The Emotional Foundations of Personality

by Kenneth L Davis

  King was also the lead author in a second factor-analytic study using all new chimpanzee subjects, combining forty-three nonzoo chimpanzees living in a large African sanctuary in the Republic of the Congo with seventy-four chimpanzees living in nine zoos in the United States and Australia. In this second study, the same forty-three descriptive adjectives were used to rate the chimpanzees, although, the adjectives for the African sanctuary chimpanzees were translated into French. Even with this heterogeneous group, including nonzoo chimpanzees with French-speaking raters, all six of the previously identified chimpanzee dimensions, including Conscientiousness, were obtained in this second study. However, the small Emotionality/Neuroticism and Openness factors did not meet statistical standards for replication, although most items loaded the same on these two factors in both studies (King, Weiss, & Farmer, 2005). The authors concluded they needed more items on these two factors. Yet, their structure of chimpanzee personality, as revealed by factor analysis, remained relatively constant across the two studies of different chimpanzees regardless of the different environments, which was consistent with a biological basis for personality and suggested that environmental effects on chimpanzee personalities may be relatively small.

  This smooth sailing was short-lived, as a shift to orangutan subjects failed to identify a Conscientiousness factor. Although there had been an earlier failure with gorillas (Gold & Maple, 1994), with two chimpanzee successes behind them Weiss, King, and Perkins (2006) tried their hand in another large-scale study with 152 zoo-housed orangutans. The same methodology and rating instrument—expanded with five additional adjectives—replicated all of the chimpanzee dimensions except Dependability/Conscientiousness and Openness. It was largely these two dimensions that blended into the fifth factor, which Weiss and colleagues labeled Intellect based on its highest loading item, “intelligent,” as well as “disciplined” from the Dependability dimension. Other adjectives identified with Dependability in the chimpanzee studies tended to load either on their Neuroticism factor, as with “impulsive,” “cautious,” “erratic,” and “unpredictable,” or on their Dominance factor, as with “persistent” and “reckless.” One hypothesis for the failure to find a Dependability/Conscientiousness dimension was that the capacity for regulating the expression of aggression and irritability was not as highly developed in the orangutan, and therefore, such traits were expressed more purely as Social Dominance.

  The authors also argued that the inability to replicate the Dependability factor in orangutans could suggest either that the Dependability/Conscientiousness capacity first emerged evolutionarily with chimpanzees or possibly that the conscientious regulation of especially negative primary emotions was less critical in the semisolitary orangutan. That Gosling and John (1999) were unable to identify a Conscientiousness factor in any species other than chimpanzees supported the argument that Conscientiousness first appeared in chimpanzees. However, the Darwinian principle of continuity as applied to mental functions in animals would argue for the possibility that aspects of Conscientiousness should be identifiable in orangutans and other even more distantly related species. Clearly, further work would be required to resolve such issues.

  Subsequently, two additional chimpanzee studies using the original forty-three King and Figueredo (1997) items fully replicated the first four factors (Dominance, Surgency/Extraversion, Dependability/Conscientiousness, and Agreeableness) but had similar problems demonstrating statistical congruence for the last two factors (Emotionality/Neuroticism and Openness). In the first case, Weiss, King, and Hopkins (2007) added 102 new zoo animals and 175 chimpanzees housed in the Yerkes National Primate Research Center. Later, Weiss and King also teamed up with a group of five Japanese scientists (Weiss et al., 2009) to examine the personalities of 146 chimpanzees living in Japanese zoos, research institutes, and a sanctuary. Despite the same problem of not being able to statistically replicate the Emotionality/Neuroticism and Openness factors found in the original King and Figueredo (1997) report, all four studies had demonstrated a statistically replicated Conscientiousness factor. In addition, the Japanese study, in which all ratings were completed using the forty-three-adjective scales translated into Japanese, also confirmed the finding from the Republic of the Congo report that chimpanzee personality ratings were not affected by the culture of the raters, which further underscored the biological foundation of personality.

  Others had pursued alternate paths to solving the chimpanzee personality puzzle. Working with Samuel Gosling, Hani Freeman wrote a dissertation that included the development and validation of a new chimpanzee personality rating scale. Her approach was to combine some adjectives from the King and Figueredo (1997) with additional trait descriptions that were unique to chimpanzees (Freeman et al., 2013). Her team tried to eliminate most redundancy and was able to reduce the list of traits to forty-one, which she used to collect ratings on ninety-nine chimpanzees at the University of Texas. In analyzing the data, several criteria finally led her to extract six factors. However, only the first five were interpretable: Reactivity/Undependability, Dominance, Extraversion, Openness, and Agreeableness. Because her Reactivity/Undependability dimension was very similar to King and Figuerado’s Dependability factor, she had once again demonstrated a Conscientiousness factor in chimpanzees. The remainder of her first five factors were also similar to the King and Figueredo dimensions, except that she lacked an Emotionality/Neuroticism factor. In the Freeman et al. study, the items that would have been expected to define a Neuroticism dimension loaded on either the Reactivity/Undependability or Dominance factors. So once again, the replication of King and Figueredo’s human-like Big Five plus Dominance chimpanzee personality structure remained illusive.

  This story of searching for distinct personality dimensions in chimpanzees using factor analysis again demonstrates the vicissitudes of factor analysis in defining a biologically based taxonomy of personality. The first four of these chimpanzee studies suggested that results more congruent with the Big Five model would likely have been obtained if they had started with a larger, more representative pool of adjective descriptors. Even though that may be true, our position is that it is time to accept the limits of factor analysis and recognize that the next round of progress in understanding the dynamics of mammalian personality will likely emerge from an affective neuroscience analysis of how the mammalian brain responds—especially how the subcortical brain responds—to significant life events. A major theme of this book is that the remarkable homology of personality traits across mammalian species has its origin in the subcortical emotional systems of the brain, and a better understanding of these systems will be essential for discovery of the evolutionary sources of personality structures, perhaps even the sources of “human nature” from emotional pathologies to self-actualization. Indeed, this has been supported recently with Affective Neuroscience Personality Scales studies of human diverse human personality disorders (Karterud et al., 2016), discussed further in Chapter 18.

  THE ANALYSIS OF ANOTHER PRIMATE

  The effort to “discover” the human Big Five in the personalities of nonhuman primates took an interesting turn when Alexander Weiss of the University of Edinburg joined with F. Blake Morton and an international research team to study brown capuchin monkeys (Sapajus apella), a New World species that diverged from the human line a bit further down the evolutionary tree but one, like the Old World great apes of Africa and Asia, that has a large brain for its body size (Morton et al., 2013). Another important consideration in the project was using a revised King and Figeredo (1997) rating questionnaire called the Hominoid Personality Questionnaire, which had now been expanded from the original forty-three to fifty-four adjectives and that allowed them to better capture the Emotionality, Openness, and Conscientiousness dimensions (Weiss et al., 2009).

  In this project, 127 captive brown capuchin monkeys from six sites were rated. The authors determined that their data did fit a five-factor solution, but they added further conceptual complexit
y to the discussion. Briefly, Morton et al. (2013) labeled their five factors Assertiveness, Openness, Neuroticism, Sociability, and Attentiveness, while acknowledging that their Assertiveness factor closely resembled King and Figueredo’s (1997) original chimpanzee Dominance factor, their Openness factor was similar to King and Figueredo’s Openness factor but also resembled the orangutan Extraversion factor (Weiss et al., 2006), their Neuroticism factor closely resembled the King and Figueredo’s Emotionality dimension, their Sociability factor shared elements of King and Figueredo’s Agreeableness and Surgency/Extraversion factors, and their Attentiveness factor most closely resembled the Conscientiousness dimension found in Japanese chimpanzees rated on the full Hominoid Personality Questionnaire by Weiss et al. (2009).

  Despite these complexities, by using an expanded set of rating items, the Morton group had remarkably demonstrated Conscientiousness in capuchin monkeys, an evolutionarily older species than chimpanzees, even though earlier efforts to identify a Conscientiousness factor in gorillas and orangutans had failed to do so. This demonstration is important from a Darwinian continuity perspective and may open the way for others to look for elements of Conscientiousness even in nonprimate species. However, illustrating the difficulties of factor analysis, they failed to fully replicate the King and Figueredo (1997) Dominance factor and their Sociability factor structure blended elements of the Big Five Agreeableness and Extraversion dimensions.

  BACK TO BASICS: ALTERNATIVE APPROACHES

  Diane Dutton of Liverpool Hope University in the United Kingdom also developed a new personality rating scale for chimpanzees, but she started from the ground up, so to speak. That is, instead of building on an existing instrument, she used the expert knowledge of people working closely with chimpanzees and a technique developed by the American psychologist George Kelly (1955) to build a new set of rating items. In short, Dutton asked raters who had worked extensively with a group of twenty-four chimpanzees housed in the Chester Zoo to independently generate bipolar descriptors of chimpanzees, such as “aggressive-submissive,” to describe differences in pairs of chimpanzee they had worked with (Dutton, Clark, & Dickins, 1997). Using such descriptors generated by expert chimpanzee observers, she intended to devise a rating scale that would be more naturalistic and perhaps more appropriate for rating chimpanzee personalities. Eventually, she produced a forty-six-item personality rating scale for chimpanzees and coordinated a study of seventy-five captive chimpanzees living in seven zoological parks in Europe, the United States, and South Africa (Dutton, 2008).

  Using factor analysis, she extracted five factors that she labeled Agreeableness, Dominance, Neuroticism, Extraversion, and Intellect. Her factors included many similarities with our Affective Neuroscience Personality Scales. For example, her Agreeableness dimension included items such as “protective toward infants” and “interacts with infants and juveniles” plus “reconciles others” and “reassures others,” all of which are consistent with the CARE system. Her Neuroticism dimension included “nervous,” “easily frightened,” “fearful for no apparent reason,” and “anxious,” all of which suggest the FEAR system. The Extraversion dimension included “playful” and not “withdrawn,” which fit with the PLAY system. The Intellect dimension included SEEKING concepts such as “inquisitive” and “investigative.” With the Dominance dimension having its highest loading on “aggressive,” which is suggestive of the RAGE/Anger system, she had scales representing each of the personality-related blue ribbon emotional brain systems except PANIC/Sadness. In other words, she may have independently discovered the emotional foundations of personality in our closest evolutionary relative.

  Interestingly, Dutton did not derive a Conscientiousness scale, an omission we touch on again shortly. However, in using rating items grounded in actual observations of chimpanzee natural behaviors, Dutton seems to have confirmed a theme we emphasize in this book: Relying on measures as close to the primary expression of emotional behaviors as possible is likely to keep researchers close to basic neural-system taxonomies of personality and keep them from deviating far from the foundational processes that help constitute even human personality.

  “CONSCIENTIOUSNESS” CONSIDERATIONS

  Conscientiousness in Big Five terms applied to humans has typically been described with such adjective pairs as “organized versus disorganized,” “responsible versus irresponsible,” “reliable versus undependable,” and “thorough versus careless” (Goldberg, 1992) but has been difficult to identify in animal studies perhaps because of the abstract nature of these terms. By contrast, in demonstrating the first appearance of a Dependability/Conscientious factor in chimpanzees, King and Figueredo (1997) selected adjectives from Goldberg’s (1990) Conscientiousness factor descriptors that seem eminently concrete and observable, such as “impulsive,” “reckless,” “erratic,” and “disorganized”—each of which Hofstee et al. (1992) showed had its strongest loadings on Conscientiousness—plus included overtones of “defiant” and “aggressive,” which might speak to the importance of regulating social dominance because it plays such a key role in chimpanzee social life. Indeed, all five chimpanzee studies reviewed above that used the forty-three King and Figueredo (1997) adjectives identified these six adjectives on their Dependability/Conscientiousness factors to some degree.

  But why was Dutton not able to find a Conscientiousness factor in her chimpanzee data? Perhaps because she unfortunately included only two clear Conscientiousness items in her forty-six statements, relying on “persistent” and “unpredictable”, which may not have had enough power by themselves to emerge as a sixth factor measuring Conscientiousness. King and Figueredo (1997) had included eight Big Five Conscientiousness items, four of which loaded most strongly on their Dependability factor, and six more Conscientiousness items were added to the revised fifty-four-item version of the Hominoid Personality Questionnaire that was used to identify a Conscientiousness factor in brown capuchin monkeys. Thus, it is likely that if the number of Dutton’s conscientiousness-type items had been expanded, her sample would have revealed a Conscientiousness dimension as well. Such are the vagaries of the factor-analytic world.

  In any case, this first clear demonstration of the Big Five human personality dimension in a nonhuman animal supported Darwin’s continuity of species principle. Indeed, having demonstrated this trait in chimpanzees and then brown capuchin monkeys opens up the possibility that rudimentary forms of conscientiousness might be observable in other mammals, such as dogs and rats. Coming back to the idea that more concrete and observable descriptors were necessary to reveal conscientiousness in chimpanzees hints that even more direct interventions and measures might be necessary to identify this cognitive, regulatory dimension in other mammals, especially in even more distantly related mammals.

  An intriguing neuroscience demonstration of what may be considered evidence of conscientiousness in rats was reported by J. P. C. de Bruin in 1990. He and other researchers had previously determined that lesions to the orbital prefrontal cortex increased social-agonistic interactions in rats, that is, made them more aggressive, which implied that the orbital prefrontal cortex exerted an inhibitory control over some brain region supporting the aggressive behavior. Research suggested the brain region was likely the hypothalamus. With this insight, de Bruin implanted electrodes in the hypothalamus of rats and observed whether aggression directed toward another rat was elicited by a small (40–60 microamp) current. Then, using a second prefrontal electrode, he concurrently stimulated the hypothalamus and various locations in the prefrontal cortex and determined that the greatest inhibition of hypothalamic-elicited aggression occurred when the prefrontal electrode was placed in a rat’s orbital prefrontal cortex. Both the lesion data and concurrent electrical stimulation data were consistent with the conclusion that the rat orbital prefrontal cortex “exerts an inhibitory control over hypothalamic sites from which aggression can be elicited by electrical stimulation” (de Bruin, 1990, p. 492). Could this inhibition
of aggressive behavior from electrodes placed in the orbital prefrontal cortex represent an induced case of conscientiousness in rats? Clearly, there is a brain mechanism in place for inhibiting rat social aggressive behavior likely stemming from the RAGE/Anger system. How this inhibitory control might be learned and expressed in rat social behavior remains to be determined. It would also be important to link orbital prefrontal activity to “conscientious” behavior in other species. In any case, this intriguing demonstration hints of a rudimentary chimpanzee-like conscientiousness inhibiting social aggression in a species much less well endowed with cortical capacity than the brown capuchin monkey.

  SOCIAL DOMINANCE

  The emergence of a prominent Dominance factor in chimpanzees could be explained in part by “the pervasive role of dominance and power-related activities in the social order of wild chimpanzees” and “may be further enhanced in zoo habitats where dispersal is not possible” (King & Figueredo, 1997, p. 268). However, some have emphasized dominance as a basic human personality dimension. The Hogan Personality Inventory (Hogan & Hogan, 2007) similarly expands the Big Five with a sixth scale labeled Ambition that is characterized by “status” and “competitiveness,” and de Raad (1999) has also emphasized dominance as a human interpersonal personality dimension. Donne van der Westhuizen, a student at the University of Cape Town in South Africa working with Mark Solms, has written and published a dissertation exploring the possibility that social dominance might qualify as a primary emotion. Even though she reviewed a great deal of biological support underlying dominant and territorial behavior, her final conclusion was that there was insufficient evidence at present to classify social dominance as a primary emotion (van der Westhuizen & Solms, 2015), for dominance needs to be learned by actual social interactions. This does not mean that there are not specific brain systems that promote dominance (e.g., perhaps RAGE, LUST, and social SEEKING). In short, even though dominance is widespread throughout the animal world (including social relations in practically all mammals), we would argue that complex social learning is essential for it to manifest in the social world, and that early social PLAY is an important factor for finding one’s place in social structures (Panksepp, Jalowiec, DeEskinazi, & Bishop, 1985). In any event, much work remains to be done in the area of social dominance in mammals, including the manner of its emergence as a key feature of primate social life.