Extraversion–Introversion D H Saklofske, University of Western Ontario, London, ON, Canada H J Eysenck{, Institute of Psychiatry, London, UK S B G Eysenck, Institute of Psychiatry, London, UK R M Stelmack, University of Ottawa, Ottawa, ON, Canada W Revelle, Northwestern University, Evanston, IL, USA ã 2012 Elsevier Inc. All rights reserved. This article is a revision of the previous edition article by D. H. Saklofske and H. J. Eysenck, volume 2, pp. 321–333, ã 1994, Elsevier Inc.
Glossary Correlation A statistical method for determining the degree and direction of relationship between psychological variables. Extraversion A major personality dimension that includes measurable traits such as sociability, activity, and dominance. Factor analysis A statistical technique that combines clusters of correlated variables into broader categories.
Extraversion: An Overview Alison and Milla are a study in contrasts. Both young women have entered their first year of university and have been assigned to share a room in the university residence. Alison is described as a ‘bubbly’ person, full of charm and wit – a real go-getter. She is active in many campus clubs and can always be found with a group of friends. She enjoys dancing and lively music, plays various team sports, loves to travel, and seems willing to try most things at least once. She is well liked and popular and often at the center of events. She has indicated that she would enjoy work that has lots of action and change to it rather than an office job. In contrast, Milla is a somewhat shy and retiring person who is much more content to work on her studies in the library or at her desk in the residence. She prefers quiet meals and going to movies or reading a book to the more outgoing activities of her roommate. She is not especially well known by fellow students as she tends to move rather quietly from class to class. She has several close friends whom she knows from her public school days. Her expectation is to major in history and English, and author books or work as an archivist. Whatever commonalities and differences exist between these two people, one particular personality dimension comes to the fore. The various behaviors, actions, and preferences would suggest that Alison and Milla vary along the personality continuum defined as extraversion–introversion. Extraverts, like Alison, manifest a relatively consistent and stable profile, characterized by such descriptors as sociable, lively, active, carefree, and sensation seeking. Milla more clearly falls toward the other end of the continuum, which defines introverts, people who are quiet, reserved, passive, controlled, and less sociable and outgoing. {
Deceased.
150
Personality A field of study that focuses on describing individual differences in human behavior and identifying the causes that characterize our unique ways of thinking, feeling, and acting across situations. Traits Relatively consistent and enduring ways of thinking, feeling, and behaving.
Extraversion is seen by many psychologists as a major personality dimension that is very important in formulating a description of human behavior and particularly of individual differences. Psychologists have created numerous concepts in an effort to explain, understand, predict, and even change or modify human behavior. But, in order to understand the nature and relevance of extraversion (E), consideration must be given to a number of basic questions that are of critical importance when examining any personality description. The discussion of E that follows focuses on the following themes: historical perspectives, current views of personality, personality measurement, the biological basis of E, cross-cultural studies, and relationships with other human characteristics.
History Efforts to describe how we both differ from each other and how we are the same certainly predate the development of psychology as a scientific discipline and a profession. In fact, the term personality is derived from ‘persona,’ the Latin word meaning mask that refers to characterization in early Roman theater. Astute descriptions of personality appear in the works of early Greek and Roman philosophers and writers such as Theophrastus and Cicero. The attempt to explain those differences in personality in a systematic way can also be traced to an early period in Roman history, notably in records of Galen of Pergamum (c130–c200 AD), who was a physician to early Roman emperors. The schema that Galen espoused, the humor theory of temperament, was remarkably influential and resilient over many centuries, and indeed, vestiges of this schema are linked to the modern personality description of extraversion. For Galen, humors were bodily fluids that influenced the health, physiognomy, and character of man. There were four primary humors, chore (bile), melanchole (black bile), sanguis
Extraversion–Introversion
lic
Emotional M el an ch o
(blood), and flegma (phlegm). These four humors were understood in the context of a general cosmological theory where fire, earth, air, and water were conceived as the four basic elements of all things. As expressions of bodily processes, psychological characteristics were influenced by the particular blend (L. temperare) of the four humors, that is, temperament. For example, steadiness and solidity was thought to depend on black bile (melancholy), whereas simple mindedness came from blood (sanguine). In the analysis of Galen’s work that survives today, it is clear that he mainly addressed medical problems. His observations of character were infrequent and bear little resemblance to the adjectives ascribed to the humor categories by later authors. Further, the humor theory as a framework for understanding health issues was displaced by progress in medical science. Nevertheless, as a descriptive personality classification schema, the four temperament categories prevailed during the eighteenth and nineteenth centuries, notably in the work of German philosophers Immanuel Kant and Wilhelm Wundt. Both Kant and Wundt developed schemas of personality adjectives that expressed the four temperament categories in terms of two concepts. For Kant, the two concepts were feelings and activity. The sanguine temperament was characterized by strong, short-lasting feelings in contrast to the melancholic temperament characterized by weak, long-lasting feelings. The choleric temperament was characterized by intense, but brief activity in contrast to the phlegmatic temperament that was characterized by weak, but enduring activity. This schema yielded four independent temperament categories based on two dimensions. For Wundt, the two concepts were strength of emotion and changeability. Although the two concepts are similar to those employed by Kant, Wundt shifted the emphasis from four independent categories to a two-dimensional system that encompassed the four types. In this schema, the choleric and melancholic are characterized by strong emotion (unstable) in contrast to the sanguine and phlegmatic that are characterized by weak emotions (stable). The sanguine and choleric are characterized by rapid change in contrast to the melancholic and phlegmatic that are characterized by slow change. In terms of semantics, the unchangeable–changeable dimension and the strong–weak emotion dimension in Wundt’s schema clearly resemble the extraversion and neuroticism dimensions, respectively, that were developed by Hans Eysenck using modern statistical methods. The relation between the classical four temperament types from Wundt’s depiction and Eysenck’s results is illustrated, using modern statistical methods of personality description, in Figure 1. During the early part of the twentieth century, there were significant advances in the application of mathematical concepts to the study of individual differences, which enabled personality description to move from the insightful categorical association of descriptive personality adjectives as discerned by Kant and Wundt to a rigorous, objective clustering of adjectives, using statistical methods. This advance in statistical methods was pioneered by several significant figures, notably Francis Galton, who developed the correlation statistic and Charles Spearman, who developed a procedure, factor analysis, for grouping correlation statistics into a common factor. During this period, the terms introversion–extraversion emerged in the personality description literature, effectively
151
Anxious Worried Unhappy
Suspicious Serious Thoughtful
Ch Quickly ol er ic Roused Egocentric Exhibitionist Hot-headed Histrionic Active
Unchangeable
Changeable
Playful Reasonable Easy-going High-principled Sociable Controlled Carefree Persistent Ph Hopeful Steadfast ne le gm ui Contented g Calm n at ic Sa Nonemotional Figure 1 Representation of the four classical temperaments using Wundt’s schema and adjectives from modern descriptions of extraversion (changeable–unchangeable) and neuroticism (stable–unstable).
replacing the terms unchangeable–changeable from Wundt’s schema. Although the original introduction of introversion– extraversion in psychological literature is obscure, it is clear that the terms were popularized in the hugely successful writing of the Swiss psychoanalyst Carl Jung. This popularity led several investigators (e.g., J. P. Guilford and R. B. Cattell) to develop psychometric tests of introversion–extraversion. This dimension is prominent in all modern personality classification systems. Neither Kant nor Wundt offered a causal basis or theory of temperament to replace Galen’s humor theory. Early in the twentieth century, there were several scientists who suggested physiological processes to account for differences in the temperament categories. Notably, Ivan Pavlov, famed for his contributions to learning and conditioning theory, applied the neurophysiological concepts of excitation and inhibition to explain the four temperament types. Introversion and extraversion types were included in an all-encompassing theory of personality developed by Carl Jung. In his theory, personality was determined by innate dispositions (instincts), culture, and conscious and unconscious processes. In 1929, from current knowledge of the nervous system, William McDougall outlined a hypothetical chemical theory that presaged our contemporary understanding of the biological basis of introversion–extraversion. In his proposal, greater cortical inhibition, mediated by a hypothetical neurochemical factor, effected the behavioral restraint and reticence exhibited by introverts. A similar course of inquiry blossomed in the biological basis of introversion–extraversion proposed by Hans Eysenck in 1967.
Contemporary Views of Extraversion Psychologists have been and are still actively engaged in the debate as to whether human personality can be best and most meaningfully described with the aid of only a few or many
152
Extraversion–Introversion
‘words.’ One can very quickly create an exhaustive list of personality descriptors commonly found in the lexicon: for example, happy, bright, friendly, kind, assertive, honest, sullen, patient, etc. At the same time, we can accurately and concisely describe individuals along such major dimensions as extraversion–introversion when we recognize that some human characteristics or behaviors share much in common with others (e.g., shy, withdrawn, nonassertive, not talkative, quiet, loner, etc.). Psychologists’ attempts to reduce human behavior to a finite set of descriptions are not an attempt to impose an artificial and nonworkable solution onto a truly complex personality description, but rather, an attempt to find the most meaningful and parsimonious means of representing human personality. As already described, extraversion has its roots in the early Galen–Wundt theory of the four temperaments but has been more recently elaborated by such notable contemporary psychologists as R. B. Cattell, H. J. Eysenck, L. Goldberg, J.A. Gray, J. P. Guilford, J. Wiggins, and P. Costa and R. McCrae. Extraversion is also included in other personality models but under different names. Further, E scales appear on all of the Eysenck personality measures, beginning with the Maudsley Personality Inventory and Eysenck Personality Inventory to the more recent Eysenck Personality Questionnaire series (EPQ and revised scales such as the EPQ-R and Eysenck Personality Profiler, EPP), where it constitutes one of the three major personality variables. The description of individuals according to a small number of personality dimensions is frequently referred to as the trait approach. Essentially, through the use of sophisticated statistical techniques, such as correlational and factor analysis, psychologists have reduced the large number of behavioral descriptions to a much smaller number of stable and generalizable characteristics that make possible the explanation and prediction of individual differences and similarities possible. This view rests on the assumption that a meaningful model of personality can be based on a smaller but more powerful set of traits that are remarkably stable across time and differing situations. Very briefly, E appears as a major personality variable in almost all the current trait descriptions, whether by name or as a result of further research investigations stimulated by these theories. An ongoing debate is whether personality is best described, for example, by the 16 primary traits suggested by Cattell, or the ‘big 5’ trait dimensions emphasized by Goldberg and Costa and McCrae, or by the ‘3 super factors’ described by Eysenck. A consistent finding from the research studies is the emergence of a robust factor that may be called extraversion. Analysis of large numbers of responses to the various personality questionnaires seems to invariably produce an E factor. The two most influential trait theories today agree that E is one of the major dimensions. Eysenck adds two other major personality dimensions labeled neuroticism (N) and psychoticism (P). Costa and McCrae add neuroticism, openness to experience, agreeableness, and conscientiousness to their extraversion factor to produce the ‘big 5’ trait dimensions of their Five Factor Model. Some of this confusion may be clarified by examining how Eysenck has developed his personality model with particular reference to E. His system is a hierarchical one and begins with
single acts or thoughts. The second level focuses on more consistently occurring behaviors and cognitions. Next comes the level of traits and finally types that reflect a composite of various correlated traits. Thus, each level is more encompassing and inclusive than the previous level. With respect to E, it is a dimension that encompasses a number of trait descriptors such as happy, cheerful, sociable, lively, assertive, and carefree. It is the observed intercorrelations between these traits that justify our postulating the existence of extraversion. This structure is shown in Figure 2. Similarly, for the other major high-order concept, neuroticism, traits that intercorrelate and thus define it are shown in Figure 3. How can we justify such a typology? Table 1 shows six questions supposed to measure E, and six questions supposed to measure N. We can collect large numbers of answers from random samples of the population, intercorrelate the answers, and check whether the patterns of intercorrelations and more parsimonious factor analysis results give us two independent factors corresponding to E and N. Should we obtain the results shown in Figure 4, it is obvious that the results support the theory. We do not have to rely on self-ratings to obtain such evidence. Figure 5 shows the results of such a factor-analytic investigation of the relation between ratings made of thousands of children in a child guidance clinic, using items of behavior. It will be seen that extraverted children demonstrate conduct problems, and introverted children personality problems. Of course, all of these children are essentially high N scorers; children with low N scores would simply be more outgoing behaviorally when extraverted, and unsociable when introverted.
E
Sociable
Lively
Carefree
Active
Assertive
Dominant
Surgent
Sensation-seeking
Venturesome
Figure 2 Some traits defining extraversion.
N
Anxious
Depressed
Irrational
Guilt feelings
Shy
Low self-esteem
Moody
Figure 3 Some traits defining neuroticism.
Emotional
Tense
Extraversion–Introversion
Questionnaire items measuring extraversion–introversion
Table 1 Questions
Key
1. Do you sometimes feel happy, sometimes depressed, without any apparent reason? 2. Do you have frequent ups and downs in mood, either with or without apparent cause? 3. Do you worry a lot? 4. Does your mind often wander while you are trying to concentrate? 5. Are you frequently ‘lost in thought’ even when you are supposed to be taking part in a conversation? 6. Are you sometimes bubbling over with energy and sometimes very sluggish? 7. Do you like surprises and spontaneous activities? 8. Are you happiest when you get involved in some project that calls for rapid action? 9. Do you usually take the initiative in making new friends? 10. Are you inclined to be quick and sure in your actions? 11. Would you rate yourself as a lively individual? 12. Would you be very unhappy if you were prevented from making numerous social contacts?
N
Introversion 0.6
0.5
Ä
N
0.3
N E E E E E E
Changeable mood Ä
0.2
0.1
0.1 -0.1
-0.3
-0.4
5 Ä7 10 Ä 4
Personality problems Ä
ÄNervous
Ä Mental conflict Intelligent Ä EmotionallyÄ Irritable Unstable Ä Lazy Ä SpoiledÄ Ä Masturbation Neuroticism Irresponsible Ä 0.5 0.6 0.7 0.2 0.3 0.4 SexualÄ ÄLacks interest delinquency BossyÄ Temper UnpopularÄ Ä tantrums Fantastic Ä lying ÄEgocentric Ä Rude Violent Ä Disturbing influence Ä Disobedient Ä Ä Fighting Ä Ä Swearing Conduct Destructive Ä Ä Lying problems Truant Ä Truant Ä (home) (school) Ä Stealing
-0.5 Extraversion
3 2 Ä6
1
Ä4 1 1
Ä Sensitive
Absent-minded Ä Day dreams Seclusive Ä Ä Ä Depressed Ä Inefficient Inferiority feelings Ä Ä Queer
N
7 Extraversion 11 Ä Ä 12 6 9Ä Ä8
1
Psychoneurotic
0.4
N N
-0.2
2
153
2
3
4
5Ä
5
6
1 Neuroticism Ä Ä 7 8 Ä 2 3
2
Figure 4 Relative position in two-dimensional space of six neuroticism and six extraversion questionnaire items.
Jung had earlier put forward the hypothesis that introverted neurotics would present with psychasthenic or dysthymic symptoms (anxiety, depression, apathy), while extraverted neurotics would show hysterical symptoms (hysterical attitude, conversion symptoms). A study of the behaviors and attitudes of hundreds of neurotics bore out this theory as shown in Figure 6. These empirical studies will give an idea of the structure of personality. A discussion of empirical findings that help us to better understand extraversion follows. It is important to know about the theories and models that psychologists have created to account for and explain extraversion. The most thorough and widely accepted is the theory put forward by the late Professor Hans Eysenck (1916–1997), who spent much of his professional career at the Institute of Psychiatry, University of
Figure 5 Two-factor representation of conduct and personality problems in children, showing breakdown into extraverted and introverted groups.
London. The ensuing discussion draws extensively from the theoretical and research contributions of Professor Eysenck. A question that is frequently asked is how psychologists measure variables such as extraversion. A number of questionnaires have been developed and are mentioned below. While a description of the structure of personality is important, it is equally important to examine its causes. Extraversion would not be a very useful person description if it only served as a summary label for a collection of human behaviors. Of considerable importance are questions related to whether extraversion is learned or whether genetic factors determine a person’s position on the extraversion dimension. What are the biological underpinnings of extraversion, and are there certain brain structures or functions that underlie an individual’s behavior along the introversion–extraversion dimension? Further points that must be addressed relate to the robustness of extraversion such as: can it be identified cross-culturally? What extraversion, or any other personality factor, tells us about other aspects of human functioning is of relevance. For example, are extraverts more or less prone to certain psychiatric illnesses? Are there differences in various performance indicators between extraverts and introverts?
154
Extraversion–Introversion
Neuroticism 1.00 Badly organized personality
0.90
0.80
0.70 Dependent
Narrow Little energy
Unemployment
interests
Schiziod Poor muscular tone
No group membership Hysterical attitude Low I.Q.
Hysterical conversion Sex anomalies
0.70
0.60
0.50 0.40 Hysteria
Hypochondriasis Hypochondriacal
Degraded work history
Pain Alcohol
0.30
0.20
0.60 Boarded Dyspepsia out 0.50 Abnormality Cyclothymic in parents 0.40 Unsatisfactory home Wartime separation 0.30 Tremor Headache Effort 0.20 intolerance
Fainting fits
Unskilled
0.80
Abnormal before illness
Age 0.10 30+
Domestic problems Somatic Bomb and anxiety exposure
0.10
0.10
0.20
0.30
Married Irritability
Anxiety
Apathy Depression Obsessional 0.40 0.50 Dysthymia
0.60
0.70
0.80
Figure 6 Behaviors and characteristics of introverted and extraverted neurotics.
The Measurement of Extraversion Psychologists employ various procedures for gathering information about human behavior. We may observe a person in a structured situation or in more natural settings. The person may be interviewed or we can ask others to tell us about the person. Various tests can be conducted to determine, for example, how an individual reacts to stressful situations or responds to tasks that require the learning of new behaviors. In some instances, we gather information that tells about the person’s internal condition such as the level of cortical arousal (EEG or MRI scans). Various biochemical influences on personality can be examined by extracting samples of hormones such as testosterone and cortisol, or neurotransmitters including monamine oxidase, dopamine, or serotonin. We may also assess personality with self-report questionnaires. It is this personality test, inventory, or questionnaire method that is most commonly used to assess variables such as extraversion. Hans and Sybil Eysenck have developed the most wellknown and frequently used personality measures of extraversion. Recent versions include the EPQ-R, for which there are both adult and child forms. This questionnaire contains items that are answered with a ‘yes’ or ‘no’ and measure the three superfactors of extraversion, neuroticism, and psychoticism. It also includes another scale that measures untruthfulness, faking, or social desirability (e.g., answering questions in a way that presents a supposedly more ‘desirable’ profile of the
Table 2 Please respond to the following statements with ‘yes’ if you agree that they more or less describe you. Answer ‘no’ if the statement is not a reasonably accurate description of you 1. I like loud music 2. I would much prefer a quiet evening at home to going to a party 3. Driving fast cars would appeal to me 4. I am easily bored 5. I enjoy meeting new people and having lots of friends 6. I make decisions in a very methodical and deliberate way 7. I would not enjoy traveling to new and different places 8. Others see me as an easygoing and carefree person There are no right or wrong answers.
person). More recently, the EPP has replaced the EPQ in both professional and research use. Since extraversion, in the main, taps traits and behaviors such as those shown in Figure 2, the kinds of items that might be found on such inventories may look like the example questionnaire shown in Table 2. While these are not actual items from the EPQ or EPP, they do reflect the kind of content and appearance found in measures of extraversion. If ‘yes’ were answered to questions 1,3,4,5, and 8 and ‘no’ to questions 2,6, and 7, then the respondent would likely be more of an extravert than introvert. Many hundreds of studies have been conducted that focus on how useful and ‘good’ these questionnaires are at assessing
Extraversion–Introversion
introversion–extraversion. Interested readers may wish to examine the journal Personality and Individual Differences, for research papers reporting results of studies with the EPQ, EPP, and related measures. The Eysenck scales show consistently high reliability (an indication of the test’s precision, consistency, and stability) and validity (an indication of whether the test measures what it was intended to measure). Unfortunately, in addition to the Eysenck measures, other scales developed to measure extraversion differ in the relative mixture of the affective, behavioral, cognitive, and goal-seeking components, which has caused some confusion in the research and clinical literature.
155
Trait models of personality are sometimes criticized for apparently claiming to explain differences in behavior by simply postulating the existence of traits based on that behavior. How do we describe social shyness? We correlate a number of questions related to that trait, extract the factor of social shyness, and then describe the observed behaviors by the very trait label based on the analysis of these behaviors! But this criticism is incorrect. In the first place, we carry out the statistical analyses in order to discover whether our original notion was correct, that is, that there was in fact only one factor of social shyness. At first sight, it may seem obvious that this is so, but, in actual fact, it may not. When we take a large sample of items relating to sociability, social shyness, and similar notions, calculate the correlations, and factor analyze the resulting table of correlations, we find, not one factor, but two. There is an introverted social shyness, defined by items that tell us that introverts do not particularly like being with other people and prefer being by themselves most of the time. However, they are not worried about being with other people and can perfectly sustain social interactions. There is also a quite independent type of neurotic social shyness, where people are actually afraid of others and avoid them for that very reason. They would like to socialize with others, but are prevented from doing so by their worries and anxieties about social relations. Thus, a statistical analysis of this kind is carried out in the main to see whether our commonsense beliefs are justified or not, and in many cases what is found is that they are not. In other words, we have to solve the problem of personality description before we can go on to attack the problem of causality. Factor analysis and other correlational methods are not meant to tell us anything about causality, but to act as tools for the discovery of a useful personality taxonomy. Having solved the problem, we may then go on to carry out the more difficult task of discovering why some people are sociable, others quite less so, and why some people are extraverted, others introverted.
that environmental factors, and particularly the influence of the family, were the major, if not sole, determinants of individual differences in human behavior. Early studies of identical and fraternal twins seemed to support such a view, but improved research methods have challenged this assertion. There are several ways of assessing the relative importance of nature and nurture. The first, and probably the most important, is to study identical twins (monozygotic (MZ) twins) who were brought up from birth in different environments, that is, were adopted. Identical twins share 100% heredity, being the product of a single ovum, fertilized by a single sperm, and then separating into two separate entities. MZ twins separated at birth share identical heredity, but have completely different environmental influences determining their behavior. We get an accurate estimate of the importance of heredity by looking at the intercorrelations between such MZ twins; if environment were all-important, there should be no correlation at all. In fact, several studies have shown quite high correlations, suggesting that heredity accounted for at least 50% of the total difference observed. Similarly, we can compare MZ twins and dizygotic (DZ) twins; that is, fraternal twins who share, on the average, only 50% heredity, and are no more alike than ordinary siblings. If heredity was important, we could expect MZ twins to be much more alike than DZ twins, and this is precisely what is usually found. So here again, we have evidence for the importance of heredity. These and other studies, on topics such as whether adopted children are more alike in personality with their biological or their adoptive parents, leave us in no doubt that heredity is extremely important. They also show that the theory implicating only the family is quite wrong. We can divide the environmental factors determining differences in personality into those between families, emphasizing the importance of family upbringing, and those within family, emphasizing the environmental differences between different children coming from the same family, such as one having a good teacher, the other a poor one, or one marrying a kind and loving person, the other an abusive, unsupportive one. Several very large-scale studies have now shown that it is the within-family environmental factors that affect individual differences in personality between children, not those between families, a finding that demonstrates that practically all the usual theories of personality are in fact inadequate, and have to be revised or replaced. Most current personality theories have now taken the implications of genetics into account. What is important to remember here, however, is that even very high heritabilities do not imply a lack of malleability; nor do they deny the possibility of changes over the life span. Heritability is a within-cohort measure that increases with a lack of environmental variability; heritability estimates say little about mean changes over time or what happens if the environment is systematically varied.
Hereditary Influences
The Link Between Biology and Extraversion
The first step in such a causal analysis must be to discover to what extent differences in behavior are determined by heredity, and to what extent they are determined by environmental factors. For many years, personality theorists took it for granted
The compelling evidence from behavioral genetic studies confirms the important contribution of heritable, constitutional factors to the development of personality. This evidence endorses the views of scholars, from Galen to McDougall and
The Basis of Extraversion
156
Extraversion–Introversion
Pavlov, who speculated on the physiological determinants of personality. Only physiological processes, under the control of DNA, are heritable. During the twentieth century, there was considerable research that aimed to identify the psychological processes and neurological mechanisms that mediated individual differences in personality. Foremost in this research enterprise was the leadership provided by Professor Hans Eysenck and his colleagues at the Institute of Psychiatry, University of London, in their studies on introversion–extraversion and individual differences in cortical arousal. From the 1950s, there was widespread interest in the idea that a continuum of neurophysiological states of arousal, ranging from sleep to alertness and excitement, was important in influencing such psychological processes as attention, motivation, and learning. Importantly, it was shown that low-amplitude, high-frequency EEG activity, associated with wakefulness and attention, and high-amplitude, low-frequency EEG activity, associated with sleep and drowsiness, were modulated by activity in the ascending reticular activation system (ARAS). The ARAS is a diffuse network of neural fibers that are innervated by collaterals from the main sensory pathways and that project to the midbrain and higher cortical centers. The discovery of the ARAS provided a plausible physiological basis for the arousal continuum and for the psychological processes involving modulation of sensory inputs that it served. Several of these psychological processes were known to differ between introverts and extraverts. For example, compared to extraverts, introverts tended to be more alert (as indicated by stronger reactions to stimulus change) and to be more vigilant in situations that demanded sustained attention. It was in this context that Hans Eysenck proposed the arousal hypothesis in 1967 as an explanation for individual differences in introversion and extraversion (Figure 7). The central hypothesis identified individual differences in behavior between introverts and extraverts, with differential thresholds of excitation and inhibition in the various parts of the ARAS. It was suggested that, compared to extraverts, introverts were characterized by higher levels of activity or lower
VB AR
AS
AAP Figure 7 Diagrammatic representation of the interaction between the limbic system (visceral brain) and the reticular–cortical arousal system. Copyright ã 1967, H. J. Eysenck.
levels of excitation in the ARAS. In general, it was expected that the responses of introverts and extraverts would be affected differently in psychological processes and psychophysiological measures that were thought to be sensitive to the stimulusmodulation function of the ARAS. The hypothesis inspired hundreds of experiments and continues to guide research on the topic even as personality theory and ARAS brain physiology advances. From this substantial body of work, some induction of the causal bases of extraversion can be made from the reliable effects that were demonstrated and from research employing different methods and paradigms that converge on common processes. First, there is considerable research, both in the field and in the laboratory, demonstrating that introverts are more sensitive than extraverts to sensory stimulation. This effect can reasonably account for some social behavior that is displayed by introverts, that is, their preference for quieter environments and solitude and avoidance of noisy environments and intense stimulation, such as partying. For example, one project investigated introversion–extraversion and the area where a person preferred to study in the library. Introverts tended to prefer sitting in quieter corners of the library whereas extraverts seated themselves closer to the noisier, well-traveled areas, although there is the view that an impulsivity effect underlies these reported differences. In another example, when given the opportunity to set the volume on the radio to their preferred listening level, on average, introverts set the volume level 20 decibels lower than did extraverts. In the laboratory, under rigorously controlled conditions, introverts exhibited greater sensitivity to sound and taste stimuli, that is, lower auditory and gustatory thresholds. Similarly, introverts show less tolerance for intense sensory stimulation, as evidenced by lower pain thresholds and noise thresholds. The behavioral differences in response to sensory stimulation are also evident in psychophysiological indices that gage the physiological response to sensory stimulation. When small metal electrodes are placed on the finger tips to record minute changes in sweat gland activity, the greater sensitivity of introverts compared to extraverts to brief moderate-intensity tones is evident in their larger skin conductance response, that is, greater sweat gland activity. From electrodes placed on the scalp, introverts exhibit larger electrocortical responses (event-related potentials) to simple auditory stimulation. Similarly, with techniques that measure the speed at which eye blinks are initiated, introverts display faster startle reflex responses than extraverts to moderate-intensity noise bursts. It was also found that introverts are characterized by faster brainstem auditory evoked potentials (BAEP) than extraverts. This effect too is indicative of greater auditory sensitivity for introverts. However, BAEP waves develop very early in the auditory pathway, prior to excitation of the ARAS, and they are relatively independent of descending cortical influences (the waves are unaffected by either sleep or arousal). Thus, the BAEP effects cannot be easily accounted for by differences in the ARAS as originally proposed. It is also important to note that the described differences between introverts and extraverts reflect differences in response to stimulation rather than tonic or base levels of arousal. This conclusion is drawn from the absence of differences when autonomic nervous system measures of arousal, such as sweat
Extraversion–Introversion
gland activity (skin conductance), are obtained prior to sensory stimulation and in conditions that have low arousal potential. Notably, differences are seldom reported with EEG measures in low-arousing conditions or during sleep. In effect, it is the reaction to stimulation, rather than endogenous differences in level of physiological activity, that differentiates introverts and extraverts. This conclusion challenges the idea that extraverts seek stimulation in order to achieve an optimum level of arousal. An important clarification of the arousal hypothesis was the distinction between energetic and tense arousal. Energetic arousal, which contrasts energy versus sleepiness, varies over the day in a clear diurnal rhythm. Tense arousal, reflecting a dimension of tension and worry versus calm and relaxed, varies in response to situational demands. While not differing in their tense arousal, introverts and extraverts do differ in the phase of their energetic arousal. Introverts have a somewhat earlier peak level of energetic arousal than do extraverts. More importantly, high energetic arousal is associated with increased activity and sociability rather than the inverse relationship that would be expected by the arousal-seeking hypothesis. Introverts and extraverts are also known to differ in a range of social behaviors that can be linked to differences in movement or motor activity. These differences in the expression of motor activity provide a plausible basis for understanding the greater spontaneity, sociability, and liveliness of extraverts. For example, extraverts are more disposed to participate in physical activities and sports than introverts. Extraverts are more talkative and initiate conversation more frequently than do introverts during interview situations, and they are more restless (fidgety) in restricted environments. In controlled laboratory experiments, faster reaction time is commonly observed for extraverts than for introverts. On a task that involves tracking a small target on a revolving disk and requires refined motor control, extraverts show less effective tracking performance than do introverts. In reaction time tasks, where participants are instructed to respond as quickly as possible, extraverts tend to make more false-positive errors than introverts. These effects indicate that extraverts inhibit the initiation and control of movement responses less readily than introverts. In reaction time tasks, it is understood that the task involves a cognitive decision to respond when a cue to respond is presented and a movement response that typically requires pressing a response button. Decision time and movement time can be separated in reaction time tasks by measuring the time from the onset of the cue to respond to the release of a home button that is, decision time, and by measuring the time from the release of the home button to the press of the response button, that is, movement time. In a series of studies, it was consistently observed that extraverts exhibited faster movement time than introverts, but that there were no differences in decision time. In effect, it was motor activity rather than cognitive activity that was the discriminating factor. Moreover, it was clear that this movement time effect was due to the faster initiation of movement for extraverts than introverts rather than the speed of the ballistic motor response. The behavioral differences in motor activity between introverts and extraverts were also explored using psychophysiological indices of movement responses. In the recording of electrocortical activity during reaction time tasks, extraverts displayed greater response amplitude than did introverts in
157
the interval between a warning signal and the cue to respond. This effect is indicative of greater response motor preparation for extraverts than introverts. It has also been observed that extraverts exhibit slower motor-neuronal recovery than introverts following the discharge of motor reflexes, an effect linked to increased dopaminergic activity for extraverts. Differences in motor activity between introverts and extraverts were also examined with an electrocortical response measure, the lateralized readiness potential (LRP), that directly assesses movement initiation processes following stimulusrelated processing. In this procedure, the duration of premotor activity, including stimulus analysis, response preparation, and some aspects of response selection, is distinguished from the duration of motor activity, independent of stimulus processing. With this task, shorter response-linked LRP latencies were found for extraverts than introverts, indicating faster speed of motor processing for extraverts than for introverts. There were no differences between introverts and extraverts, however, for stimulus-linked LRP latencies. Again, it is the motor component of the response process, which favors extraverts, rather than the cognitive analysis component that is the relevant factor in this analysis. Overall, these results indicate that introverts and extraverts are characterized by fundamental differences in the expression of motor activity. Given the remarkable differences between introverts and extraverts in sensory sensitivity and in the expression of motor activity that were consistently demonstrated, there remains the important question of identifying the neural circuits and neurochemical properties on which these effects depend. The neural generators of psychophysiological response measures have been extensively explored and do provide some insight into the neural pathways that are involved in individual differences in extraversion. Perhaps more importantly, there is good evidence indicating that dopamine, a neurotransmitter that is thought to modulate the probability and strength of behavioral responses to sensory input, may be implicated in variation in extraversion. Dopaminergic activity in the mesolimbic pathway is involved in locomotor activity and reward behavior. Increased dopaminergic activity in the mesostriatal pathway can enhance sensory sensitivity by counteracting the inhibitory effects of the striatum on the thalamus and ARAS. When alpha-methylpara-tyrosine (AMPT) was used to block the synthesis of dopamine, both decision time and movement time were markedly slower for introverts during a choice reaction time task. However, AMPT had no effect on the response time of extraverts. Similar results were observed when remoxipride, a receptor blocker that selectively inhibits mesolimbic dopaminergic activity, was used. These important experiments pioneered by Thomas Rammsayer at the University of Bern, Switzerland, indicate that the differences between introverts and extraverts in the transmission of sensory input to motor seem to be a clear function of modulation in the dopaminergic system. As an extension to this work, introversion–extraversion differences in dopaminergic activity may also be related to the sensitivity in detecting and learning cues for reward. That is, with more dopaminergic activity, one is able to learn which environmental stimuli lead to future reward. When presented with positively valenced stimuli, extraverts display more brain activation, as measured by fMRI, than do introverts. This is true for reward cues but not true for cues of punishment and nonreward.
158
Extraversion–Introversion
Cross-Cultural Research into Extraversion There are other consequences of the general theory that E and N are determined very largely by heredity, and that their behavior is mediated by biological structures in the central nervous system and the autonomic nervous system. One of these is that E and N should appear not only in those cultures where the original studies were made (essentially the European-North American culture) but should also be equally characteristic of all other countries. What is asserted is not that all these cultures would be equally extraverted, or similar in emotional lability, but rather that the same factors would emerge in these countries as were found in North America, Australia, etc. The experiments were done in over 35 different countries, including a large number of European, Scandinavian, Slavic (e.g., Russia, Hungary), Southeast Asian (e.g., Japan, China), South American, and African (e.g., Uganda, Nigeria) countries. In all these countries, over 500 males and 500 females were administered the EPQ, and the items were then intercorrelated and factor analyzed to see whether the resulting factors would be the same as the ones we found in England and the United States. It was found that in all these countries, there were practically identical structures of items, verifying the existence of N and E. Of course, not all countries had identical levels of E and N; Japan was very high on N but lower on E. The United States was very extraverted. Thus, the prediction is verified. Identical factors emerged in many different countries and cultures. Of interest is the fact that similar findings have been obtained across large samples of children from different countries such as Canada, Denmark, Greece, Spain, and Singapore.
Drug Studies and Extraversion There is another way of testing the hypothesis that personality is mediated by biological mechanisms in the central and autonomic nervous systems. If this is true, then it should be possible to shift the person on the various dimensions involved by means of drugs known to be relevant to the hypothetical underlying structures. Thus, stimulant drugs should make people more introverted because they are known to raise the cortical arousal level, while depressant drugs should have the opposite effect, making people more extraverted. It is well known that alcohol, which is a depressant drug, does make people more extraverted (except in excessive doses, of course), while stimulant drugs have the opposite effect. Similarly, adrenergic drugs lead to greater neuroticism, anxiolytic drugs to greater stability. Figure 8 illustrates these effects. Also shown is the third major personality dimension, psychoticism, which is increased by hallucinogens and diminished by antipsychotic drugs. A great deal of work conducted along these lines has yielded positive results.
Extraversion and Other Personality Factors E as a dimension of personality is never found in isolation but is always interacting with other dimensions of personality and with intelligence. Thus, an intelligent extravert behaves very
Neuroticism
Adrenergic drugs Psychoticism Hallucinogens
Introversion
Stimulant drugs
Depressant drugs
Extraversion
Anti-psychotic drugs Impulse control
Anxiolytic drugs
Stability
Figure 8 Psychotropic drugs and personality.
differently from a less intelligent one, and a neurotic extravert very differently from a stable one. Furthermore, while extraverts may have the same score on the E scale of the EPP, they nevertheless differ to some extent from each other because they may be characterized in the main by different traits. The sociable extravert is not necessarily assertive, the carefree one not necessarily dominant. These traits are all intercorrelated, but the correlations are far from perfect. Different extraverts may sometimes show different facets, and therefore subtle differences. It should never be assumed that knowing the main dimensions of personality of a given person tells us all there is to know. The hierarchical system outlined earlier simply tells us that knowing the position of a person on the major dimensions will tell us more about that person than any other similarly restricted data set. If we know a person’s IQ score as well as degree of E and N, we will know more about him or her than any other three figures could tell us. This does not mean that there is not a great deal more to be known; if we characterize the person in terms of 40 or 50 traits, we would certainly have a much more detailed picture. But much of this information would be redundant, and we would have great difficulty in absorbing it all. Having a hierarchical theory means that for different purposes, we may choose different levels of analysis. We may be particularly interested in one aspect of E, say sociability, or dominance, and under these conditions, we would obviously prefer to make use of a questionnaire that measures sociability or dominance. Thus, a hierarchical system is very flexible, for it is not committed to working only at the highest level. R. B. Cattell, for instance, preferred to focus on his 16 personality factors, although these are intercorrelated and give rise to higher-order factors very similar to E (which he called ‘envia–exvia’) and N (which he called ‘anxiety’). There is no essential contradiction between the Cattell and Eysenck systems. For all practical purposes, they simply lay emphasis on different levels of analysis; however, the relative weighting of the lower order factors, facets, or aspects does lead to somewhat different conceptualizations of extraversion. Essentially, what a proper theory of personality must do is look at all the steps, from distal causal factors, such as heredity, to proximal causal factors, such as differences in the
Extraversion–Introversion
functioning of the limbic system or the ARAS, to the major dimensions of personality like E and N. The theory must then proceed to the study of proximal consequences, such as the study of changeability already mentioned, which is best done in the laboratory. And finally, it must proceed to distal consequences, mainly in the social field, such as sociability or differences in sexual behavior. In the case of E, we have a complete chain, from DNA, through ARAS, to E, to laboratory behaviors, and then on to social behavior. It is not suggested that each element in the chain has been finally located and that no further progress is either required or possible. The concept of cortical arousal presents many difficulties, and there are many anomalies in the experimental analysis of extraverted behavior in those terms. These will be analyzed and solved in future research; at the moment, all we can say is that the theory has much going for it.
Extraversion and Social Behavior In a complete chain of events, our main interest of course is in the final link, that is, social behavior. An examination of the current research literature implicates E in a wide variety of reallife behaviors including social interactions, sexual activity, work performance, school achievement, psychiatric disturbance, and antisocial and criminal behavior. For example, it has been shown that traffic accidents are much more frequent in people who are high N and high E. It has also been shown that by using this knowledge in the selection procedure for drivers, the accident rate can be halved. A rather different field is education. There has been a great deal of interest in the last 20 years in the so-called ‘discovery learning’ or inquiry-based methods, in which students are encouraged to discover basic principles and rules. It has been found, by comparing classes taught by either discovery or reception learning methods, that there is little difference in achievement. This may be due to the fact that both methods are equally good, but it may also be that extraverts learn better by means of discovery methods, and introverts by means of reception methods, as predicted on theoretical grounds. Figure 9 shows the results of one such study, where pupils were taught by one method or the other and then were separated into introverts and extraverts.
30
30
28
28
Mean score
24 22 Reception learning
20 18 0
26 Mean score
Discovery learning
26
Discovery learning
24
Reception learning
22
Extraverts
Post-test (after one week)
The posttest was done after 1 week, and again after 5 weeks. It does show that extraverts do indeed benefit more by discovery learning, and introverts by reception learning.
Conclusion Most people underestimate the very large individual differences in personality, which heredity has created. We tend to think that most people are like us, just slightly different because of environmental events, particularly events characteristic of a given family. Such a view underestimates the very real differences, which are largely genetic, or, if environmental, have little to do with the way children are brought up in a given family. Theories of personality characteristic of the last 60 years have now been shown to be essentially contradicted by the facts of behavior genetic research. We will have to develop new theories in order to do justice to the facts as described. The above descriptions of extraversion are contemporary. While there is good reason to suggest that E will continue to be recognized as a major personality dimension, the specific findings reported here, ranging from its biological basis to behavioral manifestations, are currently being examined and reexamined.
See also: Behavior Genetics of Personality; Behavioral Genetics; Individual Differences in Temperament: Definition, Measurement, and Outcomes; Temperament and Individual Differences.
Further Reading Barrett PT, Petrides KV, Eysenck SBG, and Eysenck HJ (1998) The Eysenck Personality Questionnnare: An examination of the factorial similarity of P, E, N, and L across 34 countries. Personaltiy and Individual Differences 25: 805–819. Eysenck HJ (1990) Biological dimensions of personality. In: Pervin L (ed.) Handbook of Personality: Theory and Research. New York: Guilford. Eysenck HJ and Eysenck MW (1985) Personality and Individual Differences. New York: Plenum. Eysenck HJ, Wilson G, and Jackson C (1996) Manual of the Eysenck Personality Profiler. Guilford: Psi-Press. Furnham A, Eysenck SBG, and Saklofske DH (2008) The Eysenck Personality Measures: Fifty years of scale development. In: Boyle GJ, Matthews G, and Saklofske DH (eds.) The Sage Handbook of Personality Theory and Measurement, vol. 2. London: Sage. O’Connor KP (2008) Eysenck’s model of individual differences. In: Boyle GJ, Matthews G, and Saklofske DH (eds.) The Sage Handbook of Personality Theory and Measurement, vol. 1. London: Sage. Strelau J and Eysenck HJ (eds.) (1987) Personality Dimensions and Arousal. New York: Academic Press. Wilt J and Revelle W (2009) Extraversion. In: Leary M and Hoyle RH (eds.) Handbook of Individual Differences in Social Behavior. New York: Guilford. Zuckerman M, Kuhlman DM, and Camac C (1988) What lies beyond E and N? Factor analyses of scales believed to measure basic dimensions of personality. Journal of Personality and Social Psychology 54: 96–107.
20
Relevant Websites
18 Introverts
159
0
Introverts
Extraverts
Delayed test (after five weeks)
Figure 9 The performance of introverts and extraverts using different learning methods.
http://www.personality-arp.org/ – Association for Research in Personality. http://www.issid.org/ – The International Society for the Study of Individual Differences. http://www.smep.org/ – Society of Multivariate Experimental Psychology. http://www.aaas.org/ – American Association for the Advancement of Science. http://www.apa.org/ – American Psychological Association. http://www.spsp.org/ – Society for Personality and Social Psychology.