The Alleged Child Victim and Real Victims of Sexual MisuseHollida Wakefield and Ralph Underwager*INTRODUCTION Neither the science of psychology nor the exercise of law is perfectly accurate. The lawyer and the psychologist each agree that improved accuracy is the goal for each. Everyone involved in the justice system — judges, attorneys, accusers, victims, defendants, law enforcement, and society as a whole — benefits from increasing the accuracy of the decision-making process. As Saks and Kidd (1980-81) observe:
The science of psychology has been asked to provide information to aid in the decision-making process when there are allegations of child sexual abuse. Psychologists have responded to this request, but often in the absence of accurate and reliable information. Ravetz (1971) notes that when a field of inquiry is immature and the relationship to the events it purports to describe is tenuous, it may mask this true state of imprecision by producing aphorisms. An aphorism is a deeply felt personal doctrine presented in a superficially comprehensible fashion, which often is in error. This is what took place in the development of the system to handle allegations of child sexual abuse in the last fifteen years. Aphorisms such as "Child sexual abuse is a raging epidemic," "Society actively chooses to ignore sexual abuse," "Children don't lie about sexual abuse," "Children can't talk about things they have not experienced," "Children are no more suggestible than adults," and "Children must be believed at all costs," have been advanced as truisms and given the patina of scientific authority. On this basis, politicians, law enforcement, and social welfare authorities have constructed, in the absence of facts, a system of laws, policies, regulations, bureaucratic structures, techniques, and procedures now entrenched in our justice system and established as the way our society responds to child abuse. Any decision-making structure built on error is liable to produce error at an indeterminate, unrecognized, but significant level that causes harm (Gambrill 1990). Evaluation of the extent of error and thus the potential for harm may best be done by the application of Bayes Theorem (Fischoff & Beyth-Marom 1983). The Bayes approach rests on the principle that the degrees of belief in an ideally rational person conform to the mathematical principles of probability theory (Horwich 1982). This concept should be acceptable to a scientifically trained person or to anyone who respects science. A number of scientists have applied Bayesian inference to child sexual abuse in the interest of assessing the level of error and type of error produced by the system. Every Bayesian analysis of the decisions made by the child abuse system that we have found concludes that the most probable and most frequent type of error is false positive, that is, identifying an individual as abused or an abuser when ft is not true (e.g., Altemeier et al. 1984; Caldwell et al. 1988; Gambrill 1990; Homer 1992; Homer & Guyer 1991a, 1991b; Kotelchuck 1982; Milner et al. 1984; Paradise 1989; Realmuto et al. 1990; Starr 1979; Wakefield & Underwager 1988; Zeitlin 1987). This is true even when a 95 percent accuracy level for the decision making is assumed, as in Gambrill (1990). Starr (1979) assumes a procedure that is 83 percent accurate in correctly identifying abusive situations. Still he reports a ratio of twenty false positives to one true positive. Homer (1992) finds the ratio of false positive to false negative classificatory errors ranging across studies from 3:1 to an astonishing 2000:1. More realistic and probable lower estimates of the accuracy of the decision-making process but higher base rates produce unconscionable and unacceptable levels of false positives, ranging to a ratio of nine false positives to one true positive. Much has been made of the harm to children if there is a false negative decision, that is, a failure to identify a child as abused when it is true. This conclusion ignores the damage done to innocent people who represent the false positives generated by the system. Also, the severity of damage may be greater when a nonabused child is treated as if abused than when an abused child is not correctly identified. The justice system must begin to consider this possible reality and to ask how the accuracy of decision making may be improved. There is now enough credible scientific knowledge that we can improve the accuracy of the decision-making process. We do not need to continue making the kinds of error that have been made up to the present. There are no good reasons for being opposed to efforts to increase accuracy, only bad ones.
THE AFTERMATH OF ERROR If adults make a mistake and treat nonabused children as if they have been abused, this is not innocuous or benign. Smith (1991), a psychotherapist, describes a letter brought to him on her fourth visit by Stephanie, a 17-year-old psychotherapy client:
The next night Stephanie died from an overdose of her mother's sleeping medication. William H. Yanco, a policeman, committed suicide after a complaint was made that he had kissed a 10-year-old boy on the lips. Yanco, who had specialized in working with disturbed children for fourteen years, was judged to have died in the line of duty. Yanco wrote in a letter to his two sons, "No matter what I say, people will always be suspicious. I can't stand the way people are going to think of me" (Associated Press 1992). Douglas Tarrant killed himself after being accused of sexual abuse which he claimed was false in a videotape he left behind for his family. The l5-year-old girl who had accused him had retracted her story two days before Tarrant killed himself, but he had not been told of the recantation. (Landry 1988). A Canadian mother who had, been falsely accused of sexual abuse by her former husband got her three children returned to her by judicial order (Stuart-Mills v. Cher 1990). They had been subjected to interrogation techniques known to the husband who was a member of the Mossad, the Israeli intelligence agency. The children hated their mother so violently they cut up her clothes, tried to burn down the house, attacked her, cursed, swore, skipped school, and refused to cooperate in any way. Now, two years after the event, the mother believes she and her children have repaired the breach but it will never be the same (Stuart-Mills 1992, personal communication). The assessment of an accusation is seldom a neutral event for anyone. Mantell (1988) notes that the process of evaluating an accusation may result in more damage to the interests of the child and to the child's primary relationships than the original act in question. Tyler and Brassard (1984) report that a sexual abuse investigation can be devastating to families and children, especially to the alleged victim, who may be placed in a shelter or foster care. Diorio (1992) reports that parents whose children had been removed invariably perceived the actions of the protection system as oppressive, unfair, insensitive, harsh, cruel, and tyrannical. Baurmann (1983) found that secondary victimization was common in his sample of 8,058 sexual abuse victims in lower Saxony; he reported that for at least one-fifth of his sample, the main cause of the damage was judged by the victims to be the behavior of relatives, friends, or the police rather than actual abuse. Even if an allegation is eventually judged to be false, the family, including the alleged child victim, will have been severely traumatized by the allegation. Schultz (1989) surveying one hundred families falsely charged with sexual abuse, found that almost all reported major disruption and trauma. Davis and Reppucci (1992) surveyed eighty-five men who had claimed to have been falsely charged with abusing their children or stepchildren. Almost all, regardless of whether they had been found guilty of abuse or not, reported a variety of negative effects in diverse areas of their lives. Robson (1991) comments on the aftermath for the Scott County, Minnesota, families, in which twenty-five adults were accused of abusing their own and other children. The charges were dropped after the only one to go to trial ended in an acquittal and the children eventually were returned home. However, the families all experienced severe dysfunction and distress, and "perhaps permanent emotional damage to the accused and the accusers alike" (p.50). Weinbach (1987) remarks: "Acquittal of charges cannot undo the damage. Even unsupported charges tend to leave lingering doubt among friends, family, and associates" (p. 532). A study of thirty families where erroneous charges of sexual abuse were made was recently carried out in England by Westminster College, Oxford. The study examined the process of investigation, outcomes, and the effects on children, parents, and extended families. All suffered post-traumatic shock. The report also suggests that the type of counseling required by these families is unique and the closest model is that of victims of violence. They are disabled because an external force has assaulted them and torn their lives apart (Prosser and Lewis 1992). Children are often placed in foster care as a result of a sexual abuse allegation. Foster care is seldom positive; Besharov (1985b) reviews its effects and concludes it is an emotionally jarring experience which confuses younger children and unsettles older ones. The trauma for a child who has been sexually abused is likely to be increased if the child is removed from home (Gomes-Schwartz et al. 1990). A court-appointed panel of thirteen experts reviewed the entire Illinois child welfare system and their handling of the 23,000 children in their custody. The report establishes that the agency charged with preventing child abuse and neglect is itself abusing and neglecting thousands of children (Final Consolidated Report of Rule 706 Panel of Experts 1990). Silberberg and Silberberg (1982) observe that parents who abuse children are accused of child abuse while officials who abuse children say they are rehabilitating the child. Jones (1991) describes nine possible components of iatrogenic (doctor-induced) harm by the system when children have actually been abused. They are (1) overzealous professional intervention, (2) repeated interviewing, (3) repeated physical examinations, (4) decline in living standards, (5) defensive decision making, (6) attendance at court, (7) withholding treatment, (8) over-treatment, and (9) foster care. Jones also reports a number of studies demonstrating that there is a higher frequency of abuse in foster care families than in matched-comparison natural families (families of birth). If an abused child is traumatized by the consequences of a sexual abuse disclosure, investigation, and justice system involvement, the effects can be even worse for a child who has, in fact, not been abused. Not only may the child be placed in foster care or be separated from a parent, but the child is often put in sexual abuse therapy where the false allegation is encouraged and reinforced. The child may be forced to be fixated on feelings of having been abused and to talk about the abuse and the abuser for months or even years. Campbell (1992b) reports on a 7-year-old girl who was in therapy for two years. As part of the therapy, the child was required to keep diaries of her thoughts and feelings for nine months. There were 246 entries in the diaries, mostly descriptions of dreams. The girl's mother was portrayed negatively in 121 entries and her stepfather in 120 entries. There was a steady progression of hatred and hostility toward both. When the trial was held, the girl testified to abuse by the mother and stepfather, recounting the content of her dreams as if they were reality. The environment of a divorce/custody conflict and the development of the allegations strongly suggested that this was a false allegation and the jury acquitted both the mother and stepfather. This kind of therapy experience can be an assault on the child's developing ability to differentiate reality from fantasy. This is the kind of harm that can be done to children and families. If it is the case, as Homer and Guyer (1991a, 1991b) suggest, that for every false negative there are over twenty false positives, the sheer amount of human misery and anguish caused by the system that aims at protecting children demands an immediate improvement in accuracy of decision-making.
HOW MANY ALLEGATIONS ARE FALSE? The reports of suspected child abuse have greatly increased in the past twenty years, particularly for reports of child sexual abuse. This increase results from the broadening of the reporting laws along with the public and professional attention given to child abuse. But as Davis and Reppucci (1992) observe, the reforms that have increased reporting and made prosecution easier have also made it easier wrongfully to accuse and to prosecute innocent people. Child sexual abuse allegations present great difficulties to the justice system. Often the child is the only witness and there is no corroborating evidence of the abuse. Several highly publicized day care cases in the last several years have highlighted the problem of sorting out the truth. The McMartin preschool case ended in acquittals, hung juries, and dismissals because the jury concluded that the suggestive interviews of the children by the prosecution clouded the actual facts of the case. But parallel trials, namely the Little Rascals case in Edenton, North Carolina, and the Kelly Michaels case in New Jersey ended in convictions. Child sexual abuse allegations arising during a divorce and custody dispute are especially difficult because of the young age of the children involved, the possible motivations of adults, and the need to balance the rights, interests, and welfare of the child and the accused parent. In all cases, particularly if the child has been interviewed repeatedly, there is a question of the extent to which the account has been contaminated by the adults influence (Doris 1991; Underwager & Wakefield 1990). While many professionals believe that most allegations are true, there has been a continuing dispute about the proportion that are false, with some professionals claiming they are extremely rare (e.g., Berliner 1988; Faller 1984 & l990; Jones & McGraw 1987; Summit 1983) and others maintaining that the frequency of false allegations has become a serious problem (e.g., Coleman & Clancy 1990; Raskin &.Yuille 1989; Wexler 1990; Wakefield & Underwager 1988). Victims of Child Abuse Laws (VOCAL) cites the percentage of unsubstantiated allegations to support their claim that most allegations are false, since 65 percent of all reports of suspected child abuse turn out to be unfounded (Besharov 1985a, 1986). Others respond by saying that "unsubstantiated" does not mean the abuse did, not occur. Much of the disagreement over the proportion of false allegations comes from differences in how a false allegation is defined. Substantiation by child protection agencies, the opinion of a mental health professional or police investigator or the justice system outcome have been used by various researchers and writers. The terms substantiated and unsubstantiated cause confusion. Corwin et al. (1987), Paradise et al. (1988), and Quinn (1988) note that unsubstantiated does not mean the allegation is false. The allegation may have been true but the investigating authorities did not get enough information to substantiate it. At the same time, an ostensibly substantiated allegation may not be true. A study conducted for the U.S. National Center of Child Abuse and Neglect found that in about. half of the cases of families under the supervision of child protection, the parents had never actually maltreated their children (Besharov 1985a). In addition, the legal meaning of substantiated, founded, and indicated can differ from state to state. Some writers use the term false allegation to include all cases that cannot be substantiated (Thoennes & Pearson 1988b; Sink 1988b), while others use it only when there is a deliberate falsification and exclude misunderstandings made in good faith. In order to estimate the proportion of false allegations, the definition must be specified since if the definition excludes cases that are not deliberately fabricated, there will be a much smaller proportion of them. For example, Jones and McGraw (1987) reported that only 8 percent of the cases they studied were false. But examination of their data indicates that only 53 percent were founded, even including cases that were later recanted. Of the remaining 47 percent, there was insufficient information in 24 percent, 17 percent were unsubstantiated and 6 percent were judged to be deliberately fictitious. The 8 percent comes from dropping cases with insufficient information and recalculating the percentages. This also inflated the percentage that was defined as founded, with the result that Jones and McGraw stated that 70 percent of the reports were "reliable." The estimate of the proportion of sexual abuse cases that are false will differ according to the definitions used as well as to the experience and opinion of the person making the estimate. If true abuse is defined as all substantiated cases and false allegations are limited to deliberate fabrications, there will be a small proportion of false allegations. There will be more if false allegations refer to unsubstantiated cases. If false allegations are defined in terms of the judicial system outcome, there will be still more since not all substantiated allegations result in an abuse finding in court. In this chapter, we differentiate between false and fabricated allegations. A false allegation occurs when abuse is judged not to occur, while a fabricated allegation is limited to a purposefully and deliberately false allegation.
DIVORCE AND CUSTODY DISPUTES Many professionals believe that the largest percentage of false allegations are in divorce and custody disputes, but there is disagreement over exactly how often this happens. Thoennes and her colleagues (Thoennes & Pearson 1988a, 1988b; Thoennes & Tjaden 1990) attempted to get information on the incidence and validity of sexual abuse allegations through telephone interviews and mail surveys from court administrators, judges, custody mediators, and child protecti6n workers throughout the United States. They report that the initial survey and interviews revealed a general consensus that sexual abuse allegations in custody disputes occur in "a small but growing" number of cases (Thoennes & Pearson 1988a). They estimate that accusations of sexual abuse are found in approximately 2 percent of contested custody cases (the range across court sites was 1 percent to 8 percent). They observe that there are approximately one million divorces annually, and of these, about 55 percent, or 550,000 involve minor children. About 15 percent of these (82,500) result in court involvement due to custody and/ or visitation disputes. Their estimate of 2 percent sexual abuse accusations in 82,500 custody disputes translates into 1,650 cases of sexual abuse accusations annually within the environment of a divorce/custody dispute. Others, however, suggest a higher frequency. Esplin (1990) estimates that sexual abuse allegations occur in 15 percent of contested custody situations in Arizona. Many professionals believe that various factors in the divorce and visitation situation increase the possibility of false sexual abuse allegations (Benedek & Schetky 1985a, 1985b; Blush & Ross 1987 1990 Cooke & Cooke 1991; Ekman 1989; Elterman & Ehrenberg 1991; Fisk 1989; Gardner 1992; Green & Schetky 1988; Guyer & Ash 1986; Hindmarch 1990; Horner & Guyer 1991a, 1991b; Kaser-Boyd 1988; Levy 1989; Raskin & Yuille 1989; Ross & Blush 1990; Sheridan 1990; Spiegel 1986; Wakefield & Underwager 1991). There is disagreement over how many of these accusations are false, although most estimates range between 20 percent and 80 percent. Others, however, believe that most allegations arising in divorce and custody disputes are true. Faller (1990) suggests three possible reasons for an abuse disclosure surfacing during a divorce in addition to a false allegation: (1) the nonoffending parent finds out about the sexual abuse and decides to divorce the offending parent (2); there is longstanding sexual abuse that is revealed only in the context of divorce; or (3) there is sexual abuse that has been precipitated by the marital dissolution. Several authors (Berliner 1988; Corwin, et al. 1987; Faller 1990; MacFarlane 1986; Sink 1988b) suggest reasons why valid allegations of sexual abuse may not surface until the time of a divorce. A sexually abused child may be afraid to disclose while the family is still together. A child who has been threatened with the breakup of the family may tell once this has already happened. It is more difficult for the abusing parent to persuade the child to keep the secret once he or she is not living with the child. A child may become genuinely terrified at the prospect of spending time alone with the abuser and therefore tell in order to avoid a visit. Increased distrust between parents results in willingness to suspect sexual abuse. Some writers believe, without substantiating evidence that the divorce itself is a risk factor for actual abuse and that the father may begin sexually abusing his child because of the stress and emotional devastation of the divorce (Corwin, et al. 1987; Goodwin et al. 1989; MacFarlane 1986).
EVALUATING SEXUAL ABUSE CASES Memory, Suggestibility and Interviews of Children Understanding the nature of memory is necessary in evaluating child sexual abuse. The fact that memory is reconstruction and not recall is supported both in laboratory studies and in surveys (Bradburn et al. 1987; Dawes 1988; Goodman & Hahn 1987; Loftus & Ketcham 1991; Loftus et al. 1989). Although people introspectively believe that their memories are a process of uncovering what actually happened as though a videotape had been made in the brain and is being replayed, in reality our memories are also reconstructed in accordance with our current beliefs and feelings. Loftus and Ketcham (1991) describe the reconstruction process and how people can come to believe firmly in events that never happened:
Children are often subjected to repeated interviews, sexual abuse therapy, and interactions with adults who believe that the abuse is real. The adults may ask a series of leading questions in which they provide information to the child about what supposedly happened. They may even tell the child that they already know about the abuse. Through this process of adult social influence, the adults may create stories about abuse that can become incorporated into the child's memory. Understanding this process of adult social influence is necessary in assessing a child's statements about abuse. Although repeated and/ or suggestive interviews do not mean that a child has not been abused, they can make it very difficult to sort out what, if anything, has happened. The influence of multiple interviews over time by persons who have a prior belief about what they think happened and who ask suggestive questions is discussed by several researchers in a recent American Psychological Association study edited by John Doris (1991). A number of writers have examined the factors of memory development, cognitive and moral development of children, and suggestibility of children to adult social influence (e.g. Ceci & Bruck 1993; Doris 1991 Garbarino & Stott 1989; Lassiter, Stone, & Weigold 1987; Lepore 1991; Lindsay 1990; Loftus & Ketcham 1991; Wakefield & Underwager 1988; Underwager & Wakefield 1990). There is no doubt that children can be led to make statements about and sometimes believe in events that have not happened. This does not mean that children lie, but rather that they are influenced by the adult's agenda and beliefs (Wakefield & Underwager 1988). Research on Memory and Suggestibility of Child Witnesses After the turn of the century there were many studies done on children's memory and suggestibility. Many professionals concluded from this that children were not reliable witnesses, and demanded that children's testimony be excluded from the court record. The overall picture from the early studies is of a potentially accurate witness, who can recount events and answer non-leading questions fairly accurately, but whose report can easily be contaminated by suggestion (Goodman 1984). Later studies compared the memory and suggestibility of children to that of adults. Loftus (1979) and Loftus and Davies (1984) summarize the results of studies that demonstrate how adults' memories are influenced by suggestion. The question that has been investigated in studies of children's memory and suggestibility is how suggestible they are as compared with adults. Until recently there was some disagreement concerning whether children are more susceptible than adults to distortions of memory caused by leading questions. The older studies concluded that children were more suggestible. Later studies found that both adults and children are influenced by leading questions, but were inconsistent as to whether children were more suggestible. However, recent studies that have used younger children and attempted to employ a more natural or "ecologically valid" context have found young children to be more suggestible than adults and younger children to be more suggestible than older children (Doris 1991; Underwager & Wakefield 1990). Some recent studies have provided dramatic demonstrations of the degree to which young children can be influenced by an interviewer (Ceci & DeSimone 1992; Clarke-Stewart et al. 1989; Haugaard & Alhusen 1992; Thompson et al. 1991). One of the most consistent research findings is that, although young children can provide accurate, information, they recall less than do adults (Lepore 1991). But the less information the child gives in free recall, the sooner the interviewer will turn to using leading questions, which can influence the child and distort the story. Children may have a different perception of the interview than do the adults and therefore try to tell the interviewer what they believe the interviewer wants the to say. (Ceci et al. 1987; Cole & Loftus 1987). They may answer questions they do not understand and about which they have no information (Hughes & Grieve 1983). Although, the newer research comes closer to duplicating the situation of a child witness in an actual case of alleged sexual abuse, the situation faced by a child witness in a real case cannot be reproduced in a laboratory study. With a few exceptions the suggestions asked in the typical suggestibility experiment are much weaker than are found in an actual interview. Children are presented the misleading information once and may be given two or three misleading questions. But in the real world, interviews with young children often include many leading and suggestive questions and use significant pressure and coercion for the child to affirm abuse (DeLipsey & James 1988; Slicner & Hanson 1989; Underwager & Wakefield 1990; Warren et al. 1992). In laboratory studies children are generally tested immediately or after several days. But in sexual abuse cases in which children are required to testify as witnesses, they are often interviewed many times by a variety of people over a long period of time. Poole and White (1992) found young children were consistent if repeated but appropriately open-ended questioning was used immediately or one week after the event. However, when the children were questioned again two years later, repeated questioning increased inaccuracy. The authors noted that the children often seemed simply to make up responses. They conclude that, although children can be prompted to discuss a remote event, this procedure is not without risks. It cannot he assumed that results from studies using short retention intervals can be generalized to actual cases which often have long delays. When the child eventually testifies, the memory will consist of a combination of recall and reconstruction influenced by all of the interviews, conversations, and therapy sessions that have occurred during the delay. The longer the delay, the greater the possibility of social influence and the more the memory may consist of reconstruction rather than recall. In addition, as Raskin and Yuille (1989) note, providing testimony in a sexual abuse case may have profound effects on the child's life. Whereas in the laboratory studies, the child's statements have no serious consequences, in real life, what the child says about sexual abuse may lead to the breakup of the home, inability to see a parent, or placement in a foster home. Ceci and Burke (1993) determined that there are three conclusions which can be said to be generally accepted in the scientific community. First, there are significant age differences in suggestibility, with younger children more vulnerable to suggestion than older children and adults. Second, extreme positions, such as "Children never lie" versus "Children cannot distinguish reality and fantasy," are not supported. Young children do in fact make false claims about central events, particularly events that could he construed as sexually abusive. Third, children are capable of recalling much that is forensically relevant. However, it is of utmost importance to examine carefully and critically the conditions at the time of initial disclosure and the process since then to which the child has been subjected. Unsupported Interview Techniques Although so-called anatomically correct dolls are the most frequently used interview technique, investigators of sexual abuse also use a variety of other techniques with children (Kendall-Tackett 1992). These include the use of books, puppets, drawings, projective cards, play dough, games, and play therapy. None of these are reliable or valid techniques in assessing possible sexual abuse. Their use is likely to contaminate and influence the statements children may make. The suggestive nature of these techniques is increased when the interviewer encourages the child "to pretend." Anatomical Dolls Anatomically detailed dolls that are genitally differentiated are used by professionals from a variety of disciplines (Boat & Everson 1988; Conte, et al. 1991; Kendall-Tackett & Watson 1992). The dolls, however, have proven to be extremely controversial and there is disagreement in the professional community as to whether they should be used (e.g., Yates & Terr 1988). White and her colleagues (White, Strom, and Santilli 1986; White, Strom, and Halpin 1986) developed a structured protocol for the use of the dolls, but this has not been validated or accepted in clinical practice. Realmuto et al. (1990) used this protocol and found that their raters were unable to correctly classify the children as abused or nonabused. Only a few researchers, however, attempt to use any kind of protocol. The, American Psychological Association (APA Council of Representatives 1991) issued a statement concluding that there are no normative data for the dolls and their use, and that there are no uniform standards for conducting interviews with the dolls. But despite this, they are often used. Unfortunately, the persons actually using the dolls are often untrained, unsophisticated, and vary widely in their interpretation of children's behavior with the dolls (Boat & Everson 1988). Skinner et al. (1992) observe that dolls users are likely to develop personal norms based on their own experience and caution practitioners against equating their own experience with scientific evidence. Modeling the dolls can be a learning experience for a child. Interviewers modeling the dolls suggest that they be undressed (or undress them for the child) and label them for the child. They ask the child to show with the dolls what the accused perpetrator did and may even place the dolls in sexually explicit positions for the child. This is, in effect, a teaching experience for the child. Several studies suggest that some nonabused children engage the dolls in sexual play (Dawson & Geddie 1991; Dawson et al. 1992; Everson & Boa 1990; Mclver et al. 1989). The rationale given for using the dolls is that it enables young children who have difficulty verbalizing their abuse to demonstrate it. But there are no data supporting the belief that a child who is unable to talk about what happened can then accurately demonstrate the event using a doll. In fact, one study found decreased accuracy when young children were asked to show on a doll where they were touched compared to being asked to show on their own bodies (DeLoache 1991). Another rationalization offered is that the dolls enable an interviewer to go through a body parts naming procedure and learn the child's idiosyncratic terms for the genitals. This is often done early in the interview. When the questioning about body parts ends with the identification of the genitalia, as it often does, the child has likely been taught that the purpose of the interview is to talk about sexuality. There is no evidence to support the efficacy of this procedure. An interviewer does not need dolls to talk to children about sexuality or to learn the child's vocabulary. Props are not used to talk to children about anger, violence, play, kindness, or any other behavior about which they may have a unique, individualized vocabulary. In summary, there is no evidence that doll interviews are a reliable method for obtaining accurate information about sexual abuse. The studies that claim to show differences between the responses of sexually abused and nonabused children have major methodological shortcomings which limit any conclusions that can be drawn from them (Ceci & Bruck 1993; Underwager & Wakefield 1990, Wakefield & Underwager 1989a; Wolfner et al. 1993). In a recent review of the empirical research, Skinner, et al. (1992) conclude that distinct patterns of play of abused versus nonabused children have not been identified and that the lack of norms calls into question the forensic use of the dolls. Wolfner et al. (1993) assert that there is no scientific evidence to justify clinical or forensic diagnosis of sexual abuse on the basis of doll play. Levy (1989) argues that any statement by a child that is the product of a doll-aided evaluation should be inadmissible as evidence. Drawings Children's drawings, such as the House-Tree-Person (HTP) and Kinetic Family Drawings, as well as free drawings, are often used in assessing possible sexual abuse. The assumption is that the drawings of children with emotional problems will differ from the drawings of normal children (DiLeo 1973; Koppitz 1968; Myers 1978; Yates et al. 1985). Qualitative features of the drawings may be seen as signs that the child has been sexually abused. These may include features such as the colors used, the size and detail of body parts such as hands, and the shape of the figures. However, drawings lack validity and reliability as projective assessment devices. In a review of the Draw-A-Person test in the Seventh Mental Measurements Yearbook, Harris (Buros 1972) notes that there is very little evidence for the use of "signs" as valid indicators of personality characteristics. With children's drawings there is so much variability from drawing to drawing that particular features of any one drawing are too unreliable to say anything about them. The Tenth Mental Measurements Yearbook, (Buros 1989), in reviews by Cundick and Weinberg (p. 422-25), continues the consistent finding since the first edition (1938) that interpretations of drawings (as are often done in forensic evaluations) are not supported by data. Both reviewers note that there are no normative data establishing reliability and validity of the Kinetic Drawing System. There are serious difficulties in the few research studies that report differences between the drawings of abused and nonabused children (Wakefield & Underwager 1988, 1989a). There is no good research establishing that the drawings can be used diagnostically to substantiate sexual abuse. The only valid use of drawings lies in establishing rapport and encouraging the child to talk. For example, Burgess and Hartman (1993) describe a technique for using drawings as an associative tool in assessing and accessing stored memory. However, if drawings are used this way, the interviewer must avoid selective reinforcement and cueing of responses. Another type of drawing often used is an outline of the back and the front of a boy or a girl. The child is shown the outline and instructed to put an X where he or she was touched. There is no research on this technique. The way it is often used may well give the child the message: "You were touched, now show me where." Books and Projective Cards Children's books about sexual abuse are often used in assessment sessions with children who are thought to have been sexually abused. A typical book is Red Flag Green Flag People (Rape and Crisis Abuse Center, 1985). In this book, the child is led through a series of pages that present good touch and bad touch. After several pages, a child is asked to color portions of a figure that may have been touched. When the child colors a genital area, this is regarded as evidence that the child has been abused. However, neither this book nor any others have been validated for diagnosing child sexual abuse. They may, teach the child about what is expected to be said in the interview. Two special techniques have been developed for assessing suspected Satanic ritual abuse. The Projective Story Telling Cards (Northwest Psychological Publishers 1990) and Don't Make Me Go Back, Mommy: A Child's Book about Satanic Ritual Abuse (Sanford 1996) contain explicit pictures illustrating Satanic rituals and are used to encourage the child to describe the abuse. These unvalidated techniques are likely to frighten a child as well as teach about what is supposed to occur in Satanic ritual abuse. Play Therapy Observations A child's behavior in play therapy may be used to substantiate abuse. Such therapy is sometimes called disclosure-based and the sessions focus on reenactments and talking about the alleged abuse. Although there is little evidence that play therapy is an effective therapeutic procedure for victims of child sexual abuse, (Campbell 1992a; Underwager & Wakefield 1990; Wakefield & Underwager 1988), children are frequently given disclosure-based play therapy for sexual abuse before there has been any legal determination that sexual abuse has occurred. Jones (1991) says that the use of the term "disclosure work" itself suggests the interrogator has a preconceived bias and is not able to consider the alternative, namely, that there is nothing to disclose. A child who does not produce abuse statements or behaviors symbolically interpreted as suggesting abuse may be seen as being in denial and the questioning continues relentlessly. The play is often combined with questions and the behavior of the child in the play sessions is used to form conclusions about abuse. But there is no support for the supposition that behaviors in play therapy can be used as signs to establish the truth of past events. Campbell (1992b) notes that play therapy can influence children to accept the beliefs of the therapist and can be a contributing factor to false allegations. Children's play has been shown to be aggressive, to use pretend monsters, and to express in behavior what may appear to be concepts of anxiety, fear, and sexuality (Archer et al. 1988; Best 1983; Bretherton 1982; Bretherton & Beeghly 1982; Fein 1981; Gundersen et al. 1981) but which instead may be normal, nonpathological behavior (Trad 1990). Of male and female children ages three to six, 75 percent tell stories that are violent and aggressive (Ames 1966). Fantasies of children express fear of monsters and mishaps (Pitcher & Prelinger 1966). Boys and girls show different patterns of play behavior; boys' aggressive behavior may be expressions of friendliness (Condry & Ross 1985; Maccoby 1990). Many of the persons using play therapy and interpreting the play behavior appear to be uninformed about the research literature and the actual nature of children's play activities. Their personal and idiosyncratic values are then the basis for erroneous and unrealistic over interpretations and misinterpretations. An example is a 5-year-old boy who played with the figures of popular mythical TV heroes and depicted aggressive behaviors. A sexual abuse "validator" interpreted these behaviors to mean that the child had been sexually abused. Effective Interviews Even very young children can provide accurate and useful information if the adults know how to interview them effectively. As mentioned above, the free recall of children is accurate. If the child is interviewed carefully and leading questions and suggestions avoided, the child can provide forensically useful information. Jones and Krugman (1986) describe a case of a 3-year-old girl, who had been abducted, sexually abused, and left to die in a mountain outhouse. When found and later interviewed by the police she was able to describe what had happened and to pick out the perpetrator from a lineup. The perpetrator ultimately confessed, confirming the accuracy of the child's description. Therefore, the task of the interviewer is to tap into a child's accurate free recall by encouraging the child to tell in his or her own words what has happened. Garbarino and Stott (1989) conclude that the problem is not that children cannot give reliable information, but rather that adults do not know how to get it from them. Several professionals have suggested guidelines for conducting an unbiased evaluation, and noncontaminating interview (e.g., Daly 1991, 1992a, 1992b; Quinn et al. 1989; Raskin & Yuille 1989; Slicner & Hanson 1989; Underwager & Wakefield 1991b; Wakefield & Underwager 1988; White 1990). The interviewer must enter the interview with an open mind about what may or may not have happened. There should be a conscious decision and continued conscious determination to delay decision making. Unfortunately, mental health professionals often reach a quick diagnosis, within thirty to sixty seconds, and on the basis of idiosyncratic and unvalidated pathognomic signs (Gauron & Dickinson 1969). When this happens, the likelihood of error is increased, for once a choice is made, evidence may be fabricated to support the decision and contrary information ignored (Arkes & Harkness 1980). To prevent this, all possible alternatives should be considered until there are sufficient data to rule out those that are inapplicable. This might be done by using a form or writing out the alternatives to be considered. This approach has been shown to be effective in eliminating clinicians' bias (Arkes 1991). Steller (1990) and Daly (1992b) also recommend formulating all possible alternative hypotheses about what led to the accusations of abuse. These hypotheses are successively explored and ruled out during the interview. Throughout, the interviewer should ask open-ended questions and encourage the child to provide a free narrative. Details should be encouraged by responses such as "and then what happened." Pressure and coercion, leading questions and selective reinforcement of responses, and unvalidated techniques must be avoided. The interviewer should be fully informed about the serious consequences that may result if there is an error. For example, the reality of foster care should be known so that there is not the automatic assumption that the state can be a better and effective parent. This is a possible way to increase the response costs to decrease strategy-based errors (Arkes 1991). In-service training programs should include material that clarifies the potential for harm by the system. King and Yuille (1987) stress that the child be told that the interviewer is only interested in what the child remembers, and that admissions of memory failure and memory gaps are expected. The child should be instructed (in an appropriate fashion), "I don't want you to say something that didn't happen, but if something did happen, I want you to tell me about it." Repeated questions should be avoided since this tells the child that the previous answers were not acceptable. Discussions of "good touch" and "bad touch" should not be used since these are confusing and potentially contaminating. There is beginning research on a European procedure for interviewing children suspected of being abused and for analyzing the resulting interview. Criterion Based Content Analysis/Statement Validity Analysis (CBCA/SVA) assumes that an account based on memory of an actual event will differ in content and quality from accounts that are based on fabricated, learned, or suggested memory. The procedure requires a relatively complete statement obtained as soon as possible after the child has disclosed an incident; the interview must also be designed to obtain as much free narrative as possible. Leading questions and suggestions must be avoided. The interview is tape-recorded and transcribed for later analysis (Köhnken & Steller 1988; Raskin & Esplin 1991; Rogers 1990; Undeutsch 1989). Preliminary research on CBCA/SVA is promising. Raskin and Esplin (1991) found that the techniques discriminated between twenty interview statements of children independently determined to have been abused and twenty statements in cases determined to be false. Horowitz et al. (1992) found high reliability for the criteria used in CBCA/SVA as a whole, although there was variability for individual criteria. A professional may be asked to assess a case after the child has been interviewed by others. In such cases, the progress of the case and the procedures followed by the previous evaluators must be carefully examined in order to assess possible contamination (Wakefield & Underwager 1988; White & Quinn 1988). What happens to a child during the process of investigation arid the interim between beginning an investigation and reaching trial or dispositional hearing is crucial for making the best possible and most accurate decisions (Ceci & Burke 1993). Children are often interviewed repeatedly by a variety of trained and untrained adults, and may be put into disclosure-based sexual abuse therapy. When children have been subjected to leading and coercive interviews, their recollections may become so contaminated that it is extremely difficult to determine what happened. Therefore, it is essential to analyze all contacts with the child in which abuse was discussed. All interviews of the child should be videotaped, or at least audiotaped, since a tape is the only means whereby the procedures and information obtained during the interview can be accurately documented (DeLipsey & James 1988; Herbert et al. 1987; Raskin & Yuille 1989; Underwager & Wakefield 1990, Wakefield & Underwager 1988, 1989a). In practice, this is often not done. Many prosecutors do not want the defense to have access to a tape so that they can criticize the interviewer's techniques during the trial (Stern 1992). This is hardly a legitimate argument if the goal is accuracy and avoidance of harm to children by the system. Behavioral Indications Lists of behaviors thought to be caused by sexual abuse have been widely publicized (e.g., Cohen 1985; Council on Scientific Affairs 1985; Sgroi 1982). These behaviors are quite inclusive; nearly every problem behavior exhibited by children has been offered by someone as a sign of possible sexual abuse. As a result, when these behaviors are observed, an adult may become suspicious and begin questioning the child. Particularly when the questioning adult is untrained and believes abuse is likely, the questioning may result in answers that are interpreted as indicating abuse. However all of the behaviors mentioned on the lists are found in nonabused children. They appear in many different situations, including conflict between parents; divorce; economic stress; wartime separations; father absence; natural disaster; and physical, emotional, but nonsexual abuse (Emery 1982; Hughes & Barad 1983 Jaffe et al. 1986; Porter & O'Leary 1980; Wallerstein & Kelly 1980; Wolman 1983). Although children who are distressed for whatever reason may show behavior changes, there are no behaviors exclusive to victims of sexual abuse. A recent review of the literature concluded that with the exception of sexualized behavior the majority of symptoms shown in sexually abused children characterize child clinical samples in general (Beitchman et al. 1991). Even sexualized behavior cannot be used as proof of abuse. What children normally do sexually is more involved than most people believe (Best 1983; Gundersen et al. 1981; Langfeldt,1981; Martinson 1981; Okami 1992). Friedrich et al. (1991) asked mothers of 880 nonabused 2- to 12-year-old children to complete questionnaires concerning sexual behavior. Although behaviors imitative of adult sexual behaviors were rare, the children exhibited a wide variety of sexual behaviors at relatively high frequencies. Mannarino et al. (1991) report no differences in sexual behavior between abused girls and a clinical control group, although both scored higher than did the normal controls. Haugaard and Tilly, (1988),found that approximately 28 percent of male and female undergraduates reported having engaged in sexual play with another child when they were children. On the other hand, since many sexually abused children do not suffer significant trauma (Browne & Finkelhor 1986; Gomes-Schwartz, et al. 1990; Wakefield & Underwager 1988) an abused child may not exhibit any behavioral signs. It is thus a mistake to use the absence of behavioral signs as support for an allegation being false. Therefore, using behavioral Indicators to assess sexual abuse may result in a mistake in either direction. Besharov (1990) observes that behavioral indicators by themselves are not a sufficient basis for a report. Levine and Battistoni (1991) state that none of these indicators, in any combination, are valid without a direct statement by the child about sexual involvement or sexual knowledge. This entire procedure of attempting to infer a posteriori to an alleged prior event and claim that a present observation establishes the facticity of the alleged earlier event is the logical error of affirming the consequence. It will result in increased error and will produce harm (Gambrill 1990). In the absence of credible data supporting a strong causal relationship, this procedure should not be used. Physical Evidence Although medical examinations are common when there is an allegation of sexual abuse, the results are frequently ambiguous. Much abuse involves exhibitionism, fondling, and masturbation of the perpetrator rather than anal, oral or vaginal penetration, and therefore there will be no physical or medical evidence. However, police, social workers, juries, prosecutors, defense attorneys, and therapists have come to rely on physicians to find evidence that will either rule out or substantiate the abuse (Coleman l989; Paradise 1989). It may be that too much credence is given to medical evidence when there are no conclusive data for almost all of the physical observations that are made. Nevertheless, a physical examination is important, particularly when the allegations are of behaviors likely to result in physical sequelae. There have been attempts made to classify the physical findings in terms of how specific they are to sexual abuse (Bays & Chadwick 1993). For example, the presence of semen or sperm would be classified as a definitive finding of sexual abuse; marked enlargement of the hymenal opening, recent laceration of the hymen, or laboratory confirmation of a venereal disease would be a specific finding, and various abnormalities of the genitals would he a nonspecific finding. The examination should take place as soon as possible since genital injuries heal rapidly and physical signs will be difficult to detect if the examination is not performed immediately after the alleged event (Bays & Chadwick 1993; McCann et al. 1992). Paul (1990) describes the expected physical signs when there is abuse but cautions physicians concerning the risks of an incorrect interpretation. Sometimes the absence of an expected finding can be important, such as an intact hymen when full vaginal penetration is alleged, or the failure of the child to describe severe pain during penetration or when defecation is attempted after allegations of anal rape. Unfortunately, there is generally little consideration given as to whether the physician's conclusions are speculation or legitimate scientific evidence. A 1983 paper by Cantwell is still frequently cited in support of the belief that a vaginal opening size above four millimeters indicates abuse, although there has never been empirical support for this claim. Other normal findings, or vague and ambiguous findings, such as genital redness, are seen as "consistent with" sexual abuse. Physicians often rely on the history given by the person who brings the child in for the examination and conclude, "sexual abuse based on history." Such statements are taken seriously by others and repeated as evidence that the physical examination has substantiated the allegation. An analysis by Paradise (1989) suggests that 65 percent false positives occur when assessing penile penetration and 73 percent false positives with assessment of digital penetration. This raises serious questions about the validity and reliability of medical examinations Until recently, a major difficulty in evaluating physical findings was the absence of baseline data — that is, information about the appearance of the genitals in normal, nonabused children. This situation has now changed. Adams (1992) observes that there are now enough data on normal and presumably nonabused children to classify many findings as either normal or nonspecific. McCann and his colleagues conducted research on 267 prepubertal children, screened to rule out sexual abuse (McCann et al. 1989, 1990; McCann et al. 1990). They found a high incidence of nonspecific findings such as erythema (redness of the skin), tags, fissures, scars, adhesions, notches, thickening, and anal relaxation in their sample of nonabused children; therefore, they urge physicians to exercise caution in giving an opinion about the significance of such findings. They also found a large range of vertical and horizontal hymenal orifice diameters and report that this varied not only by age group, but according to the technique and position used to measure it. Emans et al. (1987) also found a large range of hymenal openings in their subjects and report that the genital findings of sexually abused girls were similar to nonabused girls who had other genital complaints such as vaginitis, vulvitis, bleeding, or dysuria (difficulty in urinating). Bays and Chadwick (1993) note that there are a variety of dermatologic, traumatic, and infectious conditions that may mimic findings caused by sexual abuse. Coleman (1989), after analyzing 158 medical examinations in cases of alleged sexual, abuse, reports that nearly all the findings attributed to abuse were found by McCann in substantial numbers of the normal children he had examined. Coleman, however, does not differentiate between children who were later determined to be abused or nonabused. In 216 cases we have analyzed arising in the context of divorce, similar vague and nonspecific findings were commonly interpreted as supporting the allegation. This was also true in the majority of the cases we classified as false allegations and where abuse was not found by the justice system. Sexually transmitted diseases support a suspicion of sexual abuse since sexual contact is the most common means of transmission. Sexual transmission of diseases outside of the perinatal period is rare (Bays & Chadwick 1993). However, since there can be alternative explanations for the transmission of sexual diseases, a finding of a venereal disease does not conclusively establish sexual abuse (Wakefield & Underwager 1988). In addition, the testing performed may be inaccurate or inappropriate. For example, a chlamydia screening test meant for an adult may be highly inaccurate with a child, and produce false positives because the test reacts positively to certain bacteria normally found in the intestinal tract of children (Fay 1991). The only specific and unambiguous physical findings demonstrating sexual contact are pregnancy or sperm in the vagina or anus. Krugman; (1989) concludes: "The medical diagnosis of sexual abuse usually cannot be made on the basis of physical findings alone. With the exception of acquired gonorrhea or syphilis, or the presence of forensic evidence of sperm or semen, there are no pathognomic signs for sexual abuse" (pp. 165-66).
SUGGESTED CRITERIA FOR DIFFERENTIATING BETWEEN REAL AND FALSE ALLEGATIONS The Origin and Timing of the Original Disclosure The allegations are less likely to be correct when the parent, rather than the child, initiates the disclosure (Yates & Musty 1988; Yates 1988). The disclosure may have come about only after the parent becomes suspicious after observing one of the behavioral indicators and begins questioning the child (Wakefield & Underwager 1988; Rogers. 1990). Although adolescents some times initiate false allegations for a variety of reasons, young children almost never initiate false allegations without influence from an adult. Klajner-Diamond et al. (1987) note that false allegations are most likely the result of adult indoctrination rather than childhood fantasy. A spontaneous disclosure made by a young child without evident adult influence is more likely to be true. The timing and circumstances of the allegation are also important, especially in the divorce and custody context. Although both true and false accusations of sexual abuse can occur at at any point (Blush & Ross 1987; Underwager & Wakefield 1988; Wakefield & Underwager 1991), there is a difference between an accusation that appears in a continuing marriage and an accusation that first appears during a custody or visitation battle. If the allegation appears in the midst of a custody or visitation dispute, it is more likely to be false. But if it can be determined that the divorce occurs as a result of the abuse disclosures, the abuse is more likely to be true. Sirles and Lofberg (1990) studied 128 families in which sexual abuse occurred: approximately half of these families ended in separation and/or divorce. The Age of the Child Allegations that turn out to be false appear to involve very young children. Schaefer and Guyer (1988) report that the children in their false cases were most often under five years old. Everstine and Everstine (1989) note that a younger child whose parents are divorcing may be more vulnerable to the manipulations of an angry and vengeful parent. Our analysis of 216 cases of sexual abuse allegations arising in a divorce context, which involved 325 children, found that the median age was 8.6 when the allegations were probably true, and 4.6 when the allegations were probably false. The Nature of the Allegations Schaefer and Guyer (1988) observe that in cases involving false allegations in the divorce and custody context, the allegations were often extremely vague. Often the allegation involved no specific behavior but only a vague observation that something was happening. Also, normal parenting behaviors such as bathing, toileting, and tickling may be mistakenly labeled as sexual abuse (Cooke & Cooke 1991). Rosenfeld and his colleagues (Rosenfeld et al. 1986; Rosenfeld et al. 1987) stress getting normative information on nakedness, genital touching and bathing practices before deciding whether any of these behaviors supports a suspicion of sexual abuse, since they found that many behaviors that could trigger suspicion of abuse occurred often in normal families. When there is no corroborating evidence, and the behaviors alleged are highly improbable, it is unlikely that the allegations are true (Wakefield & Underwager 1988, 1991, 1992b). There is information about the behavior of known sexual abusers (e.g., Erickson et al. 1988; Tollison & Adams 1979). Fondling is common and aggression and physical violence are rare. Bribery is more common than threat. Vaginal and anal penetration are rare in very young children because it is so painful. Girls are more likely than boys to be victims. Allegations of ritual abuse should be treated with great suspicion. Although such allegations are often directed at multi-perpetrator, multi-victim day care situations, they are occasionally seen in divorce and custody battles. Both parents are also occasionally accused of ritual abuse with their own children. However, despite hundreds of investigations by the FBI and police, there is no independent evidence supporting the existence of organized cults of outwardly normal people who engage in ritual abuse, animal and human sacrifice, murder, and cannibalism of children (Bromley 1991; Hicks 1991; Jenkins & Maier-Katkin, 1991; Lanning 1991, 1992; Mulhern 1991; Noll 1989; Nathan 1991; Putnam, 1991; Russell 1991; Underwager & Wakefield 1991a; Victor 1991a, 1991b; Wakefield & Underwager 1992a; see also Victor in this volume). Therefore, when the allegations are of very low frequency behaviors such as rape, physical violence, vaginal or anal penetration of a very young child, use of feces and urine, and/or ritual abuse with several people involved, the allegations are unlikely to he true. Characteristics of the Accused Deviant and bizarre allegations are even less likely to be true if the accused has been evaluated and found to be psychologically normal. Although child sexual abusers form a heterogeneous group, they tend to have behavioral and psychological difficulties. Studies on the Minnesota Multiphasic Personality Disorders (MMPIs) of sex offenders may show elevations on scales reflecting poor impulse control and judgment, antisocial behavior, a history of "acting out," lack of self-esteem, feelings of inadequacy, a schizoid social adjustment, much time spent in fantasies, and/or thought disorders and confusion. However, since some sex offenders produce MMPIs within normal limits — Erickson et al. (1987) found this in 19 percent of their convicted sex offenders sample — a "normal" personality profile on the MMPI or other assessment technique does not rule out the possibility that the individual is a sexual abuser. But knowledge of the accused person's personality characteristics becomes more useful as the alleged abusive behaviors become more bizarre and deviant. It is not likely that a psychologically normal individual would engage in sadistic and violent sexual behaviors. Characteristics of the Child's Statement There is general agreement concerning the characteristics of a child's statement in cases of actual abuse compared to false allegations, although little research has been done thus far on this important variable. Contextual details are seen by most professionals as important. DeYoung (1986) notes that affective details, specific action, and contextual details are found in valid accounts and that in a false allegation, the child will be unlikely to give such details. Jones and Seig (1988) observe that when the allegation is false, the child will give an inconsistent, sparse, or unrealistic account. Sink (1988a) believes that in real abuse the child will spontaneously provide contextual descriptive information and that the child's affect will be appropriate to the report. Jones and McGraw (1987) state that a child's account of actual abuse will have, relative to age, a sufficient amount of uniqueness and idiosyncrasy of detail, appropriate emotion, and reports of secrecy, coercion) or threats." Faller (1988) analyzed 103 cases in which the perpetrator had confessed to some level of the abuse. She examined the child's statements for contextual details, the description or demonstration of the sexual victimization, and the child's emotional state. She reports that a description of the sexual behavior of the sort said to be found in a true allegation as well as an emotional reaction to the sexual abuse was found in over four-fifths of the statements. Contextual details were found in over three-fourths. Over two-thirds of the allegations contained all of these characteristics. Faller therefore concluded that this study supports the clinical assumptions concerning these criteria. These characteristics are similar to several of the criteria looked for in the CBCA/SVA procedure described earlier (Horowitz et al. 1992; Köhnken & Steller 1988; Raskin & Esplin 1991; Rogers 1990; Undeutsch 1989). Therefore, the quantity, quality and contextual embedding of details and appropriate affect and emotion are useful criteria when evaluating the child's statement. However, the statement must be obtained as soon as possible from a child who has made a spontaneous disclosure. It must also be based upon the child's narrative account and, not on yes or no answers to the interviewer's leading questions. In divorce and custody cases, strong hatred expressed toward the accused parent based upon trivial and vague reasons may result from learning from the accusing parent rather than from actual abuse (Gardner 1992). Ross and Blush (1990) report that in a false allegation, the child may describe "horrible" and traumatic events while not appearing to be traumatized. Also, a young child who is very eager to talk about the abuse may have learned that adults reward such talk (Wakefield & Underwager 1988). Telltale Signals of the Parties Involved Personality disorders and other psychiatric problems have been reported in parents who make false allegations (Ross & Blush 1990; Green & Schetky 1988; Jones & McGraw 1987; Rogers 1990; Klajner-Diamond, et al. 1987; Wakefield & Underwager 1990). Faller (1990) notes that a childhood history of abuse in the mother may result in distortions of events or hypervigilance. However, as Bresee et al. (1986) point out, a troubled woman with vengeful motives may nevertheless have discovered genuine evidence of sexual abuse. Several writers have described differences in the behavior of an accusing parent when there is real abuse compared to cases of false allegations (e.g., Bresee et al. 1986; Cooke & Cooke 1991; Faller 1990; Gardner 1992; Jones & Seig 1988; Wakefield & Underwager 1991). The natural response of a reporting parent is to hope that the child was not sexually abused and to be relieved when the evaluator believes abuse is unlikely. The parent will consider other explanations for the behaviors that aroused her suspicions and she may initially report not believing the abuse, or thinking that the child was mistaken. If the evaluator concludes that abuse was likely, she will probably be upset, distressed, and embarrassed. But in a false allegation, the accusing parent may remain firmly convinced the abuse is real. She is likely to feel vindicated if her suspicions are validated by the interview or medical examination, and may even tell everyone what happened and what a terrible person her former spouse was. If the evaluator says that abuse is unlikely, this parent may seek out another professional to confirm her suspicions. She may even involve the child in multiple examinations and interviews despite the effect of this on the child (Bresee et al. 1986; Cooke & Cooke 1991; Rand 1989, 1990; Schaefer & Guyer 1988). Such an experience can be iatrogenically negative for the child (Jones 1991). Although observations of the behavior and motivations on the part of the person reporting the abuse may be helpful in the divorce and custody situation, teachers, day care workers, and therapists may also be primed to find sexual abuse. They may have attended sexual abuse workshops, learned the behavioral indicators, been told to be alert to any sign of sexual abuse, and believe that sexual abuse is rampant. They may therefore observe an ambiguous behavior, conclude that the child has likely been abused and begin questioning the child in a way that results in the child giving answers that are interpreted as supporting their suspicions The professionals who have interviewed the child may be biased. Klajner-Diamond et al. (1987) note that one factor suggesting a false accusation is a professional committed prematurely to the truth of the allegation. Blush and Ross (1990) observe that false cases are characterized by a loss of control early on when professionals decide that abuse is real before doing a careful investigation. This is seldom deliberate, although there may be an occasional professional who will collude with parents to develop false accusations (Wakefield & Underwager 1989b).
ANTISEXUALITY A final observation about the current sexual abuse system is that it represents a violent and virulent antisexualism, the likes of which the world has not known since the days of Tertullian, the second-century Christian apologist who maintained that the only proper way to be a Christian was to emasculate oneself. In a recent interview, Judith Herman, a Harvard psychiatrist and early feminist who has specialized in sexual abuse, said about molesters, "They are just evil. They do it because they want to and it gives them satisfaction" (Hawkins 1991: 50). This stance is also demonstrated in the effort to link child sexual abuse with Satanic, ritualistic abuse allegations. Money (1991) sees the antisexualism of the child sexual abuse system as a new specialty of victimology and maintains there is a strong element of man-bashing that is part of a more generalized hostile reaction to the sexual revolution of the 1960s. He also sees victimology as equating sex with wrongdoing (Money 1982). Mosher (1991) traces the development of the view of children presented in the history of American child-saving: "The rebellious child became the deprived child who became the sick child who has now become the victimized child" (p. 15). This aspect of antisexualism is accepted without criticism within professional societies and accorded respectability in the professional community (Money 1992). At the same time, professional societies support the politically correct hangovers of the sexual revolution with their stances toward abortion, gay rights, and feminist rhetoric. The earliest child abuse prevention programs were proliferated by feminists. Sexual abuse prevention programs that have proliferated throughout the country are based on empowerment theory. The orientation of empowerment theory is political ideology which has antisexualism at its core (Krivacska 1991). Antisexualism may be discerned in the language game of sexual abuse with its own peculiar idiosyncratic usage of terms such as "hurt," "touch," "feel funny," "body parts," "yucky," and "uncomfortable." The system does not permit direct language about sexuality but uses circumlocutions such as "parts covered by a bathing suit." This can only communicate to children that sex is viewed negatively and cannot be talked about freely and openly. If a young child who has little knowledge of adult sexual behavior is questioned repeatedly about deviant sexuality that child has been taught a negative view of sexuality. The focus on genitals teaches a genitalized and partial view of sex that will hinder the development of concepts of intimacy and sexuality (Krivacska 1990; Nelson 1978). A distressing trend is the tendency to label childhood sexual play as sexual abuse (Okami 1992). This criminalization of what used to be understood as relatively normal sexual play and curiosity reflects the antisexualism of the sexual abuse system. An example is Cantwell's (1988) discussion of what she terms "very young perpetrators." She does not adequately define normal sexual behavior and includes in her examples a 7 year-old who was playing sexually with a 5-year-old neighbor (p. 580). Haugaard (1990) emphasizes that there is no justification for labeling mutually enjoyable sex play as sexually abusive or for labeling one or both of the children as an abuser. But this is happening. Thompson (1992) describes a situation where consensual fondling of a 14-year-o1d girl's breasts by her 16-year-olld boyfriend resulted in the boy being convicted of sexual abuse. Very young children may be sentenced to therapy programs or to various forms of detention. In Phoenix boys as young as seven were sentenced to a treatment program for young offenders using a penile plethysmograph and avoidance conditioning. They were shown specific sexual stimuli, including heterosexual, homosexual, and pedophilic acts; if the plethysmograph registered penile engorgement they are subjected to a noxious stimulus (Young 1992). Insurance companies may pay large sums of money rather than go to trial when there are allegations of sexual play between children. We consulted in one case where an allegation of sexual play between a 4-year-old boy and a 3-year-old girl in a day care setting resulted in the parents of the girl collecting from the insurers. In another case, a man who had befriended a single mother was accused of sexual abuse when he kissed her l0-year-old boy good night and patted him on the buttocks. The man was sentenced to two years in prison. The prosecutor in the closing argument told the judge that no man should ever be able to do anything that makes a child uncomfortable and excuse it by claiming he was being affectionate.
CONCLUSIONS The real victims of sexual abuse of children include all of us, because the system we have set up to eliminate abuse of children may be doing more damage than good. Children may be harmed by the intervention. Families, including extended family members, may be destroyed. Grandparents may never see their grandchildren again. Occupations that involve work with children become suspect: Teachers, preachers, boy scout leaders, big brothers, athletic coaches, day care workers, counselors, mental health professionals, and others are watched suspiciously by a society that asks why they choose to work with children. Lonely people learn quickly to avoid all friendly actions toward children (Weinbach 1987). Men learn that in spite of twenty years of rhetoric telling them they can have feelings and be gentle, if they are affectionate with children they can go to prison. Raymond Buckey (Buckey et al. 1990), who surely has the credentials to claim being damaged by the system, spoke of how everyone was a victim in the McMartin preschool case:
Unfortunately, many professionals who have worked hard and long to raise public awareness of child sexual abuse react defensively to legitimate questions about unjustified costs and false victims (Weinbach 1987). Those professionals who believe they have been ill treated by the system have great difficulty letting go of their emotional investment in establishing their innocence. Colleagues are sharply and deeply polarized. charges and countercharges fly back and forth at personal, associational, institutional, and societal levels. It must be possible for all who are concerned with reducing the abuse of children to agree cognitively that increasing the accuracy of the process is both desirable and attainable. There is more credible and reliable information now than a few years ago. It is possible now to assert, as Ceci and Burke (1993) do, that there is a general consensus in the scientific community about some basic facts. Potentially this information can be used to develop a more accurate and reliable way to make decisions about child sexual abuse which will result in greater protection of abused children and less harm to innocent persons.
|
