Unintended Victims: Diagnosis & Treatment of Children of Domestic Violence

There has been much writing, but little research, on the childhood precursors of adult violent behavior. The triad of enuresis, pyromania, and cruelty to animals (e.g., Heilman and Blackman 1966) is probably the most frequently cited set of predictors of this sort. One survey (Justice, Justice and Kraft 1974) reviewed 1,500 references to violence in psychiatric literature, interviewed over 750 professionals who dealt with violent persons, and retrospectively analyzed over 1,000 clinical cases to ascertain the most cited childhood predictors of adult violence. The authors reported that the four “early warning signs” were fighting, temper tantrums, school problems, and an inability to get along with others. The child, in other words, is indeed father or mother to the adult.

Based on discussions with large groups of psychiatrists and psychologists, Goldstein (1974) concluded that the “agreed upon” predictors of violence were “a childhood history of maternal deprivation, poor father identification, or both; nocturnal enuresis; possibly fire setting; violence toward animals; and brutalization by one or both parents” (p. 27). Diamond (1974) comments that the conclusion of the clinicians cited by Goldstein represents the sum total of our present “scientific” knowledge concerning predictive factors of murderous violence.

Yet I have repeatedly found some, and sometimes all of these predictive factors, in individuals who have never committed even the slightest harmful act, let alone assault or murder. And I have examined offenders who have committed the extraorçlinarily brutal acts of great violence and lethality who possessed none of these factors. (Diamond 1974, p. 444.)

One of the most famous studies of the childhood correlates of later criminal behavior is Unraveling Juvenile Delinquency, published by Glueck and Glueck in 1950. While not concerned specifically with violent criminality, the Gluecks claimed that three factors—supervision by the mother, discipline by the mother, and cohesiveness of the family—were predictive of later crime in young adolescent boys. This research is among the most methodologically criticized in all of criminology, and there appears to be a consensus that the practical utility of the Glueck factors in predicting criminality is marginal at best.

Lefkowitz, Eron, Walder, and Huesman (1977) published the results of a longitudinal study entitled Growing Up To Be Violent. This research followed a sample of over 400 males and females in Columbia County, New York, from ages 8 to 19. They used peer ratings, parent ratings, self-report, and a per­sonality test to measure “aggressive behavior.” Lefkowitz and his coworkers found that “aggression at age 8 is the best predictor we have of aggression at age 19, irrespective of IQ, social class, or parents’ aggressiveness” (p. 192). Several other variables, among them the father’s upward social mobility, low identification of the child with his/her parents, and a preference on the part of boys for watching violent television programs, were statistically significant predictors of aggression at age 19. Boys who, in the third grade, preferred television programs such as “Gunsmoke” or “Have Gun, Will Travel” were rated by their peers 10 years later as three times as aggressive as boys who, in the third grade, preferred “Ozzie and Harriet,” “I Love Lucy,” or “Lawrence Welk.” What is not clear from the study is why an 8-year-old boy would prefer “Lawrence Welk” to “Have Gun, Will Travel” in the first place.

McCord (1979) has reported on a 30-year follow-up of 201 boys who participated in the Cambridge-Sommerville Youth Project between 1939 and 1945. She found that 36 percent of the incidence of later violent criminality could be accounted for by childhood predictive factors. “The boys who lacked super­vision, whose mothers lacked self-confidence, who had been exposed to parental conflict and to aggression were subse­quently more convicted for personal crimes” (McCord, 1979, p. 1481).

In what has become the most influential criminological research of the past decade (Geis and Meier 1978), Wolfgang et al. (1972) obtained information on all boys born in Philadelphia in 1945 who lived there between their 10th and 18th birthdays. Of the 9,945 boys studied, 3,475, or 35 percent, had at least one recorded contact with the police by age 18. Wolfgang et al. found that the variables of race and socioeconomic status (SES) were most strongly associated with reported delinquency: 29 percent of the whites, but 50 percent of the nonwhites, and 26 percent of the. higher SES, but 45 percent of the lower SES boys had an offense record.

“Chronic” offenders were defined as those who committed five or more violations. Six hundred and twenty-seven boys, 6 percent of the sample and 18 percent of the total number of offenders, were responsible for over one-half of all offenses committed.

Chronic offenders in the cohort had a greater number of residential moves, lower IQ scores, a greater percentage classified as retarded, and fewer grades completed than did either the nonchronic or the one-time offenders, even when race and SES were held constant (p. 248).

Wolfgang (1977) has updated his research to include data on the subjects up to age 30. Only 5 percent of the subjects had an arrest record only as an adult (i.e., after age 18 but not before).

While most juvenile offenders (61 percent) avoid arrest upon reaching adulthood, the chances of being an adult offender are almost four times greater if one had a juvenile record than if one did not. While 6 percent of the sample were “chronic” offenders by age 18, 15 percent were chronic by age 30. The probability of future arrest varied directly with the probability of past arrest: The probability of a fifth arrest (for any crime not necessarily a violent one) given four “priors” was .80; the probability of an eleventh arrest given ten previous arrests was .90. The probability of a fifth serious (or “index”) offense with four prior arrests was .36; the probability of an eleventh serious offense given ten previous arrests was .42.

During the juvenile years, the subjects reported committing 8 to 11 serious or index offenses for every time they were arrested. Adults admitted to between three and six offenses for each recorded act. The main conclusion one could draw from the research on childhood predictors of adult violence is that the distinction between “childhood” and “adulthood” is not a particularly meaningful one in terms of violence prediction. The same factors (e.g., a history of past violence) appear to influence the occurrence of future violence regardless of age. Age is relevant to the extent that the earlier one begins a career of violence, the longer and more extensive that history may be, and as one enters the 30s, maturation processes become salient. The search for factors that “imprint” a violent disposition at an early age so far has produced results that are theoretically interesting but without much practical significance for prediction in the individual case.
- Monohan, John, Predicting Violent Behavior, Sage Publications, Inc.: Beverly Hills, 1981.
The article above contains foundational information. Articles below contain optional updates.

=================================

Personal Reflection Exercise #4
The preceding section contained information about childhood prediction of adult violence. Write three case study examples regarding how you might use the content of this section in your practice.

Online Continuing Education QUESTION 11
What are four childhood predictors of adult violence? Record the letter of the correct answer the CEU Answer Booklet.