INTRODUCTION

One of the important contributions to the understanding of antisocial and aggressive behavior has come from research that spanned the areas of animal research (Cairns, Hood, & Midlam, 1985), genetic processes (Cairns, McGuire, & Gariepy, 1993), attachment (Cairns, 1966), social interactions (Cairns & Valsiner, 1984), and developmental epistemology (Cairns, 1990). Unlike much other research, this integrative vision of science was not blinded by the interdisciplinary rivalries that have both pervaded and impeded the study of antisocial behavior. Instead, by viewing genetic, biological, psychological, and social influences as equal partners in the explanation of antisocial and aggressive behavior, more significant and lasting advances can be made in this field.

In this context, important progress has been made in delineating replicable psychosocial risk factors for antisocial and violent behavior (Farrington, 2000; Loeber & Farrington, 1998; McCord, 2001; Rutter, Giller, & Hagell, 1998). Within the past 15 years, important progress has also been made in uncovering biological risk factors that predispose to antisocial behavior (Lahey, McBurnett, Loeber, & Hart, 1995; Susman & Finkelstein, 2001). Despite this progress, we know surprisingly little about how these different sets of risk factors interact in predisposing to antisocial behavior. Furthermore, although passing heuristic and theoretical references are frequently made to such interactional influences, there are remarkably few investigators who are conducting serious empirical research on this interface in humans (Raine, Brennan, & Farrington, 1997).

INTRODUCTION

Violence is a national health problem. Understanding the etiology and pathophysiology that predispose certain individuals to behave in an inappropriate, excessively aggressive way is critical to reducing the incidence of personal violence in the United States. It is clear that traumatic events in early life make children more vulnerable to future stressors and enhance the probability of antisocial behaviors (Luntz & Widom, 1994). Children exposed to early stressful environments and maltreatment prior to kindergarten are more aggressive and have more social problems than control children (Lansford et al., 2002; Sanson, Smart, Prior, & Oberhlaid, 1993). Children exposed to early physical and emotional neglect may develop learning patterns that affect the interpretation of hostile social information, resulting in inappropriate aggressive behavior and general conduct disorder (Dodge, Bates, & Pettit, 1990). Furthermore, children with conduct problems are at risk for alcohol and drug abuse, a predisposition exacerbated by the fact that alcoholism and drug-taking foster antisocial behaviors and violence (Ito, Miller, & Pollock, 1996).

Although psychosocial developmental studies on children have provided insight into the long-term behavioral consequences of early abuse and alcoholism, the biological changes that accompany exposure to these risk factors are not well understood. For ethical reasons, there can be no prospective human studies looking at changes in neurobiology and neuroendocrinology in response to early life traumas. Therefore, biological data must be gleaned from animal studies that try to model certain aspects of the human condition.

INTRODUCTION

Professor Robert B. Cairns was a giant in the field of developmental science. His theoretical vision and his empirical rigor and persistence drove a lifelong research and academic career that uniquely helped shape this emerging, dynamic interdisciplinary field. He was a remarkable teacher, researcher, and scholar who profoundly influenced countless students and colleagues and whose legacy will long be with us.

I first got to know Bob Cairns during a mutual stint at the Center for Interdisciplinary Studies (ZIF) at the University of Bielfeld in Germany during the fall and winter of 1977–78. We were both at the ZIF as part of a 9-month-long program entitled “Early Development in Animals and Man” organized by the late ethologist Professor Dr. Klaus Immelmann, who had brought together a group of prominent researchers, representing a wide range of disciplines, who studied developmental phenomena both in humans and in a variety of other mammalian and avian species. I had known about Bob Cairns and his research since my very first days in graduate school at the University of Wisconsin – at that time Jim Sackett was telling all the graduate students at the Primate Laboratory about this guy in Indiana who hauled sheep to his lab in back of his station wagon in order to run studies that demonstrated the power of contiguity conditioning in establishing social bonds.

A BIOSOCIAL PERSPECTIVE ON AGGRESSION

Problem aggression in young children is viewed as more biologically driven than antisocial patterns that appear in adolescence and thought to be influenced mainly by peer socialization (Moffitt, 1993). Early-onset aggression is more chronic, serious, and linked to long-term maladaptive outcomes. By virtue of disposition or temperament, some young children are more likely than others to behave in ways that reflect disregard for the rights and welfare of others. Several constitutional factors have been implicated in early externalizing problems. These include neuropsychological problems, hyperactivity, and difficult temperament. Different antecedents and outcomes of early versus delayed onset of antisocial behavior have been identified (e.g., Moffitt & Caspi, 2001). Even early-onset aggression has poor parenting correlates, however, suggesting a complex interplay of biological and environmental processes from the start.

In this chapter we focus on early-appearing aggression and the ways in which socialization and child sex interact with it to alter forms of expression and developmental course. We consider, as well, early-appearing prosocial behaviors that may deter aggression. We draw on two high-risk longitudinal studies that have benefited from the teachings of Robert Cairns. He and his colleagues have been at the forefront in guiding the development of theories, research designs, and analytic approaches that illuminate bio-social interactions in different species and at different points in human development (Cairns, 1997, 2000; Cairns, Cairns, Neckerman, Ferguson, & Gariépy, 1989; Magnusson & Cairns, 1996).

Adolescence has fascinated developmental scholars because the transition into adolescence involves biological, psychological, and social changes (Graber & Brooks-Gunn, 1996). At the same time, adolescence has been a focus for research on psychopathology as rates of several disorders increase dramatically during this time period. Most notably, the past few decades have witnessed volumes of studies and theories on adolescent depression, conduct disorder, and subclinical psychopathology (Steinberg & Morris, 2001). Many of these studies have sought to understand the confluence of bio-psychosocial developmental factors that result in the emergence of serious behavioral and emotional problems. In this chapter, we consider several bio-psychosocial models that have been used to explain changes in internalizing and externalizing behaviors during adolescence. Examples from our own work highlight the role of pubertal development in understanding behavioral plasticity during adolescence.

Discussions of plasticity in developmental processes have frequently focused on early development and gene–environment interactions in understanding development and behavior (Baltes, Lindenberger, & Staudinger, 1998). Despite a focus on the early periods of development, the notion that adaptation occurs in neural and behavioral development throughout life has been a cornerstone of life span developmental perspectives (Baltes et al., 1998; Cairns, 1998). Recent studies in neuroscience have demonstrated that new neural connections continue to be made across the life span (e.g., Bruer & Greenough, 2001) and specific changes in the prefrontal cortex and limbic regions of the brain occur during adolescence (see Spear, 2000 for a review).

INTRODUCTION

Increasing trends for suicidal ideation and attempts among adolescents highlight the need for identifying risk factor profiles and interventions. In a recent survey study we reported that depression, anger, number of friends, and marijuana contributed to 66% of the variance on suicidal ideation in adolescents. In other studies with depressed adolescents we were able to reduce depression and suicidal ideation by massage therapy. The model explored in this paper on adolescents' suicidal ideation and suicidal attempts is (1) that depression, aggression, and body dissociation (physical anhedonia and higher thresholds to stimulation and pain) that may have derived from early touch deprivation/abuse may mediate suicidal ideation and attempts; and (2) that suicidal ideation and attempts might be reduced by massage therapy, a treatment that has already been effective with depressed adolescents. Underlying the risk factors for suicidal ideation and attempts are abnormal physiological factors including EEG alpha asymmetry and biochemical imbalances including depressed serotonin and elevated cortisol levels. Thus, some of the intervening mechanisms between touch deprivation/abuse and suicide may be the physiological (EEG asymmetry) and biochemical imbalances (depressed serotonin and elevated cortisol levels) that accompany depression and self-destructive behavior. Massage therapy may reduce suicidal ideation/behavior via reducing depression and cortisol levels and elevating serotonin levels and thereby reducing self-destructive behavior.

TOUCH DEPRIVATION AND AGGRESSION AGAINST SELF (SUICIDE) AMONG ADOLESCENTS

The critical importance of researching adolescent suicide is underscored by the fact that suicide is the biggest killer of adolescents in the United States (Madge & Harvey, 1999), and the United States has one of the highest suicide rates among industrialized countries (Weissman et al., 1999).

Aggressiveness is a major individual difference, or “personality,” factor in most species of animals (Gosling, 2001), including monkeys (Chamove, Eysenck, & Harlow, 1972) and dogs (Svartberg & Forkman, 2002). Strains of mice and rats differ markedly in aggressiveness. In nonhuman animals, aggression is classified by the provoking conditions including: pain, fear, maternal protection of the young, an intruder in the animal's territory, mating competition, and prey to a predator. The last of these, called “predatory aggression,” differs from intraspecies or “defensive aggression” in that it does not require anger or even strong emotion. If anything, the excitement of a predator is more like the positive excitement during sensation seeking in humans. Even well-fed cats seem to enjoy hunting and when there is no real prey they play at stalking and pouncing. Human predators, however, are different in that they stalk their own species, but some of the sensation seeking motivation may be the same as in other species. Sexual sadists are generally unemotional and detached but get pleasurable excitement from the suffering of their victims (Cosyns, 1998). Humans who hunt other species are most like predator animals. Like cats, they hunt for the challenge and pleasure of the kill even when they are not at all hungry.

Human aggression is classified by the form it takes. Human aggressiveness involves three types of expression: the behavioral, as in physical or verbal aggression; the emotional, as in anger; and the attitudinal, as in hostility.

This chapter begins with a review of some of the definitions of the trait of extraversion found in Chapter 1. Eysenck (1947, 1967) traced the history of the introversion-extraversion construct back to Galen's 2nd-century types. These types were based on the “humors” of the body: phlegmatic, sanguine, melancholic, and choleric. The ancient Greeks regarded these as pure types rather than bipolar traits. In the 19th century, psychiatrists described the types in terms of psychopathologies: schizophrenic versus manic-depressive (Kraepelin, 1899) and psychasthenia versus hysteria (Janet, 1907), regarding the abnormal manifestations as the extremes of normal personality types. Jung (1933) used the terms introversion and extraverson to describe what he believed to be the major dimension of personality: an orientation to the external world of people and objects versus a primary direction of interest and motive to the internal world. He described extreme types, but he recognized that “the normal man is, by definition, influenced as much from within as without,” implying a continuum or normal dimension of personality. Freud (1920), in contrast to Jung, regarded introversion as a precursor of neurosis. The association between introversion and neuroticism was reflected in some of the early American questionnaires that showed high correlations between introversion and neuroticism. The idea of a normal introvert was not widely accepted in the extraverted American culture.

Eysenck's construct was closer to Jung's, in that he regarded introversion extraversion as a normal dimension independent of and uncorrelated with the major dimension of neuroticism.

As in the 1991 first edition of this book, I begin with a discussion of temperament and personality and their basic dimensions. This is a necessary first step in a top-down approach because one cannot begin a levels analysis of psychobiology without a classification of traits at the top level. It would be like a science of astronomy without distinctions between planetary bodies such as asteroids, planets, stars, and galaxies; geology as a science of “rocks” sorted by size; or biology that makes distinctions only between two-legged and four-legged creatures – putting humans and chickens in the same category. Classification of phenomena is basic to any science. Without it, all is chaos.

There is a difference, however, between the classification of behavioral traits and other types of scientific classification. We are not defining “types” in the sense of clear-cut assortment of individuals. The concept of continuous, normally distributed trait dimensions is not widely understood outside of psychology. I purposely use the labels “high” and “low” sensation seekers, rather than type terms, such as “Big-T” and “Little-t,” to define persons falling near the extreme ends of the continuum on this trait. Types, however, may be defined from a combination of independent dimensions as particular combinations of high or low scores on these dimensions in the same way that syndromes of psychopathology are defined by particular combinations of symptoms.

In the earlier volume, I discussed certain issues about traits that preoccupied personality researchers in the 1970s.

ADVANCES IN METHODOLOGIES

Ethical constraints restrict the scope of psychobiological methods used with living human beings. In other species, one can surgically or chemically destroy selected parts of the brain and observe the effects on behavior. Autopsy can be used to verify the brain alterations. One also can stimulate selected areas of the brain electrically or chemically through canulas to the brain areas. Brain stimulation in humans is usually done through external stimulation of the senses with stimulation varying in intensity, quality, or content. Chemical manipulation is done through infusion or ingestion of drugs, which can produce transient effects in these systems. Most psychopharmacological work with humans is done on abnormal populations with limited generalizability to normal personality variants.

Measurement of responses of specific brain areas to stimulation may be done directly, using implanted electrodes. Until recently, brain study in humans was limited to amplified electrical potentials from the scalp (EEG and cortical evoked potentials) indicating areas of cortical reactivity imprecisely located in areas under the surface electrodes. Newer brain imaging methods such as positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) allow us to view activity at all levels of the brain at specific cortical and subcortical loci.

The peripheral autonomic system has been studied using physiological methods for measuring cardiovascular (heart rate, blood pressure, blood flow), respiratory, and electrodermal (skin conductance) methods. These methods have been widely used in studies of emotion, perception, and cognition, and some have used them to study personality traits such as neuroticism, and emotional traits such as anxiety and hostility.

The diverse array of traits in the title for this chapter reflects the disagreements among personality classifiers as to the nature of “personality in the third dimension.” The Eysencks proposed the label of “Psychoticism” for this higher order factor. Both Block (1977) and Zuckerman (1989) suggested that if a clinical term was appropriate for the scale called “Psychoticism” (P) (Eysenck & Eysenck, 1976), it should be “Psychopathy” (Antisocial Personality Disorder in the current psychiatric nomenclature). The subtraits of P described by Eysenck and Eysenck (1985) include: aggressive, cold, egocentric, impersonal, impulsive, antisocial, unempathic, creative, and tough-minded. With the possible exception of creativity, these are all traits of the psychopath rather than the psychotic, particularly if one includes the endogenous depressives among the psychotic. The highest scoring clinical groups among males and females are: prisoners, drug addicts, alcoholics, and personality disorders (Eysenck & Eysenck, 1976), all of which contain high proportions of antisocial personality disorders. Psychotics, among females at least, score low on the P scale relative to other clinical groups. The concept of a P dimension came from ratings of clinical groups that largely included psychotics and neurotics, but not psychopaths (S. B. G. Eysenck, 1956), and this may account for the conceptualization of the P dimension as psychoticism. Revisions of the P scale have tended to eliminate most of the psychotic sounding items except for a few with a mildly paranoid flavor. Psychopaths often voice the opinion that everyone is “against them.”

Wilson (1998) used the term consilience, defined as that quality of science that combines knowledge across disciplines, to create a common background of explanation. More complex levels of phenomena can be described in terms of simpler phenomena. Personality psychology extends from social phenomena and what we call “personality traits” at the most complex end down to genes and their variations at the simplest level as shown in Figure 7-1. Of course, there is nothing simple about the genome, and we have only the beginning of a science of molecular behavioral genetics. Wilson sees reductionism as the goal of science:

Personality constitutes a daunting challenge for this kind of reductionism. It can be analyzed in terms of distal to proximal influences along biological and social pathways as shown in Figure 7-1. Both pathways have their origins in evolutionary history.

The first edition of Psychobiology of Personality, published in 1991, provided a comprehensive treatment of a subject about which most psychologists have only a fragmentary knowledge, if any. Most behavior geneticists know a lot about the genetics of personality but little about its psychophysiology or psychopharmacology. Conversely, most psychopharmacologists know a lot about the psychopharmacology of personality and psychopathology but little about behavioral genetics. Most social and many personality psychologists know a lot about traits and their assessment but little about the psychobiology of traits. The first edition was based on a levels approach to personality (see Figure 7-1) that attempts to explore all levels of personality, from the genetic to the trait with stops at the neurological, biochemical, physiological, conditioning, and behavioral levels of explanation. The goal was not reductionism but a connectivism that respects the mode of explanation at each level and attempts to understand, as much as possible, the connections between phenomena at all levels. Although causation is usually assumed to work up from the more basic levels to the behavioral ones, it can work in the opposite direction. Only experimentation can decide the issues of causation; correlation cannot do this, as we patiently explain to students in introductory psychology (although clinicians sometimes forget this).

The psychobiology of personality has been explored by “top-down” and “bottom-up” approaches. The top-down approach is to identify basic personality traits in humans and then see what correlates of these traits can be found in physiology, biochemistry, neurology, and genetics.

The concept of neuroticism as a temperament can be traced back to the ancient Greek physicians Hippocrates and Galen. Their descriptions of the melancholic temperament included symptoms of depression and anxiety and social withdrawal. Both attributed the disorder to both an organic basis (an excess of black bile) and to an environmental one. Hippocrates said that melancholia could be precipitated by loss of love or conflict and guilt, whereas Galen, anticipating Freud by 18 centuries, suggested it could be due to blocked sexual outlet. In his early theory, Freud also ascribed pure anxiety states (panic disorder) to the tension generated by sexual frustration.

Freud regarded neurosis on a continuum lying between neurotic character and psychosis, depending on the extent of regression due to conflict between ego and id impulses. The concept of neurosis as a disorder suggested that it could be a temporary condition rather than a basic personality dimension. Freud included hysteria among the psychoneuroses, attributing differences between it and disorders such as anxiety and obsessive-compulsive to different levels of psychosexual fixation.

During the 19th century, a number of psychiatrists, including Jung, distinguished between “psychasthenic” and hysterical disorders on the basis of a second dimension of personality, extraversion-introversion. Eysenck (1947), whose early research used mental patients as subjects, regarded the “dysthymic” (anxious and depressed) types as introverted neurotics, and the hysteric and psychopathic types as extraverted neurotics. He defined neuroticism as a basic dimension of personality that could be regarded as going from extreme emotional stability at one end to emotional instability at the other.