Back to recent texts


Jerome Kagan, PhD

Harvard University, USA

April 2012, Rev. ed.


Although the term “temperament” does not have a consensual definition, most scientists would agree on the following sense meaning: A temperamental bias refers to distinctive patterns of feelings and behaviours that originate in the child’s biology and appear early in development.1

Subject and Problems

The biological foundation of a temperamental bias is usually, but not always, genetic. In some cases it is the result of prenatal stress or infection to the pregnant mother which affects the fetus. An important source of the biological foundation of a temperamental bias is variation in the concentration and density and location of the varied receptors for the large number of molecules that can affect brain function, including glutamate, GABA, dopamine, norepinephrine, serotonin, opioids, acetylcholine, corticotropin-releasing hormone, the sex hormones, vasopressin and oxytocin.2

This claim implies a very large number of temperamental biases, most of which are unknown. Because it is not yet possible to measure the neurochemistry that is the basis of a temperamental bias, scientists rely primarily on specific behavioural profiles. The behaviours in infants and young children that are most often attributed to a temperamental bias are unusually high or low levels of irritability, motor activity, smiling, ease of regulating these responses, and a consistent tendency to approach or to avoid unfamiliar people, objects, and places.

Because the likelihood of a behaviour being expressed is controlled, in part, by the local setting and it is difficult for scientists to observe children in a large number of settings that occur in the privacy of the home some psychologists administer questionnaires to parents asking them to describe the behaviours of their children in varied contexts and assume that these descriptions are, by and large, accurate. This assumption is not always valid because some parents are not sensitive observers of their children. The correlations between parental descriptions and direct behavioural observations of the comparable  behaviours are low to modest.3,4  

Therefore, it is better practice to combine parental reports with behavioural observations.

A temperamental bias does not determine a particular future trait because life experiences affect the bias and create an envelope of possible personality traits. By the second year the child’s temperamental biases and the products of experience have produced a seamless tapestry making it difficult to detect the early temperamental biases of most children because the same behaviour could be the partial result of a temperamental bias or the product of experience alone. Not all shy children inherit a temperamental bias favouring that quality. This means that discussions of adult temperaments, most often based on questionnaire data, are open to criticism.

Research Context

Two of the many potential temperamental biases that have been studied most extensively refer to the typical behaviours of one- and two-year-olds to unfamiliar people, objects, and situations. About 10 to 20% of middle-class American children are consistently shy with strangers and avoidant of unfamiliar objects and situations. These children are called “inhibited to the unfamiliar.” These children are contrasted with the 30 to 40% who show the complementary traits of approaching most unfamiliar events, called “uninhibited.”5 These proportions are apt to vary across class and ethnic groups.

Independent studies by Kagan and Fox2,5,6 have revealed  that these two styles of behaviour can be predicted by variation in vigorous motor behaviour and crying to unfamiliar visual, auditory and olfactory stimuli in four-month-old infants. Four-month-olds who show high levels of motor activity and distress, called high-reactive, are likely to become inhibited. Infants who show low levels of motor activity and crying, called low-reactive, are likely to become uninhibited in the second year. Thus, high and low reactive are regarded as two infant temperaments that are presumed to be due, in part, to different thresholds of excitability in the amygdala.

Research Results

Continued study of these two groups through age 18 has revealed several interesting facts.  White, middle-class, American adolescents who had been high reactive infants report more unrealistic worries than most adolescents, including visiting new places, meeting strangers, entering crowds, and brooding over possible harm to the self or a parent, and more frequent bouts of depression.7 These properties are rare in adolescents who had been low reactive infants. Second, measurements of the brain anatomy and function revealed that the 18 year old high reactives had a thicker cortex in an area of the prefrontal region of the right hemisphere that projects to sites that mediate defensive postures to threat as well as an amygdala that was more reactive to the unexpected appearance of pictures of unfamiliar scenes.8 This evidence supports the belief that high and low reactive infants were born with different neurochemistries in the amygdala.

Although adolescents who had been high-reactive infants are at a slightly higher risk than most of developing social anxiety or depression, but not at higher risk for phobias to animals or blood, and low-reactives are at a slightly higher risk for asocial behaviour, these outcomes require very specific rearing conditions. Most of the children in both groups will not develop any mental symptoms or psychiatric condition.

The power of a temperamental bias is to restrict the acquisition of a particular personality trait, rather than determine any one profile. The probability that a high-reactive infant will not become an extremely sociable, spontaneous, relaxed adolescent, free of unrealistics worries and low levels of autonomic and cortical arousal is very high. However, the probability that this category of child will be a quiet, anxious introvert with high levels of autonomic and cortical arousal is low (probably about 20 percent). Thus, the biology that is the foundation of a temperamental bias functions as a constraint rather than as a determining force.

Conclusion and Implications

Parents should appreciate that each of these two temperamental types has advantages and disadvantages in contemporary society. A technological economy requires many adults who like to work alone, including computer programmers, historians, bench scientists, and mathematicians. These vocations allow adults to work in environments where they can control the level of uncertainty and keep unanticipated interactions with strangers to a minimum. High-reactives tend to avoid risk and, therefore, are less likely to drive at high speeds, experiment with drugs, engage in sex at an early age, or cheat on examinations.

Low-reactive, uninhibited children enjoy their share of advantages. Sociability and a willingness to take career and economic risks are adaptive in contemporary American society. The adolescent who is willing to leave home to attend a better college or accept a more interesting job is likely to gain a more challenging position than one who stays close to home because of a reluctance to confront the uncertainties of a distant place.

As children develop, their temperament makes a more substantial contribution to their private feeling tone than to the public personality they display to others. The chronic possession of a relaxed or tense feeling tone requires a more substantial contribution from temperamental biases than does a sociable or shy posture with others.


  1. Rothbart MK. Temperament in childhood: A framework. In: Kohnstamm GA, Bates JE, Rothbart MK, eds. Temperament in childhood. Oxford, United Kingdom: John Wiley and Sons; 1989:59-73.
  2. Kagan J, Snidman NC. The long shadow of temperament. Cambridge, Mass: Harvard University Press; 2004.
  3. Seifer RA, Sameroff AJ, Barrette LC, Krafchuk E. Infant temperament measured by multiple observations and mother report. Child Development 1994;65(5):1478-1490.
  4. Biship GS, Spence SH, McDonald C. Can parents and teachers provide a reliable and valid report of behavioural inhibition? Child Development 2003;74(6):1899-1917.
  5. Kagan J. Galen's prophecy: temperament in human nature. New York, NY: Basic Books; 1994.
  6. Fox NA, Henderson HA, Rubin KH, Calkins SD, Schmidt LA. Continuity and discontinuity of behavioural inhibition and exuberance: Psychophysiological and behavioural influences across the first four years of life. Child Development 2001;72(1):1-21.
  7. Kagan, J. The Temperamental Thread. New York: Dana Press, 2010.
  8. Schwartz CE, Kunwar, P. S., Greve, D. N., Moran, L. R., Viner, J. C. et al., Structural differences in adult orbital and ventromedial prefrontal cortex predicted by infant temperament at 4 months of age. Archives of General Psychiatry 2010; 67: 78-84.

How to cite this article:

Kagan J. Temperament. In: Tremblay RE, Boivin M, Peters RDeV, eds. Rothbart MK, topic ed. Encyclopedia on Early Childhood Development [online]. Updated April 2012. Accessed November 19, 2018.