![]()
PART I
Introduction to the Jyväskylä Longitudinal Study of Personality and Social Development
![]()
1.1A historical context for the study of aggression
Longitudinal studies were rare in the mid-1960s, but there were pioneering studies that had been initiated in the 1920s in California such as the Gifted Children Study by Terman (Terman & Oden, 1959) and two Oakland studies, the Guidance Study by Macfarlane (1938) and the Oakland Growth Study by Jones (1938). Early longitudinal studies were often motivated by educational goals; that is, the possibility of advancing or steering child development. I had been interested in human development since childhood, when I had amused myself by trying to imagine what kind of adults the children whom I knew were going to become and, conversely, what kind of children the adults I knew had been when they were young. Through reading the longitudinal study conducted at the Fels Research Institute by Kagan and Moss (1962), I became interested in personality development. Children had been followed from birth up to adolescence and on into early adulthood. The study resulted in new insights into the stability of behavior. A child’s behavior during the period from 6 to 10 years of age predicted at least moderately some aspects of that person’s behavior during early adulthood (age 24).
Although I was familiar with the longitudinal study by Kagan and Moss, it did not occur to me to start a longitudinal study when I was planning my doctoral dissertation during the second half of the 1960s. My dissertation was a so-called cross-sectional study, because I collected data only at one time point. For the study of individual development, a longitudinal study is needed. The definition of a longitudinal study is that data are collected about the same set of individuals at least twice after an interval of time. In fact, I (my name was Lea Pitkänen until 1975) wrote in the Discussion section of my dissertation: “A longitudinal study would make it possible to examine the stability of the individual patterns of behaviour” (Pitkänen, 1969, p. 190). At that time, I could not have anticipated staying on the same track for over 40 years.
What were these individual patterns of behavior which I found so compelling that I felt the need to study their stability over time? To answer this question fully, I need to first describe the historical setting in which I was planning the dissertation. What were the theoretical thoughts and research instruments that were available to me as a student during my studies from 1958 onwards? The small Department of Psychology at the University of Jyväskylä offered only a few courses to students. Studying mostly consisted of reading textbooks and writing examinations based on the information they contained.
Professor Martti Takala, the only professor at the Department of Psychology, was interested in both experimental personality psychology and differential psychology; the latter refers to the study of individual differences. The textbooks he chose for students reflected his interests. I enjoyed reading the Swedish textbook Differentiell psykologi by Ekman (1952) so much that after the examination, my professor kindly said to me that it was not necessary to learn the content of the textbooks by heart. My first English textbook was Child Development by Olson (1949), which I read slowly with the help of a dictionary because I had only taken a short course in English at school. The book greatly affected my orientation to the subject of childhood. The most interesting textbook was, however, Theories of Personality by Hall and Lindzey (1957). It included in-depth descriptions of the grand personality theories by Freud, Lewin, Allport, Rogers, and many others, and the first factor theories of personality by Eysenck and Cattell. Regarding the latter, I was impressed by Eysenck’s theory with his bipolar contrast between introversion and extraversion which describes differences in temperament between individuals.
In the field of learning theory, the prevailing theory at that time was a behavioristic stimulus response orientation that Dollard and Miller (1950) applied to personality. A basic tenet of this theory was the formation of behavioral habits as a result of reinforcement. The theory introduced me to the construct of a drive as a strong stimulus which impels the individual to behave or respond in a certain manner until the stimulus is reduced or eliminated. According to this theory, learning consists primarily of developing efficient means of reducing drive stimuli. The stimulus response theory was mainly based on animal experiments. The emergence of social learning theory as a result of the work by Bandura and Walters (1963) was a groundbreaking experience for me. It opened a new perspective on social learning. Later, the theories of motivation (e.g., Atkinson, 1964) and clinical child psychology also stimulated my thinking when I had to learn about them as part of my teaching responsibilities.
As a research assistant at the Department of Psychology, I collected data from children concerning their achievement-, affiliation-, and aggression-oriented behaviors and participated in experimental research on personality (Pitkänen, 1963b; Takala & Pitkänen, 1963). I was puzzled and dissatisfied by the inconsistent findings that different measures resulted from a certain category of behavior. I became skeptical about interpreting aggression or achievement as a need or a personality trait. I found that they were not unified constructs. I became interested in the multidimensionality of personality constructs and the dependence of behavior on situational factors. As aggression was more visible and measurable than achievement, I chose aggression for a closer examination in my doctoral dissertation. I began my independent research work for my dissertation in 1964, inspired by the book on aggression by Arnold Buss (1961). At that time, I was a mother to two daughters born in 1961 and 1964, which also motivated my study on children.
I conducted a Kindergarten Study with 216 six-year-old boys whose behavior kindergarten teachers (26) observed and rated using 32 items on aggression (Pitkänen, 1966, 1969, p. 29). For sampling these variables, I devised a “cone-shaped biscuit” model to describe different components of aggression (Figure 1.1). One dimension consisted of the intensity of aggression (the height of the cone). Two other dimensions were the motivational sequence of an aggressive act, and the direction of aggression (diameters of the cone). Different modes of aggression (physical, verbal, facial) were also depicted in the descriptive model of aggression as vectors.
FIGURE 1.1A descriptive model of aggression.
Source: Pitkänen (1969: Figure 1, p. 29). Reproduced with permission.
The motivational sequence referred to the initiation of an aggressive act (offensive meaning proactive) or a response to an aggressive act (defensive meaning reactive). The direction of aggression referred to whether aggression was targeted at another person directly or indirectly via mediating events, for instance, by gossiping about the target. These components can be used for the analysis of aggressive acts. For instance, in the act, “He hurts somebody without any reason, e.g., by tripping, pulling hair, pinching, or striking in passing,” the purpose of the act is offensive (proactive), the direction is direct, the mode is physical, and the intensity can be observed.
The results confirmed the relevance of the dimensions for individual differences. They also indicated that the most differentiating quality in boys’ aggressive behavior was the purpose of the aggression. The children who were perceived as being most aggressive were proactively aggressive, and they generally also defended themselves if attacked. Aggression for the purpose of self-defense was to some extent acceptable but it was critical that the intensity of the defense matched with the intensity of the offense which triggered it. Tremblay (2000) acknowledged that the differentiation between offensive and defensive aggression was a pioneering piece of research that other researchers have taken further using the terminology proactive and reactive aggression.
1.2Positive behavior in an impulse control model and the goals of the study
1.2.1The search for alternatives to aggression
The Kindergarten Study previously described revealed that only a few children were frequently aggressive. This observation stimulated my interest in alternative, nonaggressive behaviors which are used to cope with conflict situations, and led me to develop a new study that formed the second part of my dissertation (Pitkänen, 1969). The JYLS was built on the basis of this second cross-sectional study.
First, I needed to learn more about individual differences in social behavior. In the literature, I discovered that researchers were more interested in problem behaviors than in positive behaviors. The concepts used to describe positive behaviors in children were vague, such as describing someone as “a nice child,” but what does one actually mean by this? The boom in studying prosocial behavior did not begin until the late 1970s (e.g., Mussen & Eisenberg-Berg, 1977; Staub, 1979). The orientation toward positive psychology only began at the beginning of this millennium (Seligman & Csikszentmihalyi, 2000). In 1967, there was no model which conceptualized positive alternatives to aggression and, therefore, I had to develop it.
From animal studies, a nonaggressive option to “a fight” was “a flight,” as described in Miller’s (1959) theory of approach-avoidance conflict presented within the conceptual framework of the stimulus response behavior theory. In a conflict situation, there are two tendencies, either to approach the situation or to avoid a negative stimulus. The approach tendency is sustained by a drive stimulus that has its origin in the internal physiological condition of an organism. The avoidance tendency is motivated by fear and is an acquired drive. In human psychology from the point of view of psychodynamic theories, aggression was seen as an expression of aggressive energy and its display functioned in a “cathartic” way while reducing the energy of the aggression (Feshbach, 1964). Nonaggression was generally referred to as an “inhibition of aggression,” and it was not viewed in a positive light due to the assumption that inhibition might result in a personality pathology. Aggressive energy could be channeled by a great variety of acts, even by nonaggressive acts. Thus, it became difficult to assess which behavior was an expression of channeled aggressive energy and which was motivated by nonaggressive urges or reasons.
Researchers had tried to classify nonaggressive responses to frustrations using constructs, such as submissiveness and sympathy (Wittenborn, 1956); withdrawal (instigation alteration), limitation (frustration depreciation), and substitution (goal response alteration) (McClelland & Apicella, 1945); and the approval motive that makes the behavior of approval-dependent individuals conforming (Crowne & Marlowe, 1964). Feshbach (1964) put forward the idea that the reduction of anger and the aggressive drive might be accomplished by displaced aggression, changing the stimulus condition or its meaning, and mediating responses that are incompatible with anger. Lazarus’ (1966) analysis of coping process was along similar lines and this made an important contribution to my thinking. He distinguished three major coping reaction patterns: direct actions, defensive reappraisals, and anxiety-reaction patterns. Direct actions contained actions aimed at strengthening the individual’s resources against being harmed, attack patterns, and avoidance patterns. Defensive reappraisal involved thought processes of many kinds and anxiety-reaction patterns appeared as a threat reaction when no clear action tendency was generated.
In 1967, after an intensive period of thinking, I gained an insight into a way of describing differences in children’s behavior covering both aggressive and nonaggressive behaviors. I reasoned that the human brain allows for more variation in social behavior than just the “fight or flight” response and inhibition of anger. I speculated that a characteristic of human beings is that they can self-reflect their own behavior, intentions, and emotions and exercise control over how these are expressed. I concluded by considering the increased recognition of the interface between emotion and cognition (Schachter & Singer, 1962), that it is a human being’s capacity for cognitive control over his or her emotional behavior which makes...
