers concerned with various forms of reciprocity (e.g., Hamilton, 1964; Trivers, 1971), the function of the behavior is biological self-interest (with psychological self-interest being wide open to debate) and it is “cooperative” in terms of it benefitting another (see also West, Griffin, and Gardner, 2007). Self-sacrifice is not inherent to this point of view (see also Alexander, 1987; Trivers, 1971) as the behavior can be mutually beneficial (+/+; West et al, 2007). Indeed, mutual benefit seems to underlie the formation of human society in ways that sets us apart from the apes (Nowak, 2006; Boehm, 1999). Our “other-regarding preferences” ⁷ reflect a commitment to self and other simultaneously (see also Bakan, 1966; Boehm, 1999).

That humans can be reciprocally nice (i.e., prosocial) as a consequence of evolutionary forces is not controversial. The high-level, sometimes acrimonious discussion in top tier journals is about evolutionary mechanisms underlying cooperation (e.g., individuals vs. group; Abbot et al., 2011; Nowak, Tarnita, and Wilson, 2011), none of which have any bearing on the present piece, accentuated by a failure to distinguish biological from psychological self-interest. For our purposes, it is enough to say that one benefits from group living because solitary living bears enormous costs in many domains (food procurement, protection, rearing young; Krause and Ruxton, 2002), and these benefits naturally select behaviors that increase the probability that the group accepts us as valued group members. Valuing others is a case in point. People who valued others were more likely to treat group members with respect and goodwill than were people who did not value others (this of course is true even today). They were more likely to make friends and benefit from these friendships as we still do (e.g., social and material support, child rearing; Hrdy, 2009). As a result of a long history of humans forming mutually beneficial friendly relationships (i.e., indirect reciprocity), the desire for attachments to others is seen as a basic and foundational human need with deep evolutionary roots (Baumeister and Leary, 1995; Bowlby, 1969/1982). It is a need, that when fulfilled, pays (e.g., Silk, Alberts, and Altmann, 2003). Thus, it pays to be nice to others, especially if those others are in a position to bestow benefits. ⁸ This latter point is foundational to resource control theory.

Resource Control Theory

Resource control theory (Hawley, 1999), with roots in the work of human ethologists (e.g., Abramovitch, 1976; McGrew, 1972; Strayer and Strayer, 1976) and animal behaviorists (Bernstein, 1981; Chance, 1967; Rowell, 1974), is an evolutionary developmental theory of social dominance that, as has been set up above, is a theory of individual differences in behavior and personality, considers the developing child to be responsive to key social and material cues in what is essentially a competitive environment. As with other animal species, outcomes to these daily competitions are presumed to be highly visible. I use the function-focused term “dominance” to set up important questions about the development of behavioral forms (i.e., strategies) – which may be either negatively valenced (e.g., agonism) or positively valenced (e.g., cooperation) – used to “occupy a commanding position” (OED).

Resource control

Here, resource control refers to the extent to which individuals successfully access social, material, or informational resources relative to others. This definition is sufficiently general to include access to and attention from high status others (social), objects meeting one's survival needs and denoting status (material, e.g., food, clothing: Sahlins, 1963; van Vugt and Hardy, 2010), and valuable information regarding work, school projects, or events (informational; Hawley, Shorey, and Alderman, 2009; Keltner, Gruenfeld, and Anderson, 2003).

Resource control strategies

Meeting one's needs in a social group, or successfully competing for resources in the presence of others, can certainly be achieved directly, expediently, and without consideration for the goals and motivations of others. But they can also be met by coordinating efforts with other group members.

Coercive strategies. Following the lead of ethologists and biologists, RCT recognizes the utility of aggression, especially in contexts that have a zero-sum quality. Elaborating somewhat to accommodate humans' rather advanced cognitive capacities, deception and overt manipulation reside in this category (Byrne and Whiten, 1988). These strategies characteristically exploit individual group members or bypass the social group entirely.

As straightforward as these strategies seem, they are the source of a good deal of controversy. This friction is important for our present purposes because it points to the value-added by the present view. Namely, there is a prevalent assumption that social groups are intolerant of such brazenly (e.g., psychologically) self-serving behaviors. This assumption has been made in anthropological (Boehm, 1999), economic (Fehr and Gächter, 2000), and child development circles (Coie and Dodge, 1998). Yet, in the field of child development, this wisdom is being chipped away: There is emerging evidence that suggests that a sizeable subset of the aggressively powerful youths are highly regarded by peers and are even sought out for friendships and alliances. They are anything but ostracized. RCT holds a morally neutral view to such behavior (i.e., eschewing monikers of virtue and evil) and instead it predicts that those occupying commanding positions (i.e., those who have demonstrated competitive success, the socially dominant) will win positive regard because of their evident effectiveness in the material world. Simply put, we should find power attractive, even if this power is wielded aggressively (the social centrality hypothesis; Hawley, 1999). This point will be re-addressed in more detail later.

Prosocial strategies. Quite unlike coercive strategies, prosocial strategies' theoretical roots are in the evolution of cooperation and reciprocation (e.g., Trivers, 1971; Charlesworth, 1996). Here, competition takes on a non-zero sum quality; multiple interactants can gain in this cooperative or reciprocal context. Instead of bypassing the social group, as direct means do, indirect strategies exploit the mediating effect of the social group to access resources prosocially. These strategies too are the source of some irritation in the literature predominantly centering on confusion about the term “prosocial.”

In specific terms, prosocial behavior refers to “voluntary actions that are intended to help or benefit another individual” (Eisenberg and Mussen 1989, p. 3). Important for our purposes, especially in light of the common conflation of biological and psychological self-interest, this definition does not favor any one of a variety of possible underlying psychological motivations (egoistic or altruistic) (Campbell and Christopher, 1996; Eisenberg, 1996; Eisenberg and Giallanza, 1984). Nor does this definition rule out self-benefit. Thus, prosociality may be blatantly psychologically egoistic (e.g., performed consciously for present or future reward), subtly egoistic (winning favor from a colleague), or psychologically altruistic (i.e., motivated only by true other-oriented/selfless concerns; Eisenberg and Mussen, 1989). In reality, underlying psychological motivations are likely complex. Yet, most studies in the field of child development have tended to focus on altruism and have shown that even at an early age, children are capable of psychological altruism or the precursors thereof (Hoffman, 1983, 1994; Zahn-Waxler, Radke-Yarrow, and King, 1983). All of these motivations, however, may in fact be ultimately (evolutionarily) instrumental. Natural selection does not require you to know why you're doing what you're doing; only that your genes benefit from it, directly or indirectly. This point would be true regardless of whether one found individual level selection (direct benefit) or group selection (indirect benefit) more compelling (Hawley, 2014).

The development of social dominance and power

Bowlby understood that both positively valenced (smiling, touching) and negatively valenced behaviors (e.g., crying, protesting) are instrumental in human infants; both pay in terms of strengthening the bond to the caregiver, the first source of social, material, and informational resources. These strategies are honed and elaborated in the peer group via operant and social learning processes (i.e., selection by consequence). A child may learn that pinching is highly effective at wresting the swing from a peer. Teachers and parents, however, will attempt to teach her that smiling and sharing is more effective and socially preferable. These foundations give rise to interesting developmental questions.

What predicts dominance differentials and do they organize non-competitive social behavior? It has been repeatedly shown that it is possible to organize children in a group according to high and low dominance rank (e.g., Strayer and Strayer, 1976). Doing so only has meaning, however, if these relative competitive differentials are associated with behavior outside of the competitive context, an influence that speaks to the central organizing power of this aspect of relationships (I have argued elsewhere that the early ethological work had limitations in this regard; Hawley, 2007). Our own attempts at borrowing from the social relations model of social psychology (Kenny and LaVoie, 1984) suggested that when put in pairwise interactions in a non-competitive play situation (where there are plenty of toys for both interactants), dominance rank nonetheless mediates the relationship between characteristics of the child (e.g., cognitive age, gender, temperament) and play behavior. These path models spoke to two important issues: the predictors of relative success in toddlers and the power of these differentials to predict non-competitive social behavior (Hawley and Little, 1999).

Regarding the first issue, we began to chip away at long-held beliefs from ethology; namely, that size and gender (i.e., big males) predict social dominance (but see Lukaszewski and Roney, 2011; Sell, Tooby, and Cosmides, 2009 for a different perspective). In fact, when one includes psychological variables, we found that persistence (temperament), cognitive age (as assessed by the Bayley Scales of Infant Development), and familiarity with the setting predicted dominance well more than physical size. Moreover, all things being equal, girls were favored over boys in the preschool years.

Regarding the second issue, when paired with a child of lower rank, children were more likely to engage with the play material and issue directives (e.g., “you color here”), and less likely to imitate and passively watch (Hawley and Little, 1999). Children of middle rank behaved like dominants when paired with children of lower rank, yet deferred to and imitated those of higher rank. These relationships were moderated by the degree of familiarity of the children; the more experience they had with each other, the stronger effect of their relative ranks on their respective behavior. All children were under the age of 3 (i.e., toddlers), which suggested to us just how basic these abilities and learning processes are.

What types of behavior in children are associated with resource use? When do they emerge and what is their developmental course? One of the key and most controversial questions raised by RCT involves the instrumentality of prosocial behavior. Can prosocial behavior provide self -benefits? To this end, we set up quasi-experimental situations to look closely at what children do to access a highly attractive limited resource (i.e., a novel toy). In the course of filming and coding behavior from multiple dyadic interactions, we documented that prosocial behaviors (e.g., making helpful suggestions, issuing polite requests, offering “help” to commandeer the play material, initiating unequal trades) not only were associated with resource use (r = .53), but were also the most often employed strategy class (at twice the frequency of coercion: taking, aggression, insults, also associated with resource use). Moreover, and critical to the present perspective, both strategies were highly related to each other (r = .67), just as one would expect of behaviors sharing common function. Children employing these behaviors effectively controlled our resource over 70% of the time. Though not the most popular aspect to our program of research, we do have good evidence that prosocial behavior can be instrumental across multiple age groups (see, for example, Hawley, 2002 for evidence of its emergence in preschool; see also Green, Cillessen, Rechis, Patterson, and Hughes, 2008; Pellegrini, 2008; Roseth et al., 2011; Teisl, Rogosch, Oshri, and Ciccetti, 2012).

Are there mixed strategists? Coercive and prosocial strategies find their analog in game theoretic models under the guise of hawks (escalation) and doves (caution). As with more complex game theoretic models, RCT does not see humans as pure strategists; more interesting to us is a model where “players” bring distinct genotypes that are calibrated by early environments, with the modal (life history) pattern being one of a mixed strategy (see Hawley, 2006 for details). That is, most humans flexibly use both strategies to some degree, the relative employment of which is an important source of individual differences. Indeed, in principle, it would be optimal to be a “mixed strategist.” Those who are able to cooperate and aggress are able to minimize the cost associated with prosociality (e.g., exploitation) by drawing on aggressive behaviors to punish free riders. Machiavelli essentially described the mixed strategist on his treatise of political power (Berlin, 1980; Machiavelli, 1513/1966).

With this framework in mind, our program has focused on types of resource controlling individuals depending on their relative employment of the two strategies. The degree to which an individual employs the strategies can be measured by way of observation (Hawley, 2002), self-report, or other report (Hawley, 2003a,b). We then compare these types across dependent variables of interest to derive qualities associated with the two strategies alone and in combination. Over the last decade (Hawley, 2002; Hawley and Geldhof, 2012), we (and, more recently, others: e.g., Chen and Chang, 2012; Roseth et al., 2011; Olthof, Goossens, Vermande, Aleva, and van der Meulen, 2011) have generated five subgroups of individuals: bistrategic controllers employ both strategies to a high degree relative to peers, coercive controllers employ coercive strategies to a high degree, prosocial controllers employ prosocial strategies to a high degree, and non-controllers are low on both relative to others (i.e., they are not resource directed). Typical controllers are the largest remaining group and are average on both strategies and as such represent the modal life history pattern and a baseline for comparison. We know quite a bit about these profiles across multiple age groups, and the profiles have in turn revealed some interesting and counterintuitive (from most views) aspects of aggression and prosociality.

For example, prosocial, coercive, and non-controllers possess very distinct profiles of socially appealing traits and corresponding social success. Prosocial controllers are above average on resource control (thus reasonably socially dominant), highly socially skilled (e.g., socially perceptive, extroverted, morally astute), agreeable, and very well-liked by their peers. They know and internalize moral norms, and behave well within them (Hawley and Geldhof, 2012). Evolutionists and non-evolutionists alike would agree that these individuals, as a group, are highly socially competent peer leaders. Perhaps not surprisingly, females comprise the majority in this group. In stark contrast, coercive controllers, while being reasonably high in resource control, fit the stereotypic profile of impulsive, unskilled, and socially repellant aggressors (Hawley, 2003a; Hawley, Johnson, Mize, and McNamara, 2007). By the age of 5, they are already showing signs of social rejection by their peers. Perhaps also not surprising, given the physicality of pure coercion, males comprise the majority in this group. Traditional perspectives have handled these children well and have identified them as needing ameliorative measures.

Most informative to our perspective, however, are the individuals who employ both strategies to a high degree. These “bistrategic” controllers tend to be the most successful at resource control in all age groups. Consistent with our predictions, we have generally found bistrategic controllers to possess attributes associated with social skills; i.e., they tend to be extroverted, socially perceptive, and morally astute (Hawley, 2002, 2003b; Hawley and Geldhof, 2012). But unlike skilled prosocial controllers, they are as a group very high on traditional measures of aggression, and even direct this aggression towards their best friends (Hawley, Little, and Card, 2007). In this regard, they are much like coercive controllers. Yet, despite their high levels of aggression, and in contrast to expectations from most theoretical perspectives, they demonstrate relationship skills (e.g., intimacy; Hawley, Little, et al., 2007) and are socially attractive to peers (Hawley, 2003a,b; cf. Henrich and Gil-White, 2001). They understand moral norms well, but they have not internalized them in an emotional sense. This combination, unlike prosocial controllers, allows bistrategics to manipulate the moral atmosphere to their own advantage (Alexander, 1987; Hawley and Geldhof, 2012). Interestingly, both males and females are well-represented in this group. And both genders possess skills typically associated with the opposite sex (e.g., males are attuned to others whereas females are highly aggressive). Thus, we find bistrategics as rather androgynous behaviorally, in the sense that they embrace all human skills and strategies and can employ them flexibly and in ways that garner some degree of group approval. They may even enjoy differential reproductive success (Jokela et al., 2009).

Bistrategic controllers contradict views held by the majority that aggression does not pay in the long run because it is associated with social punishment and ostracism. On the contrary, aggression seems to work very well when employed as part of a mixed strategy. The effectiveness and social savvy of the bistrategic struck us to parallel Machiavelli's treatise. ⁹ Thus, we have referred to them as Machiavellian (Hawley, 2003a; see also Baar and Wubbells, 2011; Chen and Chang, 2012; Palmen, Vermande, Deković, and van Aken, 2011; Roseth et al., 2011). And because they not only meet their material needs more effectively than all other groups, while at the same time also meeting their social needs of power and affiliation, we have argued that their behavior pattern may represent a type of human social competence (Hawley, 2002; Stump, Ratliff, Wu, and Hawley, 2009).

If there are winners, there must be losers

Views stemming from behavioral ecology suggest that multiple strategies can achieve similar average payoffs (i.e., different forms can successfully perform different functions). Non-controllers, however, should not be seen in this light because their payoffs do not appear to be equal to other groups. Indeed, lack of resource directedness (and by extension, total lack of aggression) appears to be a losing strategy. Everything the bistrategics are, the non-controllers are not: They are low on resource control, extraversion, social skills, aggression, social status, and peer regard, and high on anxiety and withdrawal. This “niche” is cost-bearing. In fact, non-controlling school children are the gravest clinical significance (Stump et al., 2009; but see Hawley et al., 2009); they are ostracized and victimized by their peers. Thus, non-controllers confront ideological views to social competence as well. What does it mean when aggression protects you from failing to thrive?

What these profiles reveal about conventional wisdoms

First, prosocial strategies are not altruistic; they reap true benefits for the actor (see also Hardy and van Vugt, 2006; Griskevicius, Tybur, and Van den Bergh, 2010; van Vugt and Iredale, 2012). Moreover, participants readily admit that “being nice” to others is the best way to ensure that “others are nice to you.” That is, not only are these behaviors not altruistic in the biological sense, but also not altruistic in the psychological sense. ¹⁰ The self-serving nature of prosocial behavior is mostly keenly understood by prosocial and bistrategic controllers, i.e., those who use them frequently and effectively. Indeed, these individuals report high levels of material and social goals, and explicitly understand the relationships among them (Hawley et al., 2009).

Second, aggression (i.e., blatantly self-serving strategies) does not unilaterally lead to ostracism. Bistrategics clearly show this undeniable fact. You can gain positive peer regard (i.e., prestige; Henrich and Gil White, 2001) by wielding power aggressively if you also possess social skills to foresee and smooth over potential repercussions. It is not that the group does not see their aggression, because they clearly do (e.g., Hawley, Johnson, et al., 2007). College students simply report that the benefits for associating with someone who can sometimes be a very volatile person outweigh the costs. Bistrategics are fun, innovative, and know how to get things done. On the flip side, being non-aggressive does not win friends. In a sad twist of nature, non-controllers are socially at risk.

Third, relative competitive ability is less gendered than is commonly thought. When allowing for variegated strategies, females do quite well (Hawley, Little, and Card, 2008; Hrdy, 1981/1999). Form-focused approaches to social dominance would include only bistrategic and coercive controllers as power holders; that is, those who were physically aggressive. Add prosocial controllers to the mix, and the gender representation equalizes.

Working RCT into the Theoretical Landscape

Reversing the hierarchy

The present view does not explicitly call for leveling mechanisms, or normative strategies that function to equalize the hierarchy (i.e., “reverse dominance hierarchy”; Boehm, 1999) primarily because it does not equate dominance with unbridled aggression. These leveling behaviors (i.e., counter dominant behavior; Boehm, 1993) against bullying that have been ethnographically documented (e.g., gossip, ostracism, or corporal punishment) are generally interpreted to mean that humans possess a propensity for egalitarianism (see also chimps: Boehm and Flack, 2010). However, Boehm himself has stressed that this pattern might be expected within some, small, autonomous hunter-gatherer groups, but certainly not “all early human societies.” Moreover, Boehm explicitly added that “egalitarianism” in these model societies does not extend to women and children as a rule.

Aside from the obvious narrowness of a male-oriented perspective, it is easy to conclude that, if we were “egalitarian” in early human groups, then we must be “naturally” egalitarian (or “inherently moral creatures”; Gintis, 2012) or, at the very least, intolerant of social dominants. Work in the RCT paradigm seems to suggest something quite different. When aggression is mitigated with social skills, we seem to be drawn to it even before the age of 3. And why wouldn't we be? The powerful make better friends and alliance partners than the socially subordinate. Exchange theories in social psychology have illuminated these dynamics (e.g., Blau, 1964; Kelley and Thibaut, 1978), and work in developmental circles have documented relationship processes of powerful adolescents (e.g., friends of powerful individuals become more powerful themselves over time; Dijkstra, Cillessen, Lindenberg, and Veenstra, 2010; Marks, Cillessen, and Crick, 2012). These patterns occur right under our noses at scientific conferences. That is not to say that currying favor from the powerful is cost-free; powerful individuals are more likely to be aggressive, socially inappropriate, and sexually uninhibited than the less powerful (see Keltner, Gruenfeld, and Anderson, 2003 for review). In any case, these patterns too appear to be “natural.” “Egalitarianism” may be the exception rather than the rule. And so much the better if one can operate with a “veneer” of morality, as aggressively socially dominant preschoolers already do (Hawley and Geldhof, 2012). Both perspectives agree, however, that behavior patterns of group members are context specific, and a good deal of the “context” is what other group members are doing to access limited resources.

Leadership and prestige

The present view overlaps significantly with more recent views on leadership (e.g., van Vugt et al., 2008) and prestige (e.g., Henrich and Gil White, 2001), even though these lines of work hold that these structures are distinct from dominance (indeed, the “antithesis” of dominance because leadership is proposed to be a solution to intragroup conflict caused in part by dominance striving: van Vugt et al., 2008; p 188). Here, leadership, with benefits of its own, is viewed as a group resource having evolved in response to pressures for group coordination and conflict resolution. Again, I believe that the distinction relies on the overly narrow view of social dominance as aggression. As mentioned above, it has long been known that socially complex species with protracted developments “work” their hierarchies with variegated strategies (Byrne and Whiten, 1988; de Waal, 1982a,b; Drea, 2005; Hawley and Little 2003; Hrdy, 1981/1999). These strategies are especially complex in humans (e.g., Kyl-Heku and Buss, 1996; Raven and French, 1958) and undoubtedly include behaviors we would call leadership (Hawley, 1999). Prosocial strategists, for example, are exceptional at achieving their own goals while simultaneously supporting the goals of others. In the end, it may be necessary to develop separate theories for leadership (a gentler style of command than coercion) in humans and social dominance in animals. I have made the case elsewhere, however, that it is not necessary or even desirable (Hawley, 1999). In fact, many of the phenomena highlighted by the leadership literature have been documented in rudimentary form in animals (e.g., group coordination, initiating group action, maintaining cohesion, competence recognition, conflict management). Moreover, the dominance literature has never found subordinance to be a “riddle” as the leadership work has “followership” (cf., van Vugt et al., 2008). Accordingly, the present view does not hold dominance to be “the antithesis of leadership” (van Vugt et al., 2008; p 188) because it does not equate dominance with coercive domination, nor does it see coercion as an “opposite” to prosociality; instead, RCT holds them to be “two sides of the same coin” (Hawley, 2002, p. 175). These bodies of work will most certainly find more points in common than in conflict.

Conclusions

Many authors have long recognized a dualism inherent to human nature; a tension between self and other which is insufficiently solved by attending solely to the self or solely to others (e.g., pure psychological egoism or pure self-sacrifice). The present evolutionary view sees this tension as a result of competitive processes inherent to natural selection giving rise to both antagonistic and other-regarding behavioral strategies. Developmental processes (e.g., attachment, social learning) are presumed to underlie change over time, driving strategies to higher levels of sophistication and subtlety. By being mindful of the important distinction between form and function and their interrelationships, and being clear about the difference between psychological and evolutionary egoism, we can bring central human social dynamics and hierarchical structures into clearer focus.