The Emotional Foundations of Personality

by Kenneth L Davis

  What Is the Meaning of the Big Five Personality Model?

  A big question remained unanswered: What was the meaning of the Big Five personality model? Gerard Saucier and Lewis Goldberg (1996) argued that the Big Five model was phenotypic rather than genotypic. Phenotypes refer to observable characteristics (Cattell’s idea of surface traits); genotypes refer to the underlying biological causes (Cattell’s idea of source traits). Thus, the lexical origins of the Big Five model provided excellent descriptions but not deeper functional or psychobiological explanations. In their own words, “The lexical perspective leads to data that need explaining, not necessarily to the modes of explanation” (Saucier & Goldberg, 1996, p. 25). Saucier and Goldberg went so far as to suggest that the Big Five model should be called an “attribute” model rather than a “trait” model (p. 25). The term trait brought too much baggage, with assumptions about stability and genetic origin that need to be verified and not assumed. The term attribute was more neutral and made no assumptions about the cause or origin of the described temperamental characteristics.

  Also, Goldberg (1993) stated that the Big Five were “dimensional” as opposed to “categorical” in nature. In other words, they were gradients of existence rather than distinct categories such as different kinds of fruits and vegetables. There is general agreement that levels of these five dimensions can be measured with various instruments but not on what constitutes each of these dimensions. Plus, it has not been well determined what limits there are to the categories of variables used in their measurements. For example, to what extent should evaluative terms, physical characteristics, or world views be used, and how might they alter the model that was based on descriptive adjectives? (We discuss this further in Chapter 13.)

  Goldberg also preferred to use the term Big Five for the personality attribute model, which is derived from lexical data and remains descriptive rather than explanatory. He cited John and Robins (1993) for suggesting that the term Five-Factor Model (FFM) is an explanatory model that assumes the five factors correspond to specific biological origins. However, it seems unlikely that even the scientific community will be able to agree on this subtle distinction in terminology. In general, the term Big Five attribute model seems more neutral and open to modification than FFM, but we use both terms, guided by the preferences and usage of the various investigators we cite.

  For Saucier and Goldberg, the Big Five are “dimensions of perceived personality” (1996, p. 42). It is not a theory; rather it is more a conceptual taxonomy of convenience for mapping personality variables. In any event, the dimensions constitute robust replicable findings, which can provide the basis for future research. The Big Five approaches can inform those who strive to understand personality at deeper neuropsychological levels but represent only a beginning rather than any statement about the underlying neural or psychological structures of personality based on relevant functional systems of the brain and mind.

  THE CATTELL TRADITION: HOW WILL THE UNDERLYING BIOLOGY BE REVEALED?

  In the tradition of Raymond Cattell’s dreams, there are some like Robert McCrae and Paul Costa who see the FFM as a new empirically based personality theory that is supplanting the older more clinically based theories. They argue that the lexically derived FFM provides some support for human rationality, which stands in contrast to radical behaviorism and the psychoanalytic theories that focus on diverse unconscious processes (McCrae & Costa, 1996). Specifically, they view the factor-analytically derived FFM as having revealed the wisdom encoded in our spoken language. In other words, personality features that were important to daily social life have “trait terms” associated with them. Like Cattell, they believed that factor analysis could statistically extract the “latent” personality features embedded in language and thereby reveal the underlying biological source traits. However, this has yet to be achieved, for that requires kinds of causal/constitutive (rather than just correlational) brain research that, despite the neuroscience revolution, barely exists at the present time.

  Their faith in the reality of these language indexed dimensions led McCrae and Costa (1996) to promote the FFM as an answer to the age-old problem of parsing the foundations of human personality. Thereby, they have clearly moved beyond the simply surface descriptive phase of inquiry, into the beginning of neurogenetic and neurobiological explanatory phases, without, however, having provided any roadmap or empirical clues on how such levels of explanation might be achieved. Outside of genetic heritability (Loehlin, 1992), supporting biological evidence for the FFM has come largely from modern brain neuroimaging studies, the specifics of which have sometimes been inconclusive or contradictory (as we discuss further in Chapter 16). However, brain regions frequently implicated have been cortical structures, such as the prefrontal areas, and basal ganglia structures, such as the amygdala (Canli, 2006; Mincic, 2015), which together would focus on tertiary and secondary rather than primary processes.

  Researchers taking a more direct and less inferential approach are those pursuing linkages between brain, behavior, genetic, and psychological variables (Montag et al., 2013; Weaver, 2014; Weissman, Naidu, & Bjornsson, 2014). However, with the genetic-linkage analyses, which are currently proceeding ever more rapidly and efficiently (because of the vast armamentarium of modern genetic tools), the issue of how the sources of heritability will need to be linked to actual brain systems is just beginning to be addressed (e.g., Montag et al., 2013). Until the emergence of cross-species emotional neuroscience (Panksepp 1982, 1998a), the specific BrainMind emotional origins of these five factors had remained elusive.

  In sum, the personality structure erected by McCrae and Costa has so far remained a surface analysis that rests largely on their trust in the factor analysis of linguistic descriptors, with the conviction that the more fundamental evidence will emerge with future research. Obviously, that will be a neuropsychologically complex story that will require a synthesis of critical lines of human and animal research. Although not yet widely recognized, animal brain research is critical for deep progress. At the very least, it will require a scientific understanding of the basic evolutionarily constructed “Darwinian” emotions that are best illuminated by cross-mammalian brain emotion studies. How those neuromental functions interact with the higher cognitive apparatus (especially language), at this point in scientific history, most critically requires deeper and deeper understanding of brain research, particularly those aspects that are shared homologously (by shared genetic factors) between humans and other animals.

  Five Factors and Looking Forward to the Neurobiological Foundations of Personality

  Factor analysis is a wonderful tool. However, like someone once said, if your only tool is a hammer, every problem looks like a nail. If factor analysis is the only tool for exploring personality, are we limited to explaining personality with tightly intercorrelating clusters of items that lend themselves to factor analysis? Or should factor analysis be used as the tool it really is, to simply help guide systematic research, with the caveat that there may not always be a one-to-one correspondence between statistically derived factors, based on the analysis of language, and biological systems that actually control our experiences—our diverse feelings, perceptions, and thoughts?

  Like Cattell, Costa and McCrae seem to have leaped far ahead of the evidence, postulating inherited biologically based dimensions relying mainly on statistical criteria. In addition to the five factors measured by their revised NEO-PI personality assessment (Costa & McCrae, 1985), they included six narrow “facets” for each of the five higher level dimensions, reporting thirty facets altogether, and have suggested that the “psychological literature . . . provides a rich source of conceptualizations for specific facets of each of the five domains” (McCrae & Costa, 1996, p. 62). While there may easily be thirty facets operating at the tertiary level of the human personality, we believe it is unlikely that the neocortex creates new facets on its own without support from the subcortical emotional core for the “feelings” of
existence that we share with the other animals.

  Thus, is it really possible to “carve nature at its joints” statistically without examining the organism itself, as is beginning to be done with cross-species affective neuroscience approaches? We propose that the foundational neuroscience of the mammalian brain is revealing the primal emotional processes that need to be understood in order to help personality theory take the next big step toward understanding the neurobiological foundations of some of the most important affective underpinnings of human personality. We believe Goldberg was right: The Big Five is best thought of as a phenotypic, descriptive system rather than the genotypic, explanatory theory some would like it to be. We believe that affective neuroscience strategies have already provided some of the likely neurobiological foundations of personality that are currently missing from the five statistically derived personality dimensions.

  It was probably inevitable that the lexical hypothesis would produce overly abstract traits. Is language itself not an abstraction? Yet, if Charles Darwin, Paul MacLean, and all the comparative affective neuroscientists that followed have taught us anything, it is that emotions appeared in the animal world long before spoken language. Why should we be surprised that our tertiary-level words, when statistically clustered, sometimes represent only metaphorical approximations of the primary-level capacities with which we confront the world and thereby experience evolved emotional qualia in various life contexts?

  It may be that the Big Five attribute model does a reasonably good job with Extraversion (gregarious “hanging out,” an elaboration of the PLAY/Joy emotional brain system), Agreeableness (kindness and warmth, the affects of CARE/Nurturance on the positive pole), the regulatory trait of Conscientiousness, and even Openness to Experience (curiosity and enthusiasm emerging from the SEEKING system). However, language and the Big Five seem quite inadequate when describing our capacity to experience powerful negative emotions, which are commonly transitory but which can become crystallized as affective ways of being in the world: RAGE/Anger, FEAR/Anxiety, and PANIC/Sadness. At present, the Big Five/FFM simply lumps these three affectively negative emotions together as forms of Emotional Instability. It seems unacceptable to relegate anger, fear, and separation distress as mere FFM “facets” of Neuroticism/(low) Emotional Stability, especially given the suffering (antecedent to diverse psychopathologies) created when any one of them becomes overly sensitized or imbalanced. Surely our anger and frustration, our fears and anxieties, and our social pains and distresses deserve as rich an assessment, as primal personality dimensions, as our various positive emotional systems. To emphasize their subcortical and prelanguage origins, all of these primal emotions in animal models survive removal of the upper brain (neodecortication).

  The reason for the factor-analytic clustering of the negative emotions into a single factor is not clear. It could have to do with multiple negative emotions occurring simultaneously and not always coming in neat, isolated bundles. A good example comes from Don Klein’s research on panic attacks (Preter & Klein, 2008). Even after medication (e.g., the tricyclic antidepressant imipramine) had stopped the occurrence of the panic attacks, his patient reported little relief (Klein, 1981). It was only after the anticipatory “fear” of having another panic attack was eliminated that the patient felt a complete sense of relief—of having been “cured” of those sudden lapses of a sense of social security that good social attachments provide. A similar example came from a friend who said that she was afraid her mother was going to die, which combined the fear about how difficult life would be without her mother’s help with the anticipation of the separation distress she would also experience without her mother’s companionship. It may be that two negative emotions become combined in this way, with the negative emotions augmenting each other and building an overall tertiary (upper brain), negativistic state of mind, which can be difficult to sort out later in rational, descriptive terms. In other words, the neocortex, which cannot generate emotional feelings by itself, may operate in a more general dimensional (positive vs. negative) affective way, while lower BrainMind regions have more categorical emotional systems. It may also be that humans have various strong negative emotional experiences before we are old enough to possess adequate language and the conscious awareness to express the distinct feelings clearly to ourselves and others.

  These are among many possibilities, but regardless, we are left with a major limitation of the lexically and hence factor-analytically derived Big Five/FFM as an overall theory of personality. While it is apparent that the unique primary emotions of RAGE/Anger, FEAR/Anxiety, and PANIC/Sadness are supported by some overlapping systems (e.g., for acetylcholine, norepinephrine, and serotonin neuromodulators), they are also functionally separable brain systems (using various neuropeptides), which have differentiable effects on our behaviors and feelings and thereby promote the solidification of different personality traits. As we have already seen, each generates distinct psychopathological trajectories, such as explosive disorders, specific phobias, generalized anxiety, depression, and panic attacks, the fear of which may promote chronic anticipatory anxiety. Thus, these powerful and discrete primary negative emotions are unsatisfactorily lumped together in the Big Five as a lack of Emotional Stability. We now need new brain-based organizational visions such as affective neuroscience provides, to generate diverse objective criteria for the causal parsing of some of Cattell’s source traits and thereby bringing more neurobiological specificity and conceptual order to the Big Five model.

  We have finally reached a time when neuroscientific knowledge is rich enough to begin taking the next major steps in understanding the mental forces that govern personality development. Thus, it may be best to appreciate Big Five theorizing for its highly rigorous level of descriptive utility than to try to make it into a major explanatory personality theory in its own right. Affective neuroscience provides a clear trajectory for taking new steps in that direction and thereby expanding the Big Five model into a more comprehensive coherent psychobiological representation of human personality, hopefully, one that can link more effectively towards understanding psychiatric disorders. In sum, providing clearer visions of our genetically derived, brain-emotional “tools for living” may allow personality theories to become ever more closely linked to variations in our genetic heritage. Furthermore, our emerging understanding of fundamental affective-emotional brain systems holds the potential of providing a new set of objective standards for carving the foundational nature of personality at its affective joints and thereby further illuminating what it means to be human.

  CHAPTER 13

  The Clarities and Confusions of the Big Five

  We have met the enemy and he is us.

  —Pogo (Walt Kelly, Pogo cartoonist)7

  BIG FIVE PERSONALITY THEORY, as discussed in preceding chapters, is clearly a top-down, verbal, tertiary-level approach to understanding personality. It is openly based on the lexical hypothesis that all important human behavior has already been well described in human language, especially adjectives, and that personality can be fathomed by the clustering of these descriptors of human mental states. This chapter deals with developments in this top-down statistical approach to personality since Lewis Goldberg’s research that effectively established the Big Five as the most robust human personality model (Goldberg, 1990, 1992).

  Given the statistical and language basis of the top-down generated Big Five, it is perhaps not surprising that variants of the Big Five model would eventually emerge. Several of these alternates are briefly reviewed here, to show that there are chinks in what once appeared to be the solid armor of the Big Five, and that increasing numbers of top-down researchers no longer see the Big Five model as a final solution to the human “personality problem.” What seemed to be a robust, almost unassailable theory of personality—the Big Five—has become more like a ball of yarn that is gradually unrolling. Challenging the Big Five’s hegemony are claims for a Big Two, a Big Three, two different Big Six m
odels, a Big Seven, and even a Big One (Saucier & Srivastava, 2015).

  Even though this development may seem a bit bizarre, since the era of Raymond Cattell (see Chapter 12) and Hans Eysenck (see Chapter 14), there have been other personality models with more than or fewer than five dimensions (e.g., Douglas Jackson’s Personality Research Form, and Robert Cloninger’s Tridimensional Personality Questionnaire described later in Chapter 14). What is different about the studies featured here is that, like Goldberg, these researchers also used adjectives to generate their competing personality models. However, as we describe here, these new personality models were often “discovered” by using variations in the sets of trait-descriptive terms used by Goldberg in his seminal Big Five work.

  As one might imagine, given each of the various tertiary top-down approaches, small differences in language and culture can come strongly into play. Although we do not take sides on these controversies, we do note that all variants of the Big Five have led to conclusions based on statistically obtained clusters in adjectival measures of human personality. From the language analysis perspective, it is pretty obvious that trait-descriptive adjectives and nouns, including words that described highly consistent patterns, were sustained in all human languages. Thus, these approaches are by no means theory-free. They are simply capitalizing on how different human groups at different times used such descriptors. Our only aim here is to briefly review what, from our perspective, are the more significant new challenges confronting the once monolithic five-factor solution to human personality assessment.