- About Us
by James Q. Wilson of The Atlantic Monthly
Psychologists are helping parents to control bratty behavior by teaching them how and when to use rewards and punishments
“When I met him, he was six and a half years of age. There was nothing about his appearance that identified him as the boy who had set the record.” The words are those of Gerald R. Patterson, a family therapist at the Oregon Social Learning Center, in Eugene. The “record” to which he referred was the frequency–measured with painstaking care by the Learning Center’s staff–with which Don, a small boy, displayed rotten behavior. Nearly four times a minute while in his home, Don would whine, yell, disobey, hit, or shove. When he was not at home, telephone calls from teachers and merchants would mark his progress through the neighborhood: “He left school two hours early, stole candy from a store, and appropriated a toy from a neighborhood child.”
Don had a “sleazy look about him,” Patterson wrote, “like a postcard carried too long in a hip pocket.” His violent outbursts were frightening; any simple request or minor provocation would trigger obscene shouts, attacks on other children, or assaults on the furniture. His mother was tired, depressed, and nearly desperate as a result of coping unaided with this monster–no babysitter would take on the job of minding Don, whatever the pay. She nevertheless persevered, changing his wet sheets, bathing and dressing him, even feeding him, all the while talking to him in tones that vacillated between cajolery and scolding, murmurs and shouts. That produced only temporary compliance. When the father was home, things were not much different. The shouting and fighting between Don and his younger brother continued, occasionally punctuated by the father slapping both children.
Children like Don are the youthful precursors not only of difficult teenagers but sometimes of delinquents and adult criminals. The progression from violent, dishonest youngster to violent, dishonest teenager is not automatic, but it is common. Though child psychologists disagree about the extent to which personality in general is shaped by early childhood experiences, there is not much dispute over the fact that male aggressiveness tends to be stable from the early years on. Donald West and David Farrington, of Cambridge University, followed 411 London boys from childhood to adulthood, and concluded that those who were hostile or aggressive at the ages of eight or ten were likely to be hostile or aggressive when sixteen or eighteen. A Norwegian psychologist, Dan Olweus, reviewed this study and many others, and concluded that there is a great deal of continuity in male aggressiveness and misconduct. Aggressiveness tends to appear very early, often by the age of three, and to persist for many years in ways that are clearly reflected in the reports of parents, teachers, and police officers, and even in the reports of the young men themselves.
What is remarkable about this persistence is that it occurs despite the fact that aggression is often punished. Don’s parents hit him; we can safely assume that his teachers scold him; and it is likely that in time the police will arrest him. Some explain this persistence with the argument that punishment actually causes aggression: young men and women who are hit or abused learn to hit or abuse in return; they model themselves on the behavior of others, leading to a “cycle of violence.” Others believe that the child brings to the family some constitutional dispositions, acquired by heredity or from the accidents of prenatal or birth traumas, that not only cause him to behave more aggressively than others but cause his parents to treat him differently from his brothers and sisters. Still others think that it is not punishment that causes aggressiveness but punishment poorly conceived and badly executed. Whatever the case, contemporary research, by Patterson and by scores of others, has placed the family at the center of any effort to explain and reduce unruly or violent behavior.
The role of the family in producing such behavior seems to be a greater mystery to social scientists than to parents. If one asked the average citizen why some people are more likely than others to misbehave, he or she would very likely blame family experiences. If one then asked why children from the same family differ in their behavior, the citizen would probably say that because children have different temperaments, they react differently to the same treatment, or are treated differently by their parents, or both.
When Sheldon and Eleanor Glueck, at the Harvard Law School, published in 1950 the results of their ten-year study of delinquent boys in the Boston area, they concluded that if one held age, race, neighborhood, and (roughly) intelligence constant, delinquency appeared to be the result of an interaction between certain constitutional traits (body type and temperament) and a family environment in which one or both parents were indifferent or hostile and followed lax or erratic disciplinary practices.
There were criticisms of the Gluecks’ research. Some suggested that the observations of the boys may have been colored by the observer’s knowledge of whether or not a boy was delinquent. Others noted that some of the factors thought to be most important in producing delinquency, such as the degree of affection and the disciplinary practices of the parents, were not observed directly but inferred from statements by the boys and their parents. Still others objected to the Gluecks’ efforts to predict who would become delinquent given a certain family background; such prediction, it was argued, might unfairly label a decent boy a troublemaker.
But if the only problems with the Gluecks’ findings were these and similar methodological difficulties, one would have expected better studies to be mounted that would test their conclusions more rigorously. Few such studies were done. The reason, I suspect, is that the beliefs of social scientists about what constituted the appropriate objects of research changed. By the mid-1950s, the family was no longer the major unit of analysis in criminological research, having been supplanted by the gang and the social system.
Owing in part to an apparent rise in gang violence in the 1950s and in part to the advent of a newer group of criminologists, various theoretical perspectives that assigned a less prominent role to the family (and no role at all to constitutional traits) became dominant. One view held that children naturally conform to the values to which they are exposed; delinquents are simply exposed to values different from those the rest of us learn. The social class of the family or of other boys on the street will be an especially important determinant of what values boys accept. Lower-class boys may come to disparage the conventional ethic of success and embrace instead the values of toughness and immediate gratification.
Another view, first formalized in 1949 by Robert K. Merton, of Columbia University, held that man is governed by the relationship between his aspirations and the means available to achieve them. Families may encourage boys to want a good car and social respect, but if society denies the boys legitimate means for obtaining them, they might react by stealing to get the car and committing violent acts to earn the respect. Richard A. Cloward and Lloyd E. Ohlin, both than at Columbia University, applied this view to delinquency in 1960 by asserting that economically disadvantaged adolescents turn to crime as a way of achieving material goods and avoiding the shame that accompanies economic failure.
Other writers claimed in the 1950s and 1960s that delinquency was less the result of having learned from peers to commit crimes or of being driven to crime by frustrated expectations than the result of being arbitrarily labeled a criminal by society and its institutions. Young people who violate some arbitrary social convention are called criminals; once treated as a criminal, they become criminal.
Travis Hirschi, of the University of Arizona, observed that these perspectives sought to explain crime not by reference to manifestly wicked influences but by reference to supposedly good ones. Whereas an older generation of criminologists, of which the Gluecks were very much a part, thought that evil causes evil (for example, bad families produce bad children), many members of the newer generation were attracted to the idea that good–for example, the desire for achievement, the class structure, the economic order, or the decisions of the police and judges–produces evil. This view was most clearly implied by Merton: “A cardinal American virtue, ambition,’ promotes a cardinal American vice, deviant behavior.'”
The critical stance of many criminologists was reinforced by reformist impulses that led them, implicitly or explicitly, to explain social problems by reference to those factors that are, or appear to be, susceptible to planned change. The 1950s was a period of worry about gang violence, the 1960s one of attention to civil rights and economic opportunity. It was natural to assume that if one wished to deal with these problems, one had first to make them the focus of one’s analysis. Gangs, race, social class and criminal justice were concepts that seemed close to the crime problem and amenable to change. By contrast, the family was given less attention because of two defects: it was not on the public agenda, and it was not clear how its practices might be altered.
The study of the family did not stop, of course, but it did begin to lose it connection with the study of crime. Research on child development, for example, was a vigorous enterprise, but one not much concerned with crime (though it was concerned with behavior such as hostility and aggression, which we now know to be strong indicators of later criminality). There were exceptions: the sociologist William McCord and Joan McCord continued to trace well into adulthood the unhappy effects on later behavior caused by the discordant families in which a group of Cambridge and Somerville (Massachusetts) boys had been raised in the 1930s and 1940s.
The family was placed once again near the center of criminological attention by the masterful study of Travis Hirschi, published in 1969 under the title Causes of Delinquency. It began, as do most path-breaking works, by showing how we had been asking the wrong question. Most criminologists had been trying to explain how a naturally good person, or at least a naturally compliant one, might be led by social forces–gangs, schools, the police, labor markets–to commit crimes (or “deviant acts,” a phrase used by those criminologists who were anxious to avoid giving the impression that “crime” was anything more than the label middle-class society attached to behavior it didn’t like). Hirschi argued that the first question is not why men break the law but why they obey it: conformity, not deviance, is what is most in need of explanation.
His answer was that people obey rules to the extent that they form a bond to society. That bond is composed of four elements: attachment, chiefly to the family; commitment, by which Hirschi meant a prudent regard for the costs of wrong actions and the benefits of right ones; involvement in conventional activities, such as school; and belief in the moral validity of society’s rules. Hirschi tested his theory against the predictions that could be derived from rival theories by gathering information about the delinquent acts and personal backgrounds of several thousand junior and senior high school children in and around Richmond, California.
The number of delinquent acts, as reported by the children themselves, was powerfully influenced by the children’s attachment to their parents. The closer the mother’s supervision of the child, the more intimate the child’s communication with his father, and the greater the affection between child and parents, the less the delinquency. Even when the father held a low-status job, the stronger the child’s attachment to him, the less his delinquency. Other factors also contributed to delinquency, such as whether the child did well in and like school, but these factors were themselves affected by family conditions. Hirschi’s evidence supported his “social control” theory of the sources of delinquency better than it supported any alternative theories, such as those that say delinquency results largely from the influence of peers or the social class of the child: after one takes into account the bond with the parents and doing well in school, the contribution of parental class nearly vanishes, and that of delinquent friends sharply drops.
A very large body of data has demonstrated beyond much doubt the powerful effect on aggressiveness and delinquency of being raised in a family that is discordant, lacking in affection, or given to inappropriate disciplinary practices. The best studies are those that have followed a group of boys while they were growing up. The longest-running inquiry was the Cambridge-Somerville one, which began in 1937; data from it are still being published. From these industrial communities, 650 boys were selected to participate in a program designed to prevent delinquency. Half were chosen because teachers or social workers thought they were likely to become delinquent, the other half because they were thought to be normal boys. They entered the program at age eleven, on the average, well before most had committed any delinquent acts. The object was to test the efficacy of a counseling program in preventing delinquency, and to that end the boys were randomly assigned to treatment and control groups. There is little evidence that the counseling worked, but that is not what interests us here. Rather, the study is valuable because of the relationship it found between family background–as reported by counselors, teachers, social workers, and doctors–and behavior.
In 1955, eighteen years after the program began, William McCord and Joan McCord analyzed the voluminous records of these boys “blind”–that is, without knowing in advance which boys had been convicted of crimes and which had not. Thus, whatever biases may have existed in the Gluecks’ research, owing to their prior knowledge of who was delinquent, could not have affected the McCords’ investigation. The McCords found that the delinquent boys were about twice as likely as the nondelinquent ones to come from homes where parental disciplinary practices had been rated as erratic or lax. Delinquents were also much more likely to come from homes with a quarrelsome rather than an affectionate or cohesive atmosphere. The combined effect of these two factors–warmth (or its absence) and consistent discipline (or its absence)–was powerful: all the boys from quarrelsome families with erratic discipline, but only one fourth of those from cohesive families with consistent discipline, were convicted of a crime.
Lest someone suppose that the courts were more likely to convict boys from unhappy than from happy homes, the McCords analyzed separately the backgrounds of those boys who were described by counselors or others as aggressive but who had not officially been called delinquent. They found that these aggressive boys had essentially the same family backgrounds as the convicted delinquents.
It is possible that these differences in family background reflect how the boys were originally selected for inclusion in the study; children chosen because they appeared especially good or especially bad might differ more in family circumstances than would a random sample of boys. Indeed, it is possible that some of the boys were chosen because of what the teachers or others already knew about their families. There is not much evidence in the Cambridge-Somerville project that this was the case, but it cannot be ruled out entirely. Fortunately, another long-term study, this one in England, was carried out in ways that eliminate this potential source of bias. West and Farrington followed the careers of 411 boys chosen at random from a working-class section of London, starting at age eight and continuing for seventeen years. The boys were given batteries of tests and interviews, and their parents were interviewed about once a year until the boys were teenagers. West and Farrington gathered information about self-reported, as well as officially recorded, delinquency. By the time they reached the age of twenty-five, a third of the boys had criminal records. About thirty of the boys became persistent repeat offenders; in fact, those thirty accounted for more than half the convictions of the entire group.
The delinquent boys were more likely than the nondelinquent ones to have low IQS and to have parents who were cruel, neglectful, or passive. Boys described as aggressive by teachers or by themselves had the same background as the officially reported delinquents. Parental behavior could not explain all of the observed delinquency, however. If the father had a criminal record, this materially increased the chance that the son would be delinquent, independent of family income and prenatal behavior. In all, five factors seemed to play the largest role in predicting which boys would become delinquent: low intelligence (as measured by both verbal and nonverbal tests), large family size, parental criminality, low family income, and poor child-rearing practices.
What is indisputable in the London study is the impact of adverse child-rearing practices. West concluded: “A particularly noticeable characteristic of the parents of many of the delinquents in the study was carelessness or laxness in matters of supervision. They were less concerned than other parents to watch over or to know about their children’s doings, whereabouts and companions, and they failed to enforce or to formulate fixed rules about such things as punctuality, manners, bedtime, television viewing or tidying up.”
All this may strike the average parent as so obvious that it seems hard to believe that so much time and money had to be spent to establish it. Before making exasperated comments about professors reinventing the wheel, however, a parent reading this should be aware that there are at least three other possible explanations for the observed connection between family circumstances and youthful criminality. One is that delinquency is caused by the labels society attaches to children. A youngster from a discordant home is more likely, in this view, to be arrested and charged with a crime that if committed by a child from a happy (or an influential) home would result in the child’s release in the custody of his parents. No doubt different kinds of children are treated differently by governmental institutions. There is evidence, for example, that a delinquent from a broken home is more likely to be incarcerated than one from an intact home. But the labeling process cannot explain away the influence of families on misconduct generally, because, as we saw in the studies mentioned so far, the connection between discordant or unaffectionate families and delinquency exists even when our only knowledge of the delinquent act comes from the child himself (Hirschi; West and Farrington) or from counselors instead of police officers (the McCords).
A second explanation for the link between family life and criminality rests on the possibility that the child may have inherited from his parents some cognitive deficiency or temperamental predisposition that causes his aggressiveness. In this view, the discordant family is not so much the cause of the misconduct as either a reaction to it (a mother slaps or yells at children who since birth have been fussy or irritable) or irrelevant to it (the delinquent is born, not made). There are good scientific grounds for believing that the child is not a blank slate on which families, friends, and society can write at will. There is some constitutional, possibly genetic, predisposition to certain forms of misconduct, as shown by elaborate studies in Scandinavia and elsewhere that reveal a connection between the criminality (or the alcoholism) of the biological father and the subsequent criminality (or alcoholism) of the son, even though the son was adopted at a very early age and raised by noncriminal foster parents. Important as these constitutional factors may be, their exact operation is not well understood, nor do they account for all of the individual differences in criminality that we observe. Familial factors play an important role.
A third explanation is that discordant or unaffectionate parents cause aggression by their use of punishment, especially physical punishment. One famous study, by Robert Sears, Eleanor Maccoby, and Harry Levin, concluded that “the more severe the punishment, the more aggression the child showed.” Barclay Martin, after reviewing the evidence about how young (aged two to five) children go abut seeking attention from their mothers, noted that the studies suggested that punishing attention-seeking behavior seemed to increase rather than decrease the frequency of the children’s demands. The clear implication of such findings is that the use of punishment is at least as likely to teach children, by their observations, to be aggressive as it is to teach them, by being conditioned, to be pacific.
But, as Martin and others have pointed out, there are some reasons to be skeptical of such conclusions. First, most are based on the mother’s report of the child’s behavior and of her discipline; both may be seriously in error. A mother trying to raise a difficult child may overstate how often punishment is used while a mother raising an easy one may understate how often it is used; these reporting errors would create a false positive correlation between punishment and aggression. Second, most such studies neglect the way the child is predisposed to behave. An active, restless, fussy child may be punished more frequently than a placid, easygoing one. The greater the punishment of the former may in fact reduce the rate of such misconduct, but because that rate is already high, it can appear that more frequent punishment “causes” more frequent aggression.
Most important, the effect of punishment (or affection) on behavior will depend not on how much is used but on how it is used. For a reward or a penalty to have significant effect on behavior, it must be used in a consistent and contingent manner. Suppose a child who nags his mother for attention often gets her attention, but on occasion is slapped instead. The child will come to believe that he can usually get his mother’s attention by nagging, even though he is sometimes slapped. Sometimes he wins, sometimes he loses, but if he values her attention enough, or if he gets his way more often than he gets slapped, he will redouble his nagging. If nagging does work often, the amount of it can be very great and thus the number of times he gets slapped can be very great as well. This can cause the unwary observer to conclude that slapping has “caused” the nagging.
Or the parent may slap the child without regard to whether he has broken a rule, because the parent is grouchy, or irritated by behavior that ordinarily is permitted, or acting under the erroneous belief that a rule has been broken. Such punishment, not being clearly contingent on behavior defined by some rule, is not likely to produce adherence to any rule; on the contrary, to the extent that it is random or erratic (in the child’s view), it may lead him, just as social-learning theorists argue, to believe that violence is a normal and acceptable method of expressing one’s feelings.
Speculating about all these possibilities is no substitute for directly observing children’s behavior and trying to change it. Family therapists have been doing this, in one way or another, for decades. In the 1950s, and to some extent today, a therapist presented with an aggressive, sneaky, or hostile child would assume that the youngster was aggressive because he was anxious or frustrated or because he had a weak ego, often as a result of the parents’ neuroses. The therapist would use individual psychoanalysis or counseling to “cure” the parents and, sometimes, nondirective play therapy to help the child, or would perhaps stage a psychodrama that involved the parents and the child simultaneously. In general, little of this seemed to make a difference, though there were individual success stories.
Beginning in the 1960s, a new approach was tried. Owing to the rising influence of behavioral psychologists, foremost among them B. F. Skinner, family therapists began looking at a child’s behavior as learned on the basis of the rewards it received. A young boy would engage in rotten behavior if he found it useful. If he got what he wanted from his parents and teachers by yelling, shoving, and hitting, no one should be surprised to discover that he would continue to yell, shove, and hit. A variant of this approach, pioneered by Albert Bandura, held that a child would also learn to yell, shove, or hit when he saw other people doing this and getting away with it. The clear implication of these theories was that the therapist should try to reward the child for doing the right thing instead of the wrong thing. There was already evidence that such alleged “sicknesses” as bed-wetting, stuttering, and fear of snakes could sometimes be cured by rewarding the opposite behavior; why not reduce aggression by rewarding obedience?
Soon, the principles of behavioral psychology were being applied in schools, foster homes, juvenile institutions, and even adult prisons. This application came to be called “behavior modification,” a term that fell into some public disrepute because a few such programs used drugs or electric shock to punish–and, it was hoped, extinguish–undesirable behavior. But the great majority of these programs sought chiefly to reward desirable behavior by, for example, giving a person a valuable token every time he did something right. (Such programs were called “token economies.”) Other methods were employed as well, all having in common the effort to affect behavior by changing its consequences.
There have been countless demonstrations of the fact that applied behavioral psychology can change the behavior of individuals in schools, clinics, and prisons. Students have been induced to study harder, children to throw fewer tantrums, and prisoners to obey institutional rules. But in the great majority of cases, these changes disappear soon after the individual has left the institution. While behavior modification was an effective way to make life easier for people running the institutions, there was not much evidence that it was doing anything to protect the community.
Perhaps the best-known of these programs tried to improve the behavior of delinquent or troublesome youths by placing them in foster homes run on behavior-modification principles. The idea was to use psychological theory to construct a better family like for children who had an unsatisfactory natural home. Achievement Place, in Lawrence, Kansas, is a foster home led by two professionally trained foster parents, in which six to eight delinquent adolescents live. Through instruction reinforced by a token economy, the leaders of Achievement Place hope to teach the delinquents to take care of their living quarters, study hard in school, improve their social skills, become punctual, and otherwise display self-control and conformity to rules. Since its inception, the Achievement Place strategy, now called the “teaching-family model,” has been followed in many parts of the country; by last year, there were about 170 such group homes in operation, staffed by foster parents trained at six regional centers. Moreover, the system has been used not only for delinquents but for autistic or emotionally disturbed youths, the mentally retarded, schizophrenic adults, and even the aged. The bibliography of books, articles, and dissertations written about the teaching-family model runs to nearly fifty single-spaced pages.
There seems to be little doubt that the teaching-family system improves the behavior of even serious delinquents while they are in the group homes. But according to a comprehensive review of the system, completed in 1981 for the National Institute of Mental Health by Richard R. Jones, Mark, R. Weinrott, and James R. Howard, there is little evidence that it alters the rate at which the youths commit offenses one year after their release from treatment, or that it has any greater effect on delinquency than do other community-based programs.
The reason is obvious. When a delinquent leaves the foster home and returns to his discordant family and the criminal street life, most of the rewards he receives will once again reinforce, or fail to discourage, impulsive, aggressive, or illegal activity. The key to breaking this unhappy cycle is to improve life in the delinquent’s natural home. Since the late 1960s, there have been a number of efforts to do this.
Various programs have been established to teach the parents of troublesome youths to be better parents. Among the first was one carried out in Tucson, Arizona, by Roland Tharp and Ralph Wetzel and reported on in 1969. Believing that the parents of delinquents often failed to reinforce good behavior and instead denigrated or nagged their offspring, Tharp and Wetzel tried to persuade parents and children to manage their conflicts more constructively. To make it easier to manage day-to-day affairs with a minimum of discussion and hence of arguments, Tharp and Wetzel designed a “behavioral contract” that would set forth the rights and obligations of parents and children in a particular family. For example, instead of arguing over whether he could use the family car, a boy would enter into a contract with his parents whereby he would get the car at stated intervals in exchange for performing certain chores and being home at a specified time. These and other efforts at family contracting, or, as it is sometimes called, contingency management, were based on the assumption that rewards should be stably linked to desired behavior on the basis of an agreement that seems fair to both sides. The youths whose families followed a behavioral contract showed a reduction in offenses, an improvement in grades, and an improvement in behavior, but the significance of these changes is hard to assess inasmuch as there was no control group with which the youths in the experimental program could be compared.
A variation of this approach to altering family life was tried in Salt Lake City by James F. Alexander and Bruce V. Parsons, then of the University of Utah. Several dozen boys and girls who had been involved with the juvenile court for various minor delinquencies–truancy, shoplifting, drinking, and running away from home–were referred to a clinic at the university. There, the youths and their families were randomly assigned to one of four groups: a behavior-modification program, a conventional group-discussion counseling program, a church-sponsored family-counseling program, and a no-treatment group. The behavior-modification program involved inducing family members to talk more constructively with one another, to tolerate interruptions, and to agree on certain standards of conduct and the consequences of conforming to those standards. Over the ensuing year and a half, the youths who had been in the behavior-modification program were only half as likely to reappear in juvenile court as were those who got no treatment or who were in the conventional counseling programs; moreover, the brothers and sisters of the youths in the treatment program were also less likely to get into trouble with the law than were the siblings of the youths in the control groups. The number of families involved was so small and the kinds of delinquencies averted so minor that the Utah project cannot be regarded as settling the question of whether family life can be altered in ways that will reduce serious delinquency, but it provided a promising lead.
That lead has been most thoroughly explored by Gerald R. Patterson and his colleagues at the Oregon Social Learning Center. Unlike many other therapists, they have concentrated on reducing deviant behavior among pre-adolescents rather than on rehabilitating teenage delinquents; in fact, much of their research suggests that beyond a certain age, changes in the family environment have little, if any, effect on young persons, because peer pressure becomes more important than parental rewards. And unlike some programs, the Learning Center’s is not based on the assumption that families need to communicate more (or less); what is crucial is the way in which parents define, monitor, and control the behavior of their children in routine settings.
The key to Patterson’s approach–and that of several like-minded therapists, such as Robert Wahler, in Tennessee–is the belief that many parents do not know how to raise children. What makes bad families bad is not neurosis or indifference as much as incompetence. When parents have an unruly, violent, sneaky kid on their hands, it is not because they like him that way, or because they don’t care how he behaves, or because they have failed to solve his Oedipal conflict; it is because they fail to tell him clearly how he is expected to behave, fail to monitor his behavior closely to ensure that he behaves that way, and fail to enforce the rules with appropriate rewards and penalties, promptly and unambiguously delivered. The child may actually suffer from a weak sense of self, but such a subjective state is not readily changed directly. There is little good evidence that therapists can change the behavior of Don, or of children like him, by first changing their attitudes; indeed, it would be a miracle if they could get so hostile and violent a child as Don even to listen to a discussion of his attitudes. Instead, one changes behavior first and hopes that changes in underlying attitudes will follow.
Since the mid-1970s, the Learning Center has treated more than 250 families whose children, ranging in age from three to fifteen, had been referred to the clinic because most of them had set fires, stolen property, picked fights, or upset their teachers or parents; some were victims of child abuse. In the child’s home, the therapists carefully measured all that occurred between a child and his parents or siblings every six seconds for nearly an hour. About half of the child’s possible ways of behaving were termed “aversive,” meaning unpleasant; these included whining, yelling, teasing, and hitting.
Their central conclusion was that the families of problem children differed from those of normal children not so much in whether they punished too much or too little as in whether they knew how to punish. The parents of antisocial children both used more punishment (scolding, shouting, threatening) than did the parents of normal children and failed to make their use of penalties contingent on the child’s behavior. More precisely, these parents were less likely than others to state clear rules, monitor compliance, and punish violations. Instead, they “nattered” at the child, occasionally and unpredictably interrupting the nattering with a slap or a loss of privileges. Patterson and his colleagues suggested that nattering instead of effective discipline occurred in part because the parents were less attached than other parents to their children, in part because they did not know how to control behavior effectively, and in part because they felt overwhelmed by a succession of minor problems that cumulatively amounted to a crisis. The irritable parent who does not use discipline effectively tends to produce aggressive children; the indifferent and ineffective parent tends to produce larcenous ones.
The notion that there is a defect in parental skill, as opposed to personality, mental health, or economic resources, that accounts for failures to socialize young children may not be a surprise to many normal parents but it is a revolutionary conclusion in the field of family psychology. As Patterson observes in his recent book, Coercive Family Process, some psychologists, confronted by a young person who becomes violent or criminal, have sought explanations “at least as dramatic as the phenomena they purport to explain,” and so we are treated to accounts of a primal instinct for aggression, a lurking Oedipal complex, or a shattering divorce. Rather than “cataclysmic episodes, flood tides of rage, or crumbling defense structures,” what in fact is happening, Patterson writes, is the mismanagement of “coercive family processes,” processes made up of events that are “inherently banal.”
Among these banal events are the routine interactions of parents and children as they convey, by word, tone, gesture, and expression, their approval or disapproval of the behavior of others. Though the literature on family socialization is heavy with discussions of the merits of “love” versus “punishment,” most of the socialization process is carried out by the often reflexive display of attention or irritation, interest or disinterest, and approval or disapproval.
The failure of parents to use reasonable reinforcements contingent on steadily monitored behavior places the child in a situation in which he comes to understand that he cannot control by his own actions what happens to him. When one receives penalties unconnected to one’s own behavior, one experiences a kind of stress that Martin E. Seligman has called “learned helplessness,” just as when one receives rewards that are unearned, one develops “learned laziness.”
The treatment at the Learning Center was to teach these troubled parents how to set clear rules, monitor behavior, and make rewards contingent on good behavior and punishment contingent on bad behavior. By rewards, the therapists meant not necessarily giving presents in exchange for some major instance of good conduct but routinely responding in pleasant and supportive ways, or with “points” exchangeable for small privileges (such as ice cream for dessert), to pleasant language and helpful behavior. And by punishment they meant not only assessing major penalties for major misdemeanors but promptly and consistently penalizing unpleasant and destructive language and conduct. An especially favored penalty, the value of which was established by repeated trials, was “time out”–that is, being sent briefly, usually for five minutes, into seclusion in another room, usually the bathroom, without recriminations or long lectures.
Teaching these common-sense methods was difficult, but not nearly as difficult as motivating the parents to put them to use. Both instruction and motivation required extraordinary clinical skill and patience, neither of which was easy to sustain. Patterson does not assume that parents will learn how to raise children by reading a book or listening to a lecture. If that were all there was to it, they would not be in trouble in the first place. Benjamin Spock’s reassuring advice to young parents-“You know more than you think you do”–is probably correct for most mothers and fathers, but it is not correct for those who, owing to their temperament and personal problems, are parental disasters. These people sense that something is badly wrong with the way they are raising their children, but they don’t know what to do about it, or, worse, they often think they know, so they persist in trying to control children with vague rules that are frequently changed and erratically enforced. Moreover, many parents do not always share the therapists’ enthusiasm for the principles of behavior modification. To these mothers and fathers, “reinforcing socially desirable behavior” sounds too much like bribing kids to do what they are supposed to do anyway, and “relying on positive rather than negative reinforcers” seems to be spoiling the child by sparing the rod.
The only way to convince such parents is to show them that the rules they learn during many hours spent in the Learning Center actually work. I watched videotapes of a woman being trained to cope with her out-of-control young son. The child went to bed only when he felt like it, insisted on sleeping with his mother (she had no husband), rarely obeyed even the most reasonable commands, spread his excrement all over the living-room walls, was a terror to other children who tried to play with him, and seemed destined to be a terror to his teachers. The first task was to make the mother realize that he was not minding her in important ways because he was not minding her in small ones. Every day for one hour she had to count the number of times the boy failed to obey an order within fifteen seconds of its being issued and report the results to the therapist. This led the mother to become aware of how many times she was issuing orders and how long she was waiting to get results.
At the second session, she was taught how to use rewards to increase the frequency with which her son minded her. Every time he minded within fifteen seconds, he was to get one point; at the end of the day, the points were added up and converted into some small but valued privilege, such as a special dessert. Many parents, impatient, want to use big rewards to achieve big, quick improvements in behavior, but this is discouraged: such a strategy debases the currency, fails to inculcate the habit of minding, and does not reduce routine parental nagging.
At the third session, the mother was taught how to use “time out” as a means of discipline. She was told that whenever her son did something wrong, she should immediately tell him why it was wrong and order him to go to time out–five minutes alone in the bathroom. She resisted doing this, because it forced her to confront all of her son’s rule-breaking, and to do so immediately. She preferred to avoid the conflict, and the angry protests. She especially resisted using this means to enforce her son’s going to bed at a stated, appropriate time; she was a lonely, not particularly attractive woman, and it was clear to the therapist that she wanted her son to sleep with her. In time, the woman was persuaded to try this new form of discipline and to back up a failure to go to time out by the withdrawal of some privilege (“no TV tonight”). As the weeks went by, the woman became excited about the improvements in the boy’s behavior and came to value having him sleep alone in his own room. Not long after, she acquired a male friend.
Some parents come to the Learning Center having relied on severe physical punishment to control their children. The staff members do not approve of spanking, but neither do they try to talk parents out of it, unless it is clearly abusive. Instead, they show the parents that another method works better: a lesser penalty, such as time out or a loss of points, given immediately for every minor infraction of the rules. Parents who physically punish their children usually do so only for major offenses that occur at the end of a long sequence of minor ones. These big penalties erratically employed only against major misconduct are less effective than small ones regularly used to control minor misconduct.
The Learning Center double-checks on the progress of the parents by observing the family in action, making frequent phone calls to the home to see how things are going at that moment, and gathering information from teachers. Increasingly, the researchers at the center have been testing their own work using the only scientifically sound method–comparing its results to those achieved by other methods applied to similar families. Nineteen families of very aggressive children were randomly assigned to either the center’s parent-training program or some conventional form of therapy available in the community. In 1982, Patterson and two associates, Patricia Chamberlain and John B. Reid, were able to show that the children in the parent-training program displayed a sharp drop in the frequency of aversive behavior, while those in the conventional programs showed, on the average, no change.
The limit, it turns out, is the amount of time, effort, and therapist skill at the disposal of the clinic. Before Patterson’s group demonstrated what it could accomplish, efforts to prove the lasting value of parent training had produced mixed, and often negative, results. Some programs reported great success, but it turned out that their measure was the parents own reports that things were getting better. Parents always like to think things are getting better. Observations by outsiders, such as schoolteachers and therapists, would often show that things weren’t getting better at all. Other programs produced no benefits, no matter who was reporting, but it would turn out that these programs employed young, relatively inexperienced therapists, usually graduate students, and lasted only a short time. Patterson has been able to show that training parents takes a great deal of time and skill–a family with only minor problems may be in treatment for six weeks, one with moderate problems for sixteen weeks, and an especially troubled one for much longer. The most experienced therapists produce the best results, but only if they are continuously supported by a caring and knowledgeable group of colleagues.
The Patterson group is coming to understand how complex is the relationship between the economic circumstances of a family and its child-rearing problems. Most of their clients have low to moderate incomes. Half or more are single mothers. Economic disadvantage, especially when faced by a lone mother, can be a great source of stress, which exacerbates the normal difficulties of raising a child. But the connections among economic circumstance, parental personality, and the child’s behavior are not simple or obvious. The mothers in the Learning Center program tend to be maladjusted personalities, as measured by a variety of tests. Their economic and marital difficulties may have contributed to this, or, just as likely, their personalities have made it hard for them to succeed in marriage or in the labor market. If their material circumstances improved, the problems of their daily lives would be somewhat reduced, but owing to their personalities, their capacity to manage the child-rearing stresses common to all families would probably remain low.
It is tempting to speculate about whether the growth in crime experienced by this and other nations in the past few decades can be explained by changes that have occurred in family life. It is easy to believe that the increase in the number of single-parent families, working mothers, or unattended, “latchkey” children lies at the root of the increase in disorder and delinquency. It may, though I am struck by how weak or contradictory the evidence is. For example, a review by Elizabeth Herzog and Cecelia E. Sudia of eighteen studies of the relationship between broken homes and crime carried out between 1950 and 1970 found seven that claimed father-absent homes produced more delinquency, four that claimed they produced less, and seven that came to equivocal conclusions. The Gluecks, among others, claimed that broken homes made a difference; the McCords and Travis Hirschi, among others, claimed they did not. Moreover, even if the rise in American crime rates could be explained by the increase in female-headed families here, the growth in crime in European countries–a growth that parallels ours almost exactly, though beginning from a lower base–cannot be explained in this way, because these countries have not had a comparable explosion in the number of female-headed homes.
Other things being equal, single-parent families, or two-parent families with working mothers, probably experience greater trouble in raising children than two-parent families in which the mother stays home–but other things are rarely equal. Why one parent is absent is surely more important than whether he is absent. As Michael Rutter, of the University of London, notes, based on his own research in England and that of others elsewhere, families that are broken by divorce or desertion are more likely to affect the child’s behavior adversely than those broken by the death of one parent. Homes that are broken because of parental discord are unhappy places in which the child experiences not only the stresses associated with separation but those produced by the discord that led to the separation.
Many intact families are also discordant. When New York City families, some living on welfare and some not, were studied by a group of researchers based at Columbia University, it was found that delinquency was best explained among both kinds of families by “parental coldness.” But this coldness was most strongly associated with delinquency among the welfare families because, as the authors put it, the “welfare child is in double jeopardy”–when there is no father present, the harmful effects of maternal coldness are more serious. Here, as in all aspects of family life, what causes what is never quite clear. A discordant or broken home may harm the child, but a troublesome child can cause some homes to become discordant. Candyce Russell, a sociologist, found that the parents of babies who were more demanding–who cried a lot or had feeding problems–were more likely to experience discord and crises than the parents of quiet infants.
Changes in the structure of the family may have contributed to the growth of crime, though of late crime rates seem to have been stable even as the number of children raised in broken homes has continued to increase. It is more likely that broad social forces–the advent of the baby boom, the increase in personal freedom and mobility, the spread of the ethos of self-expression, and the increases in the net benefits of crime (the value of loot minus the risk of being caught and punished)–have produced both increases in crime and changes in family life.
There is no reason to assume that parents have become less intelligent about how to raise children. For hundreds of years, all over the globe, parents who have never heard of Dr. Spock, much less of behavior modification, have been raising tens of millions of perfectly decent children. If crime and disorder have increased, can it be that parents who once knew how to raise children have forgotten? It seems unlikely. But it may be that social changes have made parents somewhat less interested in being parents, and have provided children with more ways to evade parental control. More mothers and fathers may have found children to be a burden as the traditional social and moral supports for family life have become more precarious and the opportunities for distraction and entertainment outside the family have become greater. Parents who once just got by as child-rearers now find themselves slipping over the edge as it becomes harder, or less necessary, just to get by.
Give the deep-seated and poorly understood causes of discord and incompetence among families, it is hard to imagine mounting any national program to deal with the problem directly. One shudders at the thought of developing, as some politicians too eager for the national spotlight have proposed, a “national family policy.” But at the same time, one also wonders whether, starting small and in a frankly experimental mode, we might not do more to prevent the frenzied or apathetic incompetence of so many families from producing monsters like Don.
The fact that much skill and time are needed to change families means that this method of reducing delinquency, like most methods, could not easily be made into a national program, or even one that would serve a large city. But it is conceivable that the elements of the program might become part of an effort to prevent bad parental practices from developing in the first place. The principal barrier to success in helping troubled families is that they resist help. It is not the complexity of the parenting techniques that requires so much time and effort to be devoted to therapy–the techniques are simplicity itself, based, as they are, on common sense. Rather, it is the client’s resistance to adopting these techniques, or even to confronting the fact that his or her own practices are disastrous, that requires so great an investment of resources.
But suppose that these techniques were learned by young people before they become parents and enmesh themselves in the vicious cycle of nagging, exasperation, appeasement, and random hitting. Suppose the techniques of scoring behavior from videotapes of actual families in trouble were taught in school, using the same equipment researchers now rely on–the engrossing paraphernalia of hand-held computer terminals and video displays with which the Pac-Man generation has become so familiar. Suppose the changes in behavior that occur as a result of adopting sensible parenting techniques could be seen (on a TV set) by young people. It is conceivable that they might find this experience more interesting and this information more useful than what they often get from lectures they are now obliged to hear on sex education or personal hygiene, to say nothing of courses in civics or woodworking. This suggestion–not yet tried, to the best of my knowledge, anywhere–may turn out to be but another example of misplaced hopes and wishful thinking. But the costs to society of weak, discordant, incompetent families are so great as to make the experiment worth the effort.
At the time this article was written, James Q. Wilson was the Shattuck Professor of Government at Harvard. This article was based on research he was doing in collaboration with Richard J. Herrnstein on criminal behavior.
Reprinted with permission. Copyright October 1983, The Atlantic Monthly.