Yet, it remained difficult for psychologists to believe that only five factors were sufficient to account for the behavioral complexity observed in humans. Digman himself (1996) confessed that he assumed that more than five factors would be required to account for the childhood personalities he was studying and that both he and personality psychologist Warren Norman had fallen prey to this temptation in the 1960s. In short, how could Cattell hope to account for the seemingly countless variations of human personality with just a few factors?
In any case, Cattell, the pioneer, didn’t know how many factors to expect. This was exactly the problem he was exploring as he attempted to define personality space using his factor-analytic approach. In his Sixteen Personality Factor Questionnaire (16PF), which became his most widely used and enduring standardized psychological test, he settled on sixteen scales (see Table 11.1), and he had excluded several additional possible scales to end up with his final sixteen. Further, Cattell was not alone in creating complex psychological test models. Harrison Gough later settled on twenty scales for his California Psychological Inventory, the Guilford-Zimmerman Temperament Inventory had ten scales, and Douglas Jackson’s Personality Research Form had twenty scales, one for each of Henry Murray’s twenty basic needs (see Chapter 5). Even the relatively revised Neuroticism Extraversion Openness Personality Inventory (NEO-PI), which is marketed as a Five-Factor Model personality instrument, has thirty scales: six separate facets for each of the five main factors.
A FOUNDATIONAL ISSUE: THE LEXICAL HYPOTHESIS
The lexical hypothesis embraces a faith that the spoken language encompassed a profound wisdom: the ability to describe and communicate all human interpersonal behavior that had been important for survival (for a review, see Saucier & Goldberg, 1996). In reporting his derivation of sixty key variables from the 4,503 trait terms, Cattell wrote that “all aspects of human personality which are or have been of importance, interest, or utility have already become recorded in the substance of language” (1943, p. 483). He later similarly wrote that “by the pressure of urgent necessity, every aspect of one human being’s behavior that is likely to affect another has come to be handled by some verbal symbol” (1957, p. 71). Furthermore, for Cattell, factor analysis was the tool to extract this “latent” understanding of the human experience that was embedded in spoken language. Parenthetically, all this may be true about tertiary-process cognitive elaborations on basic survival issues, but our view is that the primary-process emotional survival themes need to be explicitly addressed for personality scales to be optimally useful to psychiatrists and other mental health professionals.
The scientific problem that Cattell and others hoped the lexical hypothesis with factor analysis would resolve was how many distinct categories of behavior one needed to adequately describe human personality. However, Cattell also dreamed that his factor-analytically derived “source” traits would represent the fundamental “dynamic causes” of human behavior. Yet, as we have already pointed out, this marvelous statistical descriptive system has limitations when it comes to arriving at fundamental primal categories of emotional tendencies. With this statistical tool it is difficult to discern where psychologically relevant primary emotional brain systems end and higher-order tertiary personality elaborations begin. Both of these problems—identifying distinct categories of behavior and identifying the fundamental causes of personality differences—are personality issues that perhaps affective neuroscience can address more accurately than statistical factor analysis. And as soon as one identifies the foundations of personality systems useful for understanding emotional problems, often conceptualized as psychiatric disorders, one has the possibility of bridging basic brain research and clinically relevant human problems in new ways, as we describe in Chapter 18.
Cattell’s Sixteen Personality Factors
Cattell ventured into the personality world armed with a rich lexical hypothesis and Thurstone’s “improved” factor-analytic methods. Cattell found twelve dimensions that he could observe readily using both self-report questionnaire data (Q-data) and observer ratings (L-data). However, there were four personality dimensions he was only able to find using Q-data, which he then named Q1 through Q4. Cattell published the 16 PF, short for 16 Personality Factors, as summarized in Table 11.1—sometimes using very imaginative “cryptic” terms to describe his scales.
The Underlying Complexity of the Big Five: 16PF Source Traits
Cattell further factor analyzed his sixteen primary factors to reduce the number to eight second-order personality scales (not to be confused with our secondary-process in the Nested BrainMind Hierarchy presented in Chapter 5). Cattell later reported a second-order solution that further simplified personality profiling to five second-order factors that fit the Big Five personality model. While Cattell recognized the value of recognizing these broad themes, he consistently maintained that “greater accuracy of prediction is necessarily obtained by using primaries . . . [and] the psychologist is strongly advised to keep to the full spectrum of primaries” (Cattell, Eber, & Tatsuoka, 1970, p. 127). Again, Cattell’s use of the term primaries referred to his basic sixteen factors and should not be confused with primary-process emotions.
Cattell’s factor-analytic successors further refined his factor analyses into a clearer Big Five model, which is the subject of the next chapter. However, in closing this chapter, we take a glimpse at Cattell’s factoring of more objective personality measures, which he labeled “Testing-data” or “T-data.”
Instinctual Drives and Ergs
Cattell probed closer to the biological basis of personality, attempting to explore more “innate drives.” The term instinct was very much out of favor in Cattell’s time—and still remains out of favor to this day among many behavioral neuroscientists. So, for innate drives Cattell preferred to use the term erg, a unit of energy, a term he borrowed from physics. As explained by Cattell, an erg was “an innate psycho-physical disposition which permits its possessor to acquire reactivity (attention, recognition) to certain classes of objects more readily than others, to experience a specific emotion in regard to them, and to start on a course of action which ceases more completely at a certain specific goal activity than at any other” (1950, p. 199). This complex definition generally agrees with the affective neuroscience definition of primary emotions and aligns with the simple “APT to ACT” acronym, which summarizes four components of primary emotions: Affect, Perception, Thought, and ACtion Tendency. Cattell’s “specific emotion” would generally correspond to affect; “attention, recognition” to perception and thought; and “a course of action” to action tendency. Thus, Cattell’s erg components correspond closely to Panksepp’s primary emotions, and for McDougall they would define emotional instincts (see Chapter 4). As is common in science, it is often hard to translate between different terminologies.
Table 11.1. Brief Descriptions of Cattell’s Primary Source Traits Published in the 16PF
Factor Low Score Descriptions High Score Descriptions
A Sizothymia: Reserved Affectothymia: Outgoing
B Intelligence: Dull Intelligence: Bright
C Low ego strength: Easily upset High ego strength: Emotionally stable
E Submissiveness: Humble, easily led Dominance: Assertive, aggressive
F Desurgency: Sober, serious Surgency: Enthusiastic
G Weak superego strength: Expedient Strong superego strength: Conscientious
H Threctia: Shy, timid Parmia: Socially bold
I Harria: Tough-minded, realistic Premsia: Tender-minded, sensitive
L Alaxia: Trusting Protension: Suspicious
M Praxernia: Practical Autia: Imaginative
N Artlessness: Forthright, genuine Shrewdness: Astute, polished
O Untroubled adequacy: Self-assured Guilt proneness: Insecure
Q1 Conservativism: Traditional ideas Radicalism: Liberal ideas
Q2 Group adherence: Sound follower Self-sufficiency: Prefers own decisions
Q
3 Low self-sentiment: Undisciplined High self-sentiment: Controlled
Q4 Low ergic tension: Relaxed High ergic tension: Frustrated
Adapted from Cattell, H. B. (1989).
Cattell measured ergs using Testing data (T-data), which was a third kind of more direct objective data, in contrast to subjective data estimated through self-report (Q-data) or observer ratings (L-data). Examples of T-data were galvanic skin responses, word fluency tests, idea fluency tests, reaction times, and the capacity to find a particular shape embedded in a more complex drawing.
Cattell approached the identification of ergs in the same way he had come to approach other psychological questions: relying on factor analysis, although he cautioned that, “As to the actual number and naming of ergs in human beings, another decade of systematic research . . . might reasonably be expected to give us a complete picture” (Cattell & Child, 1975, p. 37). But by 1975 he concluded that adequate factor-analytic research had consistently identified ten ergs, with a few other proposed candidates. He also noted that his list was closer to McDougall’s list than to those of Freud or Murray (Cattell & Child, 1975). Those ten, along with a yet-to-be-well-defined “Laughter” erg, are listed in Table 11.2 and compared to Panksepp’s blue ribbon emotions. It should be noted that these comparisons are based on definitions, not specific data.
Table 11.2. A Comparison of Cattell’s Ergs and Panksepp’s Primary Emotions
Cattell’s Hypothesized Ergs Panksepp’s Affect or Emotion
Goal Title Emotion
Food-seeking Hunger HUNGER (homeostatic state, not central to personality)
Mating Sex LUST (marginally central to personality)
Gregariousness Loneliness PANIC/Sadness
Parental Pity CARE
Exploration Curiosity SEEKING
Escape to security Fear FEAR
Self-assertion Pride Power/dominance (not well documented as a brain system)
Narcissistic sex Sensuousness ? (a possible higher dimension of LUST)
Pugnacity Anger RAGE
Acquisitiveness Greed ? (a possible higher dimension of SEEKING expressed in hoarding tendencies)
Laughter (proposed) Amusement PLAY
Adapted from Cattell and Child (1975, p. 40) and Panksepp (1998a).
The agreement between the two lists is striking—yet the affective neuroscience analysis emerged from basic brain research in animals, without any linkage to or awareness of Cattell’s ergs. Moreover, Cattell considered his factor-analytically derived ergs to be the main mammalian drives although his research had no connection to neurobiological brain research. They were conceived by Cattell to be in the “motivational” domain, being basic biological impulses or urges that constituted the energizing forces of personality.
In retrospect, it seems likely that Cattell’s T-data were very substantially tapping into what affective neuroscience approaches would consider primary-level emotions and motivations, along with HUNGER, an affective bodily need (homeostatic feeling). Cattell’s T-data seem to be accessing the primary-process and secondary-process of the Nested BrainMind Hierarchy, with the Q- and L-data likely measuring language-derived traits at the tertiary cognitively elaborated level. In other words, much personality assessment using self-report (Q-data, i.e., language) may be describing personality characteristics emerging at Panksepp’s tertiary-process level, which remains difficult to link empirically to the primary-process level of affective experience. However, Cattell’s ergs, estimated from objective T-data, seem to align quite readily with the primary genetically endowed emotions, which again confirms that relying on measures as close to the primary expression of emotional behaviors as possible is likely to keep researchers close to the basic foundational processes that help constitute human personality.
The similarity of Cattell’s ergs with Panksepp’s primary emotions continues a pattern of conceptual consistency from Darwin and McDougall to the neuroscientific work of Panksepp. Darwin and McDougall added objectivity to their selection of instinctual emotions by requiring the emotions to be observable in animals as well as humans. Panksepp used several objective criteria in making his selections: being able to neurophysiologically evoke the respective behavior patterns with electrical deep brain stimulation (DBS), specifying the anatomical brain locations associated with the various emotions, providing evidence for neurochemical systems regulating the emotions, and demonstrating that arousal of these brain systems with DBS evoked psychological feelings in animals, as evidenced by the rewarding and punishing effects (of these artificially evoked brain states (namely, animals would work for or learn to escape such DBS-induced brain states; this constituted the evidence for affects in non-speaking animals, and correspond to affective shifts seen in humans during comparable brain stimulation for more detail, see Panksepp, 1985, 1998a; Panksepp & Biven, 2012). Thus, three complementary approaches—Darwin and McDougall’s comparative behavioral studies, Cattell’s objective psychological analyses ot T-data, and Panksepp’s neuroscientific research—have independently arrived at very similar conclusions regarding the primary affective feelings underlying human adaptive reaction tendencies that are commonly called personality traits. It is especially noteworthy that Cattell required that the drives and emotions (ergs) be interpretable from the factor analysis of T-data (directly measurable objective data). It now seems that when complex language-mediated descriptions are not involved in the research, a consistency emerges in the search for fundamental origins of large-scale adaptive, psychologically relevant psychobehavioral processes that are built into all mammalian brains.
Suffice it to say that wherever there is convergence on very similar organizing principles through the use of vastly different and independent scientific approaches, one has substantial assurance that nuggets of lasting knowledge are being unearthed. McDougall, a social psychologist, saw reliable patterns of animal behavior in the natural world. Cattell, a cognitively oriented psychologist, tackled the vast complexities of human personality and found similar fundamental processes especially when objective measures of human behaviors and feelings were analyzed mathematically through factor analyzes. Panksepp, a clinically-oriented neuroscientist interested in fundamental sources of human affective life so important for understanding psychiatric disorders, found evolutionarily ingrained neural systems for emotionality that exist in all mammals, and even in more distantly related creatures. And Darwin, who gave us an evolutionary view of the emergence of animal and human nature, would surely smile, for much of this was anticipated in his last two books: The Expression of the Emotions in Man and Animals (1872), where he said in the introduction that he had been thinking about the subject since 1838, and the second “revised and augmented” edition of The Descent of Man, and Selection in Relation to Sex (1874). All of this work combined tells us much about the origins of human nature, and it is remarkable that traditional psychological personality theory can be linked to the rigorous neurobiological study of primal animal emotional systems. Even though there has been much chaos and argumentation in this field, as is normal in science, when one backs away from the historical disputes, the emerging pattern is both realistic and beautiful.
CHAPTER 12
The Big Five
The Essential Core of Cattell’s Factor Analysis
Psychology has not unjustly been accused through the last half century of filling libraries with theories more academic than scientific.
—Raymond B. Cattell and Dennis Child, Motivation and Dynamic Structure
ALTHOUGH CATTELL KINDLED a factor-analytic revolution in the scientific analysis of personality, it took quite a while for it to catch on big time. One of the reasons may have been the early lack of computer power. Calculating factor analyses was an ordeal—a distinctly unpleasant task by hand, even with the time saving but irritatingly noisy mechanical calculators of the day. Even the early computers were limited in the number of variables they could analyze at one time. Then there were the massive stacks of punch cards (each about th
e size and texture of an airplane boarding pass), amassed in heavy boxes that contained the data, delivered by hand to the local computer operators, who fed them into the computer with IBM card readers. Users of university and business computer systems had to wait for hours, if not days, for results that arrived as massive fanfolded paper computer printouts. Those massive computers of yesteryear were way less powerful than the average laptops of today, which in seconds can produce polished graphical outputs of factor analyses from huge correlation matrixes.
However, another reason for the slow acceptance of Cattell’s rigorous and sophisticated statistical breakthrough was the degree of skepticism about factor analyses and their potential for solving real problems about anything as complex and dynamic as human personality. After all, the only thing factor matrices typically provided were numerical results about people’s responses to questions experimenters posed to them about their personalities. Factor analysis simply looked for mathematical grouping of types of answers to those questions. The results were statistical correlations estimating the degree of relatedness among questions, and it was still up to the scientist to make meaning out of those numerical estimates. Computers of that era did not have capacities to describe the meaning of numerical relationships—nor do the powerful machines of the current era, even IBM’s computer Watson6, which can beat humans at their own intellectual games.
Gordon Allport (1897–1967), a highly respected Harvard psychologist, was just one prominent early critic of factor analysis who skeptically wrote that “an entire population (the larger the better) is put into the grinder, and the mixing is so expert that what comes through is a link of factors in which every individual has lost his identity . . . seldom do the factors derived in this way resemble the dispositions and traits identified by clinical methods when the individual is studied intensively” (1937, p. 244). This concern reflected the fact that, at least since Sigmund Freud’s revolutionary depth-psychological views about human nature, personality theory had emerged largely from clinical insight, and few believed that statistics alone could provide substantive contributions (because of the above-noted interpretive issues), let alone supersede, intense clinical analysis.
The Emotional Foundations of Personality Page 24