In referring to the classifications of temperament conceptualizations presented in Table 3.1, Kagan's theory can be described as a causal, monodimensional (typological) one tending to an emotion orientation, strongly concentrated on infancy and early childhood. In spite of Kagan's long career in developmental psychology, his interest in temperament, stimulated by his own experience as well as by ideas and findings in the literature, began less than two decades ago.
As delineated by Jerome Kagan (1989b, 1994; Ellis & Robbins, 1990), two events developed his interest in temperament research. First, Kagan and Moss (1962) conducted an ex post analysis of data from the Fels Longitudinal Study which comprised about 100 normal subjects followed from early childhood to adulthood. One of the conclusions from this analysis was that the only individual characteristic to remain stable across the developmental period under study was shy, timid behavior as opposed to outgoing, sociable behavior. Influenced by the environmentalists Zeitgeist Kagan and Moss interpreted these stable individual differences in terms of environmental factors. Second, a study conducted about 15 years later by Kagan, Kearsley, and Zelazo (1978) on 3- to 29-month-old children showed that Chinese children differ essentially from Caucasian children in some behavior characteristics that could not be explained in terms of rearing practices. Chinese children were quieter and more fearful than Caucasian children. This fact alone bolstered Kagan's belief that it was the biological factor that contributed to the behavioral differences between the two samples under study.
Before his involvement in studies on temperament, Kagan (1974) had already put forward a temperament hypothesis for the interpretation of changes in infants' behavior following presentations of discrepant objects. He suggested that temperament dispositions, probably genetically determined, contribute to whether a child confronted with unexpected events tends to react with smiling or crying.
At least three different lines of research influenced Kagan's interests in temperament: Jung's ( 1923) theory of biologically determined extraversion-introversion, Thomas and Chess's studies on temperament which had shown the uniqueness ofbehavioral style from early infancy, and animal research. Animal research had given evidence that the tendency to approach or to avoid novelty is biologically determined (Royce, 1955; Schneirla, 1965; Scott & Fuller, 1965).
According to Kagan (1989a, p. 668; Kagan, Snidman, Julia-Sellers, & Johnson, 1991, p. 332), the concept temperament refers to inherited profiles (categories, qualities, types) of behavior and biology which are present in the infant and which mediate different phenotypic displays depending on childhood experiences. Kagan (1989a, 1994), who does not deny the existence ofmany temperament categories or traits, developed his theory and research around one dimension, the extremes of which result in two qualitatively different categories—the inhibited and uninhibited temperament. These two categories refer to the child's initial reaction to unfamilar events (people, objects, situations). The encounter with unfamiliar events develops a state of uncertainty, which may be compared to a state of stress (Kagan, 1983), to which children react in different ways.
A child who is consistently shy, quiet, cautious, emotionally reserved, and timid, when confronted with unfamiliar events, is characterized as having an inhibited temperament. A child who under the same conditions is consistently sociable, talkative, affectively spontaneous, and minimally fearful, has an uninhibited temperament (Kagan, 1989a, 1989b, 1994; Kagan & Snidman, 1991; Kagan, Reznick, & Snidman, 1988). In an unselected Caucasian population there are about 10% inhibited children and 25% uninhibited children. The two categories are considered by Kagan and his collaborators not as a dimension but as the extremes of a continuum, with qualitatively different temperament characteristics.
The constructs inhibited and uninhibited to the unfamiliar refer to children who fall at the extremes of a phenotypic continuum from shyness and restraint to sociability and affective spontaneity. (Kagan, Reznick, & Gibbons, 1989, p. 838)
These constructs have behavioral, genetic and physiological patterns specific for the inhibited and uninhibited temperaments.
The two temperaments are relatively stable. Longitudinal studies conducted for several purposes in different cohorts of children aged from 2 months (Kagan & Snidman, 1991) to over 7 years (e.g., Kagan et al., 1988; Reznick et al., 1986) showed that the majority of inhibited and uninhibited infants selected from Caucasian samples did not change their temperament characteristics even in late childhood. The categories of inhibited and uninhibited temperament typical of children are somewhat analogous to the approach-withdrawal individual characteristics encountered in animals, especially in monkeys (e.g., Stevenson-Hinde, Still wellBarnes, & Zunz, 1980; Suomi, 1987), as well as to the extraversion-introversion dimension studied in adults (Kagan, 1989b; Kagan & Reznick, 1986).
Biological Bases of Inhibited and Uninhibited Temperament
Kagan (1982b, 1989a, 1989b, 1994) and his coworkers (Kagan, Reznick, & Snidman, 1987; Kagan et al., 1988) developed a theory of the biological bases of inhibition-uninhibition which had an essential influence on the majority of studies conducted in Kagan's laboratory.
Individual differences in the threshold of reactivity in the limbic system, especially the amygdala and hypothalamus, and the systems connected with the latter (pituitary-adrenal axis, reticular activating system, and sympathetic chain of the ANS) are regarded as the physiological bases of the two temperament categories. Inhibited children are characterized by lower reactivity thresholds in these systems.
The physiological signs that are characteristic of inhibited children could be due to tonically lower thresholds of reactivity in these brain structures. As a result, the inhibited children show increases in muscle tension, a rise and stabilization of heart rate, pupillary dilation, or increased cortisol to minimally unfamiliar or challenging events, whereas most children would not show these physiological reactions to the same relatively innocuous experiences. (Kagan et al., 1987, p. 1469)
In Kagan's (1994; Kagan et al., 1988) laboratory a whole set of physiological and biochemical markers of behavioral inhibition has been used. Heart rate, heart rate variability, heart rate acceleration, pupillary dilation, and urinary norepinephrine level reaction to psychological stress have served as indices of sympathetic reactivity. For measuring the activity of the hypothalamic-pituitary-adrenal axis cortisol levels from samples of saliva were taken. Skeletal muscle tension of the larynx and vocal cords measured by such indices as vocal perturbation and variability in the fundamental frequency of verbal utterances were used as measures of activity of the limbic system. Since the correlations between the physiological measures were rather low, varying between-. 20 and +.30 (with the exception of heart rate and heart rate variability), an aggregate index of physiological activity was often used as the physiological marker of inhibited temperament. "There was a substantial positive relation between this composite physiological index and the index of inhibition at every age (r = 0.70 with the index at 21 months, and r = 0.64 with the index at 7.5 years of age)" (Kagan et al., 1988, p. 170).
Behavioral inhibition as measured by Matheny (1989) in a longitudinal study of MZ and DZ twins aged from 12 to 30 months reared together showed that MZ twins as compared with DZ twins have significantly higher intrapair correlations. This speaks for the importance of the genetic factor in determining individual differences with respect to this temperament category. A study conducted by Emde and colleagues (1992) in which several temperament characteristics were measured in 200 pairs of 14-month-old twins has given support for this statement. Data from this study show that inhibited temperament as measured by means of behavioral observations had a heritability score of .62, one of the highest scores among the temperament characteristics under study.
Assessment Procedures Used in Kagan's Laboratory
In contrast to most researchers on temperament in children, Kagan and his coworkers did not use psychometric procedures for diagnosing the inhibited versus uninhibited temperament. When assessing children's temperament, according to Kagan (1994; Ellis & Robbins, 1990), one cannot rely on the reports of parents or teachers; therefore, in Kagan's laboratory observation of children's behavior under standardized conditions was the main method for studying temperament (see Garcia-Coll, Kagan, & Reznick, 1984; Kagan, 1989b; Kagan & Snidman, 1991; Kagan et al., 1989; Reznick et al., 1986).
The behavioral sessions during which temperament characteristics were assessed consisted of different episodes, depending on the child's age and the specific aim ofthe study. Separation from the mother, reactions to unfamiliar objects, such as unusual toys, robots, and sounds, and to an unknown person or group of persons (children, adults) were considered the most critical situations provoking behavior typical for the inhibited and uninhibited temperament. Observations and judgments of temperament characteristics were conducted by more than one examiner. The intercoder reliability measured for the behavioral indices of inhib-ited-uninhibited temperament, for example, latency to play, latency to first approach, and time proximal to mother, was mostly very satisfactory, reaching the value of .80 to .90. The behavioral study was almost always accompanied by psychophysiological recordings taken after or during the behavioral sessions. For school-age children classroom behavior recorded during different experimental settings was also taken as a temperament measure.
Since different situations provide a different set of incentives for the manifestation of inhibited-uninhibited behavior, an aggregate index of inhibition was often used by Kagan and his coworkers. For example, in a study conducted by Reznick and colleagues (1986) on 5.5-year-old children, average standard scores of the following five indices were used for calculating an aggregate index of inhi bition: peer play inhibition, laboratory inhibition, school inhibition, risk avoidance in laboratory settings, and number of times the child looked at the examiner). The correlations among these indexes are given in Table 3.4.
In general, assessment of inhibited or uninhibited temperament was based on a complex of behavioral and physiological indices. Standardized batteries consisting of age-specific laboratory situations were often supplemented by parental reports and maternal interviews.
Only a few studies have been conducted by Kagan and his associates in which the inhibited-uninhibited temperaments were related to behaviors that were not considered to be expressions ofthe two temperament types. In a recent study, Kagan and colleagues (1991) demonstrated a relationship between children's inhibited temperament and allergic symptoms in first- and second-degree relatives. Relatives of inhibited children reported more often, as compared with those of uninhibited children, the occurrence of such allergic symptoms as hayfever and eczema. This finding suggests a genetic background that relates shy and timid behavior in children to factors influencing immunological vulnerability to selected allergies.
Kagan and his coworkers considered cognitive functioning in unfamiliar experimental settings (objects and persons) as well as physiological patterns recorded during these states in terms of stressor and stress (Kagan, 1983; Reznick et al., 1986). Findings suggest that inhibited children, being shy and restrained, experience higher stress than uninhibited children under unknown and unfamiliar situations. According to Kagan the difference between inhibited and uninhibited children bears some resemblance to Pavlov's distinction between weak and strong types of nervous system. Inhibited children have a weak nervous system and uninhibited children a strong nervous system (Reznick et al., 1986, p. 677). It may be expected that studies relating Kagan's temperamental categories to behavior under
TABLE 3.4. Correlations among the Major Behavioral Indexes of Inhibition at
5.5 Years ofAge
TABLE 3.4. Correlations among the Major Behavioral Indexes of Inhibition at
5.5 Years ofAge
Was this article helpful?