Evolved individual differences: Advancing a condition-dependent model of personality

Highlights

• Personality taxonomies do not explain the origins or structure of personality.

• Condition-dependent evolutionary models may help fill this explanatory gap.

• Conditional mechanisms’ outputs depend on the benefits of alternative trait levels.

• I advance novel hypotheses about trait variation and covariation.

• I conclude with future research directions for explaining personality differences.

Abstract

The field of personality psychology offers a wealth of robust empirical research and a successful descriptive taxonomy, but neither explains the origins of the structure of human personality nor elaborates a generative framework for predicting the specific conditions that evoke the development of distinct personality traits. Exploration of traditional personality constructs within an evolutionary adaptive individual differences framework may help fill this explanatory gap. Personality traits exhibit functional features and patterns of variation expected from psychological adaptations designed to solve survival- and reproduction-related challenges recurrently faced during our species’ evolutionary history. Condition-dependent evolutionary models of personality have been proposed for decades, but only recently have begun to see empirical investigation. These models posit that species-typical psychological mechanisms take as input cues from the individual’s phenotype that would have been ancestrally linked to differential cost–benefit tradeoffs of alternative personality strategies, and produce as output personality trait levels with the greatest probabilistic net benefit for the individual. This paper elaborates a more nuanced conceptual framework that builds on earlier conceptualizations of condition-dependent traits to yield new and untested hypotheses about personality trait variation and covariation. It then describes clear future research directions for empirically investigating these readily testable hypotheses.

Introduction

At present, the field of personality psychology offers a wealth of robust empirical research and a successful descriptive taxonomy, but does not answer why personality differences take on the structure that they do, or elaborate a generative framework for predicting the specific conditions that evoke the development of distinct personality trait levels. An adaptationist evolutionary psychological approach, which proposes that many human behaviors, cognitions, and emotions are the output of psychological mechanisms designed to solve distinct adaptive problems (Buss, 1995), may offer a cogent predictive framework for identifying the causal processes responsible for the development of personality traits and the social contexts that activate them.

Within an adaptive individual differences framework, different personality traits can be conceptualized as functional strategies that help solve specific problems recurrently faced by members of a species during its evolution (Buss, 2009). In the study of humans, this adaptationist perspective is generally applied using a “top-down” approach (Buss, 1995). First, the researcher identifies a specific challenge to survival or reproduction in ancestral environments. Second, the researcher articulates the behaviors that would have helped solve this adaptive problem, as well as the cognitive processes and emotions that would have motivated these behaviors. The researcher then conducts empirical tests for evidence of these hypothesized, functionally specialized cognitive, affective, and behavioral design features.

Personality psychology has historically operated outside of such an a priori predictive theoretical framework, focusing more on the statistical structure of individual differences than on the potential evolutionary functional origins of those differences (Buss, 1987, Buss, 1990, Buss, 1991a, Buss, 1991b, Buss, 1996a, Buss, 1996b, Buss, 1999, Buss et al., 1992).

Recent work (e.g., Kanazawa, 2011, Penke et al., 2007, Verweij et al., 2012) has explored multiple potential evolutionary models for individual differences in personality. However, theorists have largely overlooked the possibility that species-typical psychological adaptations produce individual differences in personality, instead favoring models that assume more direct gene → personality effects. Given the relative neglect of the powerful, but under-utilized tool of condition-dependent adaptations, this paper places a particular focus on adaptive individual differences that emerge from condition-linked differences in the costs and benefits of alternative personality traits.

Although personality and evolutionary psychology have traditionally had different foci of empirical inquiry, an exploration of the evolutionary functionality of the Five-Factor Model (FFM, Costa & McCrae, 1985), one of the most widely validated models of human personality and whose dimensions are exhibited in a diverse array of non-human animal species (see Gosling and John, 1999, Nettle, 2006, Smith and Blumstein, 2008), illustrates how an adaptationist conceptual framework may be fruitfully applied to the study of human individual differences. The high pole of extraversion in humans, for example, may be conceptualized as an interpersonal strategy that can increase mating opportunities (MacDonald, 2006). High levels of extraversion could lead to increased mating opportunities both directly by engaging potential mates and indirectly by leading to the formation of friendships and social alliances that facilitate increases in status and ascension in the social hierarchy (Denissen and Penke, 2008, Nettle, 2005, Nettle, 2006). Data from non-human animals offer evidence consistent with this hypothesized function of extraversion; bold behavior by Trinidadian guppies (Godin & Dugatkin, 1996), zebra finches (Schuett & Dall, 2009), and collared flycatchers (Garamszegi, Eens, & Török, 2008) is associated with increased mating success. A hypothesized function of high agreeableness is that it facilitates successful collective action by leading individuals to deeply engage in and focus on cooperation to achieve group goals (Denissen & Penke, 2008), an interpersonal orientation that is also invaluable in a long-term mate. Indeed, in some nonhuman species, signals of non-aggressive strategies appear to increase individuals’ desirability as long-term mates (see Ophir and Galef, 2003, Ophir et al., 2005). High levels of conscientiousness are hypothesized to promote successful pursuit of long-term goals such as good health and longevity by means of determination, self-discipline, and delayed gratification (Denissen and Penke, 2008, Nettle, 2006), and the creative problem-solving capacities exhibited by individuals high in openness to experience may lead to enhanced status and increased mating opportunities (Haselton and Miller, 2006, Lewis et al., 2014).

Even high levels of neuroticism, a personality trait that has historically been framed exclusively as “maladaptive” (Grant, 2011, p. 42), may at least partly reflect the output of evolved psychological mechanisms. Several theorists have proposed that humans possess evolved psychological mechanisms designed to elevate neuroticism levels as a functional response to the threat of social exclusion (Denissen and Penke, 2008, Nettle, 2005, Nettle, 2006). Higher levels of neuroticism are associated with endogenously driven attentional shifts (Flehmig, Steinborn, Langner, & Westhoff, 2007), which may guide attention toward negative social outcomes such as relationship exclusion or dissolution. Such selective attention to potential threat cues (Gallagher, 1990, Hemenover and Dienstbier, 1996) and focusing on negative information (Hemenover, 2001) may result in greater sensitivity to potential negative social outcomes (Grant, 2011, Kuppens and Van Mechelen, 2007), as well as greater worry and anxiety in response to potential relationship threats. In turn, these cognitive and affective states may motivate behaviors such as vigilance and guarding of one’s relationship partners to protect limited relationship opportunities (Denissen and Penke, 2008, Nettle, 2005, Nettle, 2006).

The key idea is that traditional personality constructs are not only amenable to exploration within an adaptive individual differences framework, but also exhibit characteristics expected of psychological adaptations designed to solve fitness-relevant problems recurrent in human ancestral environments.

Although a given strategy on a particular personality dimension may serve reproductive benefit-linked functions, each strategy also carries potential costs. In non-human species, exploratory or bold behavior may increases risk of predation (guppies: Dugatkin, 1992, Godin and Davis, 1995; theoretical model: Wolf, van Doorn, Leimar, & Weissing, 2007; for review, see Smith & Blumstein, 2008), and among humans, extraversion can similarly carry fitness-relevant costs – extraverts are disproportionately represented in hospitals with injury or illness (Nettle, 2005) and their pronounced sensation-seeking can lead to traumatic injury (Field & O’Keefe, 2004) and legal trouble (Ellis, 1987). Similarly, high levels of agreeableness can carry fitness costs; individuals who avoid conflict are less desirable as mates in a variety of species, including humans (e.g., fighting fish: Doutrelant & McGregor, 2000; Midas cichlid: Barlow, 1986; humans: Lukaszewski & Roney, 2011), and high agreeableness may lead individuals to forgo their own objectives and risk social exploitation (Judge, Livingston, & Hurst, 2011). Although high levels of neuroticism may cognitively and affectively motivate an individual to protect limited social opportunities, high neuroticism is associated with impaired somatic health (Cohen and Williamson, 1991, Glaser and Kiecolt-Glaser, 2005, Herbert and Cohen, 1993, O’Leary, 1990) and can place burdensome strain on social relationships (e.g., Buss, 1991a, Neeleman et al., 2002). Even high conscientiousness, a trait rarely regarded as undesirable, may lead individuals to forgo unanticipated, but valuable opportunities. Importantly, this includes opportunities that could dramatically increase reproductive fitness, such as opportunistic short-term mating (Schmitt, 2004). Although high openness is associated with greater creativity, it is also associated with social withdrawal, delusional thoughts, and risk for schizophrenia and related disorders (McCreery and Claridge, 2002, Nettle, 2009). In short, the pursuit of any given personality strategy is associated with both potential benefits and potential costs (Buss, 1990, DeKay and Buss, 1992).

The benefits of pursuing a given personality strategy depend on whether an individual faces the adaptive challenge the personality strategy is designed to help solve, how effective the strategy is in solving the adaptive problem for that particular individual, and the benefits that accrue to the individual by successfully solving the problem (Buss, 2009, Denissen and Penke, 2008, Nettle, 2006, Penke et al., 2007). The costs of the strategy depend on the potential costs inherent to the strategy itself (e.g., extraversion-associated injury risk) and the likelihood of the individual incurring those costs, as well as the opportunity costs to the individual – the benefits that the individual would obtain by pursuing an alternative strategy.

Crucially, these variables influencing the cost–benefit tradeoffs of a given personality strategy differ across individuals as a function of their condition. An organism’s condition refers to its phenotypic quality (e.g., physical attractiveness, strength, see Lukaszewski & Roney, 2011), and reflects the organism’s “ability to efficiently convert energy into fitness-enhancing traits and outcomes” (Lukaszewski, Larson, Gildersleeve, Roney, & Haselton, 2014), or “overall fitness budget” (Gangestad et al., 2010, Tomkins et al., 2004).

A condition-dependent evolutionary psychological model posits that species-typical psychological mechanisms take as input condition-linked cues predictive of differential costs and benefits of alternative personality strategies in ancestral conditions, and produce as output the personality strategy of greater probabilistic net benefit for the individual, given his or her condition (see Tooby & Cosmides, 1990 for their seminal theoretical discussion of this issue; see also Buss and Greiling, 1999, Nettle, 2006, Wolf et al., 2007).

Considering the condition-dependent costs and benefits of extraversion helps to illustrate this concept. Extraversion draws attention to one’s characteristics (Anderson and Shirako, 2008, Ashton and Lee, 2007, Nettle, 2005), which can advertise one’s positive attributes, but also initiate competition and provoke conflict with rivals (Lund, Tamnes, Mouestue, Buss, & Vollrath, 2007). The benefits of broadcasting information about one’s physical characteristics are not uniform across individuals – it is more beneficial for attractive than unattractive individuals. Similarly, engaging in potentially conflict-initiating social behaviors is more costly for less physically formidable individuals (McDonald, Navarrete, & Van Vugt, 2011). Consequently, the net benefits of extraversion are likely higher for stronger and more attractive individuals (Benson, Karabenick, & Lerner, 1976). If the psychological mechanisms responsible for the introversion–extraversion continuum calibrate the manifest behavior they produce according to cues linked to these costs and benefits, then we should expect stronger and more attractive individuals to exhibit higher levels of extraversion. Indeed, individual differences in attractiveness and strength are predictive of individual differences in extraversion (Lukaszewski & Roney, 2011).

Relationships between individuals’ psychological and morphological attributes have been documented since early in the history of personality psychology (e.g., Kretschmer, 1931, Sheldon, 1940, Sheldon, 1942). However, without an articulation of the process by which individuals’ morphological condition and personality are connected, these condition-personality links lack a cogent explanation – let alone a powerful theoretical tool for predicting which facets of an individual’s condition should be linked to which personality dimensions and in which direction.

At first blush, the postulation that humans possess species-typical psychological adaptations that calibrate personality levels according to an individual’s condition might not immediately appear to accord with the known stability of individual differences in personality (Caruso, 2000). However, considering the patterns that condition-dependent personality mechanisms would be expected to produce reveals how a condition-dependent model of personality actually predicts stable individual differences. To illustrate this point, let us further explore the hypothesis that species-typical psychological mechanisms calibrate individuals’ neuroticism levels according to the individual’s threat of social exclusion, which varies at least partly as a function of the individual’s condition (Denissen and Penke, 2008, Nettle, 2005, Nettle, 2006.

Neuroticism’s stability may be at least partially accounted for by temporally and cross-situationally stable links between individuals’ morphology-based desirability as a relationship partner and their social exclusion. Throughout the lifespan and across social domains, individuals who possess a less desirable morphological phenotype experience less social acceptance – and exhibit higher levels of neuroticism (Mathes & Kahn, 1975). Less attractive newborns are viewed as less intelligent, likeable, and good (Stephan & Langlois, 1984), and a less attractive child’s transgressions are both treated as more severe and more likely to be attributed to dispositional issues (Dion, 1972). In preschool, less attractive children are conferred lower social status (Vaughn & Langlois, 1983), and less attractive and athletic individuals experience more rejection by their peers in childhood and adolescence (Vannatta, Gartstein, Zeller, & Noll, 2009). In adulthood, less attractive individuals are discriminated against as potential mates and experience a more adverse mating environment (e.g., Buss, 1989, Buss and Barnes, 1986, Waynforth, 2001). This lifelong exclusion of individuals with less desirable morphological profiles may at least partly account for the cross-situational and temporal stability of neuroticism. If (i) humans possess psychological mechanisms that calibrate individuals’ neuroticism levels according to cues linked to their likelihood of being social excluded (Denissen & Penke, 2008), (ii) individuals’ likelihood of being socially excluded is stably linked to their desirability as a social partner, and (iii) their desirability as a relationship partner depends – even if only partly – on their profile of morphological attributes, then we should expect those individuals with less desirable morphological profiles to exhibit higher neuroticism across time and context. In short, we should expect species-typical, condition-dependent psychological adaptations to produce stable individual differences in personality (see also Buss & Greiling, 1999 for their discussion of “enduring situational evocation”).

Moreover, if this postulation that condition-dependent mechanisms produce stable personality differences as a consequence of taking as input condition-linked cues that inherently exhibit high levels of stability (e.g., physical attractiveness, formidability), then individuals might be able to infer others’ personality trait levels at above chance levels even at zero acquaintance (e.g., merely by looking at photograph of the individual). Indeed, there are multiple extant studies in support of this speculation. Shackelford and Larsen (1997) photographed participants, took measurements of participants’ facial symmetry from these photographs, and used these symmetry assessments to predict participants’ personality levels. They found a positive association between individuals’ facial symmetry and extraversion, a finding that has subsequently been replicated by Fink et al., 2005, Pound et al., 2007, the latter of whom provided convergent evidence for this finding by employing a distinct method. More recently, Holtzman, Augustine, and Senne (2011) expanded on this research by examining the relationship between individuals’ facial symmetry (as assessed from static photographs) and a broader range of personality dimensions. They found significant associations between facial symmetry and 44 dimensions of individual differences, ranging from the Big 5 factors to individual facet levels, and from socially aversive to pro-social traits. Collectively, this group of findings suggests that individuals’ personality level may be able to be inferred, with some degree of accuracy, merely from photographs of the individuals. More broadly, these findings are unambiguously consistent with the overarching proposal that condition-dependent psychological mechanisms can produce stable individual differences in personality as a consequence of calibrating their output according to stable, condition-linked variables.

The proposal of species-typical personality mechanisms is also ostensibly inconsistent with the heritability of personality traits (Penke et al., 2007). The heritability observed in personality traits would not necessarily be the pattern expected of species-typical adaptations; selection often leads to zero heritability as a consequence of favoring the genetic variant with the greatest replicative fitness to the point of the extinction of alternative genetic variants. Stated differently, directional selection in favor of the highest-fitness genetic variant reduces variation at that genetic locus until that allele becomes a virtually universal feature of a species’ genome. By definition, then, species-typical adaptations should exhibit virtually zero heritability at the genetic loci involved in the building of that adaptation.

A number of evolutionary biologists and geneticists have interpreted the heritability of manifest personality as evidence for variation in the genes involved in the construction of the psychological mechanisms responsible for producing personality (Nettle, 2006, Penke et al., 2007).

There are several known polymorphisms associated with individual differences in personality. For example, neuroticism is associated with the short allele of the DRD4 exon III polymorphism (Tochigi et al., 2006); the number of CAG repeats at the androgen receptor (AR) gene in men (Westberg et al., 2009); the serotonin transporter promoter 5-HTTLPR short allele (Sen, Burmeister, & Ghosh, 2004); and the GABA(A) Alpha 6 receptor Pro385Ser Pro allele (Sen et al., 2004). Extraversion is also associated with the number of CAG repeats at the AR locus (Lukaszewski and Roney, 2011, Westberg et al., 2009).

However, (1) models of personality based on genetic polymorphisms leave the lion’s share of inter-individual variation in personality unexplained (e.g., see Westberg et al., 2009), and (2) it is unclear whether these genetic polymorphisms are associated with differences in the psychological mechanisms responsible for personality, or with individual differences in the characteristics that serve as input cues into personality mechanisms and thereby influence their output.

If the psychological mechanisms that produce personality are designed to take as input cues associated with differential costs and benefits of alternative strategies, and at least a subset of these cues includes heritable, non-personality attributes such as physical attractiveness and strength (Lukaszewski & Roney, 2011), then we should expect personality to show heritability in behavioral genetics studies – as the output of species-typical, condition-dependent mechanisms.

Lukaszewski and Roney’s (2011) discovery of a link between the AR gene and extraversion illustrates this point clearly. If (i) individuals’ physical attractiveness and strength vary as a function of alternative genetic variants at the AR gene, (ii) the costs and benefits of extraversion vary as a function of individuals’ physical attractiveness and strength, and (iii) the humans possess psychological mechanisms that calibrate manifest extraversion levels as a function of cues linked to differential costs and benefits of alternative strategies on the introversion–extraversion continuum, then we would expect – on a priori grounds – extraversion to exhibit heritability – even as the output of universal psychological mechanisms (see Tooby & Cosmides, 1990, for their seminal discussion of this concept of indirect, or “reactive,” heritability).

Recent research asserts that a mutation-selection balance model – but not a condition-dependent model – can account for heritable variation in personality (Verweij et al., 2012). This research, however, betrays the misconception that these two models predict different patterns of heritable variation. Verweij et al. (2012) compare observed patterns of heritable variation to those expected under three models: balancing selection, selective neutrality, and mutation-selection balance. Under a balancing selection model, genetic variation underling personality differences is maintained by changing selection pressures across environments. For example, an allele that predisposes individuals to higher levels of neuroticism could be favored in environments in which relationship stability and fidelity are low, but would be outperformed by an allele associated with lower levels of neuroticism in environments in which relationship stability and fidelity are high. However, a condition-dependent mechanism that calibrated individuals’ neuroticism levels according to their desirability as a relationship partner, local levels of relationship stability and fidelity, or other inputs linked to differential costs and benefits of neuroticism would have been favored by selection over a mechanism insensitive to these cues. Given the known capacity of selection to craft such condition-dependent mechanisms, it is not highly plausible that heritable variation in personality can be predominantly explained by balancing selection.

A selective neutrality model, on the other hand, attempts to account for heritable variation in personality by proposing that genetic variants underlying personality differences are unaffected by selection because they are unrelated to fitness outcomes. However, given that personality differences are linked to fitness-relevant variables such as attractiveness (Lukaszewski & Roney, 2011), mating behavior (Zietsch, Verweij, Bailey, Wright, & Martin, 2010), and number of offspring produced (Jokela, Kivimaki, Elovainio, & Keltikangas-Jarvinen, 2009), selective neutrality is a rather implausible explanation for genetic variation in personality. It is perhaps unsurprising, then, that Verweij et al. (2012) conclude that neither selective neutrality nor balancing selection can account for observed patterns of heritable variation in personality. However, they erroneously categorize a condition-dependent model as being subsumed under selective neutrality, and conclude that extant data are consistent with mutation-selection balance.

This mischaracterization of a condition-dependent model leads to a critical oversight: the failure to realize that the patterns of genetic variation expected under a mutation-selection balance model and a condition-dependent model are identical. A mutation-selection balance model acknowledges that directional selection favors, among existing genetic variants, the allele associated with the most optimal trait levels, and eliminates alternative alleles. Although selection is constantly reducing genetic variation by purging these lower fitness alleles, this is an iterative process across generations that is inexorably accompanied by an influx of mutations that introduce new genetic variation. Verweij et al. (2012) are not unjustified in their conclusion that the observed heritable variation associated with personality is most consistent with mutation-selection balance. However, they jump to the unwarranted conclusion that the genetic loci under mutation-selection balance are specifically those that code for the mechanisms that produce personality.

An equally plausible alternative is that the loci under mutation-selection balance are those that code for the traits that serve as inputs to personality-producing mechanisms. Both of these perspectives represent mutation-selection balance models, so are thus equally consistent with extant data. However, the latter view is a condition-dependent model – a condition-dependent model is a mutation-selection balance model.

To illustrate this important point, let us consider polymorphism at the AR locus. Different alleles at the AR locus are associated with individual differences in attractiveness and strength, as well as individual differences in extraversion. If the AR gene is directly involved in the building of psychological mechanisms that calibrate extraversion levels, then we would not necessarily expect morphological correlates downstream from the gene, such as attractiveness and strength, to exhibit links to extraversion independent of the gene itself. Alternatively, if humans possess psychological adaptations designed to calibrate extraversion levels according to condition-based cues like attractiveness and strength, then we would expect links between these input cues and the output of these mechanisms. Strongly consistent with this condition-dependent model, individual differences in attractiveness and strength predict extraversion levels – above and beyond genetic polymorphism (Lukaszewski & Roney, 2011). This relationship between genetics, physical attractiveness, strength, and extraversion offers a clear empirical example of how species-typical, condition-dependent psychological mechanisms can produce heritable personality outcomes.

In sum, a condition-dependent model of personality proposes that if psychological mechanisms calibrate manifest personality trait levels according to condition-linked inputs such as physical attractiveness and strength, and these attributes exhibit heritability in behavioral genetic studies, then patterns of heritable variation in personality may actually reflect heritable variation in these inputs, rather than in the psychological mechanisms sensitive to those inputs.

There are two crucial points for personality researchers to note. First, there is no strong reason to believe that a single model must account for all heritable variation in personality. Human personality exhibits heritable variance on manifold dimensions. There may be different evolutionary origins for variance on each of these dimensions, and within each dimension, different models may account for distinct proportions of variance. Evidence suggesting an important role of mutation-selection balance in the evolution of human personality is not tantamount to evidence of an absence of other selective and non-selective forces at work.

Second, because a mutation-selection balance model and a condition-dependent model predict identical patterns of heritable variation, no pattern of heritable variation in personality could adjudicate between a mutation-selection balance model and a condition-dependent model. Instead, demonstrating functional design of the psychological mechanisms that produce personality differences ultimately will be necessary to discriminate between these alternatives. Testing for this discriminative evidence represents a key future research direction for an evolutionary approach to personality, as discussed in greater detail below.