THE DEVELOPMENT OF VIOLENT BEHAVIOR
In this chapter, we will review some of the features of normal development drawn from the fields of psychology and evolutionary psychology that we believe to be most consequential for the development of the capacity for serious or frequent violent behavior. We do this in part to introduce the more advanced developmental constructs to readers from the criminology side, who may be unfamiliar with them. We also do this to provide background so that the reader can better understand our reasoning for selecting the differential predictors examined in depth in later, substantive chapters. Thus the processes and constructs we feature in this chapter are incorporated to justify our decision to focus on intelligence, education factors, parental attachment, parental warmth and rejection, and child abuse in Chapters 4-8.
Stability and Continuity of Aggression
Stability in aggression and correlations between early aggression and later delinquency have been studied extensively. Olweus (1981) avered that “... aggressive and related ‘acting-out’ behavior shows a high degree of stability often over long periods of time” (p. 152). Pettit (1997) concludes that it “is now well accepted, with several reviews indicating a level of trait-like stability in aggression that perhaps is rivaled only by the intelligence quotient” (p. 288). Significant correlations between earlier measures of aggression and later ones, or evidence of a subgroup of chronic offenders, have been found in virtually all of the major longitudinal studies of aggression (e.g., Ellickson & McGuigan, 2000; Farrington, 2003; Hamparian, 1987; Huesmann, Eron, Lefkowitz, & Walder, 1984; Jessor, Donovan, & Costa, 1991; Kolvin, Miller, Fleeting, & Kolvin, 1988; Kosterman, Graham, Hawkins, Catalano, & Herrenkohl, 2001; Landsheer & van Dijkum, 2005; Laub, Nagin, & Sampson, 1998; Loeber, Stouthamer-Loeber, Van Kammen, & Farrington, 1991; Magnusson, Stattin, & Duner, 1983; Moffitt, 2003; Moffitt, Mednick, & Gabrielli, 1989; Patterson, 1992; Pulkkinen & Pitkanen, 1993; Robins, 1978; Sampson & Laub, 1993; Tracy, Wolfgang, & Figlio, 1990; Tremblay & LeMarquand, 2001; Tremblay, Masse, Perron, Le Blanc, Schwartzman, & Ledingham, 1992; Werner & Smith 1992). Some of the data suggest continuity of aggressive tendencies well into adulthood (e.g., Huesmann, Dubow, Eron, & Boxer, 2006). Pettit (1997) summarizes a variety of studies and estimates that about half of any sample of antisocial children will become antisocial adolescents and half of those will become antisocial adults.
Stanger and colleagues have made the case that aggressive (violent) behavior is more stable than general delinquent behavior (e.g., McConaughy, Stanger, Achenbach, 1992; Stanger, Achenbach, & Verhulst, 1997; Stanger, MacDonald, McConaughy, & Achenbach, 1996). Others have reported this as well. Loeber (1982) showed that serious antisocial behavior is more stable from childhood to adulthood than minor delinquency. Broidy and colleagues found that a diagnosis of “physical aggression” in childhood was significantly associated with later violent and nonviolent delinquency; the finding was more consistent for the violent than the nonviolent (significant at one more site) (Broidy, Nagin, Tremblay, Brame, Dodge, Fergusson et al., 2003). Huesmann et al. (1984) report that childhood aggression was by far the strongest predictor of age 30 criminality in their longitudinal data set (p. 203), and Kosterman et al. (2001) found significant correlations between childhood fighting and later adolescent violence. Teacher ratings of aggression at ages 8-10 were significantly correlated with conviction for violence in adulthood in the Cambridge sample (Farrington, 1991) and Lynam, Piquero, and Moffitt (2004) report that violent subjects, when compared to nonviolent delinquents, had higher levels of previous childhood conduct problems.
Many authors accept the premise that early developmental factors are important in the causal chain that results in adolescent and adult violent behavior.