It is of little surprise that some degree of CEM is common in the general population. A majority of parents (45% to 86% across studies) report engaging in acts that most would define as emotional or psychological maltreatment, such as yelling, insulting, or threatening their children (Straus, Hamby, Finkelhor, Moore, & Runyan, 1998). This finding heralds back to the dilemma that is the focus of this paper: when it comes to non-physical acts toward children that leave no visible mark, how do we distinguish between "normative," but problematic, parenting behaviours and the more extreme, harsh methods that may require state intervention?
In the following section we highlight findings from this review that point to a number of similarities in parental risk factors, childrearing behaviours, and potential harm to the child among parents who use poor parenting methods versus those who are considered emotionally maltreating. These similarities indicate that these two types of parenting behaviours cannot be clearly distinguished from one another on the basis of etiology or impact on the child alone. However, both can be viewed as examples of relational disorders that differ in terms of their chronicity, severity, and potential harm. Second, we examine the major differences between poor parenting and CEM to emerge from these findings, and consider the importance of taking actual or potential harm to the child into account. Together, this knowledge leads to promising directions in advancing operational distinctions that have merit in everyday matters of family intervention and child welfare.
Numerous similarities emerged from this review of the nature, causes and impact of poor parenting and child emotional maltreatment. These similarities are summarized as:
Emotional maltreatment and poor parenting methods both impair the parent-child relationship in a manner similar to the concept of relational aggression. Relational aggression has most often been used to describe behaviour of children and youth; however, it applies to parent-child interactions as well. Relational aggression has been defined as "harm to another through purposeful manipulation or damage to the relationship" (Nelson & Crick, 2002; p. 163), as well as psychological control (i.e., "the assertion of parental authority through the use of emotionally manipulative techniques" (Steinberg, 2005; p. 72). In effect, the toxic quality of the parent-child relationship stemming from poor parenting, and to an even greater extent from emotional maltreatment, undermines the child's formation of healthy representations of relationships. Over time, this disturbance in the child's sense of self and his or her view of relationships with others increases the risk of poor achievement of additional developmental milestones, such as peer relations, emotion regulation, and others.
Because they occur in the context of critical relationship roles, child emotional maltreatment and poor parenting can both be considered types of relational disorders. The parent-child relationship is particularly salient during periods of stressful role transitions for parents, such as the postnatal attachment period, the early childhood and early adolescence "oppositional" periods of testing limits, and times of family instability and disruption. Caregivers' failure to provide nurturing, sensitive, available, and supportive care, especially during critical periods, is a fundamental feature of both poor parenting and emotional maltreatment. The notion of relational disorders describes these two types of parenting behaviours as well as parental- and partner-violence, and is gaining support in relation to upcoming revisions to the Diagnostic and Statistical Manual (Lebow & Gordon, 2006; Wakefield, 2006). Relational disorders signify the importance of the actions that occur between or among individuals, such as the parent-child relationship, that are associated with distress or impairment or the potential for such harm (Heyman et al., 2009).
Although they share many similarities, two primary factors distinguish emotional maltreatment from poor / dysfunctional parenting:
Whereas there is considerable agreement regarding the extreme nature of parental acts described as emotional maltreatment, disagreement exists as to the second issue: whether or not real or potential harm to the child is a necessary definitional feature. In particular, there is debate as to whether or not "harm" should be included as part of the definition at all. For example, during a presentation to members of the think tank, Dr. Nico Trocme suggested that a definition of CEM based on evidence of harm or a strict criteria for "reasonable potential of harm" to the child might be easier to operationalize and guide intervention (than one based on parental action alone). However, other participants expressed a different view (PHAC, 2009; pp. 16-17):
Despite these arguments, there is growing consensus that a definition of CEM must recognize and include actual or potential risk of harm to the child, in order to distinguish such behaviour from poor parenting and to warrant child protection action (Heyman & Slep, 2006). Notably, this approach presumes that potential or actual harm to the child is a requirement of the definition, but evidence of harm is not. Placing emphasis on potential harm averts the tautological predicament of defining CEM on the basis of harm (e.g., "emotional maltreatment is when the child is emotionally harmed"). Determining potential harm must also take into account the wide range of children's needs, such as children with special vulnerabilities, disabilities, or developmental differences (i.e., the impact may be significant for adolescents as well as toddlers, perhaps for different reasons).
Two models are presented below, representing complementary strategies to translate these conceptual similarities and distinctions into specific guidelines and boundaries for acceptable, poor, and emotionally maltreating childrearing behaviour methods. The first model describes a hypothetical continuum depicting a range of behaviours from positive, to poor, to emotionally abusive or neglectful. The second model presents a categorical distinction between poor parenting and CEM, using specific empirically derived criteria for certain parental acts and child outcomes. These two conceptual models have important, complementary implications for both universal and targeted prevention, as noted in the subsequent recommendations.
As we have seen, parental lack of control involves cognitive and emotional responses to children's behaviour. These processes, in turn, result not only from individual characteristics, such as personality disturbance, cognitive styles, and mood disorders, but also from important situational factors, such as intimate partner violence, unmanageable stress, and cultural expectations. A continuum view of CEM as an extreme disturbance of childrearing emphasizes the nature of socialization practices and norms that condone or inadvertently permit the use of harmful or potentially harmful methods with children, at the expense of more positive, healthy childrearing methods.
As shown in Figure 1 (Appendix 1), at the child-centered end of this continuum lie the more appropriate and healthy forms of childrearing actions that promote child development. Competent parents encourage their child's development in a variety of ways, and match their demands and expectations to the child's needs and abilities. Because most, if not all, parents scold, criticize, or show some degree of insensitivity to their child's state of need on occasion, it is important to acknowledge how such normative actions are set far apart from more strident and abusive methods. Moreover, child discipline often requires firm control, with its accompanying strong verbal statements and negative affect (e.g., scorn), which needs to be acknowledged as part of a parenting continuum.
Poor parenting methods occupying the middle range of this continuum reflect greater and greater degrees of irresponsible and potentially harmful childcare. Parents who show any measurable degree of these actions towards their children often need instruction and assistance in effective childcare methods, and may be at-risk for maltreatment. Notably, this middle range does not include parental acts that would typically lead to a child welfare response.
Finally, childrearing methods that violate children's basic needs and dependency status fall within the scope of emotional maltreatment. These parenting methods or particular acts towards the child are consistent with most definitions of CEM, and represent harsh, insensitive, and ineffective forms of childrearing. Examples of this type of parenting include use of excessive criticism and verbal harassment, use of unacceptable disciplinary techniques, lack of physical or verbal affection toward the child, failure to provide developmentally appropriate stimulation or opportunities to the child, exposure to domestic violence, and similar trauma-inducing experiences directly or indirectly caused by caregivers.
These and many similar instances of parental inadequacy or ineffectiveness often warrant professional involvement and child protection. Because determination of such events involves professional judgment, definitions of emotional maltreatment will vary somewhat in accordance with the purpose of the assessment and intervention concerns. From a continuum perspective, this ambiguity is considered to be necessary and acceptable in view of the current state of knowledge and the presumed advantage to the child and family in seeking assistance for wide‑ranging problems (as opposed to labelling or punishing family members). A continuum model is particularly advantageous for educating the public about acceptable and unacceptable forms of parenting. However, this model may be less useful for determining breaches in child welfare statutes that require intervention. To formulate clear guidelines for responding to dangerous or harmful parenting acts it is necessary to develop valid and reliable definitions of CEM, as indicated by the categorical model below. These definitions should be reviewed and updated in conjunction with progress on community standards of minimal care and research on the impact of parenting methods.
Research over the past decade conducted by Richard Heyman and Amy Slep at the State University of New York at Stony Brook, with the involvement of the U.S. Air Force Family Advocacy Program, has produced a reliable, categorical approach to defining CEM and other forms of "family maltreatment," including partner abuse, poor parenting methods, and sexual abuse. The premise of this strategy is to identify and describe potentially significant syndromes, of which child emotional maltreatment would qualify, with the eventual goal of determining their validity and utility.
The family maltreatment criteria were developed in a multi-stage process summarized in several publications (e.g., Heyman & Slep, 2006; 2009; Slep & Heyman, 2006). Essentially, they approached this task in a manner similar to that used to develop and revise the Diagnostic and Statistical Manual (DSM), whereby initial diagnostic criteria are based on expert consensus and field tested to refine their reliability and validity. With the help of the US military over 10 years, these steps comprised (a) examining the content validity and field usability of a set of maltreatment criteria already in use; (b) creating a unifying concept for what constituted an above-threshold problem; (c) reviewing and adapting (where appropriate) existing operationalizations; (d) field testing and refining criteria, assessments, and decision-making process; (e) testing criteria's use in wide-scale dissemination; (f) creating criteria-informed screeners and structured clinical interviews; and (g) examining the content validity of the final criteria.
by adopting a DSM-style framework, Heyman and Slep's categorical definitions
profit from consensus in the mental health field that any "disorder" requires
evidence of certain recognized symptoms or behaviours, as well as evidence of
actual or potential harm. Thus, their criteria for partner abuse, physical
abuse, poor parenting, and emotional maltreatment
(among other forms of family maltreatment) underscores the necessity of both a specific type of act (for example, use of physical force for physical abuse) and a significant impact or high potential for significant impact (such as shooting a gun at a spouse but not hitting her). They specifically avoided the notion of "intent" or the determination of outcomes in the future; rather, to be reliable and valid, this strategy poses specific injuries, fear reactions, or reasonable potential for psychological harm or developmental disruptions that can be determined at or near the time of the incident(s).
Heyman and Slep's categorical, diagnostic strategy involves a structured clinical interview, whereby aspects of the allegation or report are recorded in as much detail as possible (e.g., "describe what happened as if you were watching a movie"). Moreover, their assessment protocol provides a systematic assessment of impact on the child, as well as factors that affect the potential for such impact. These details are then provided to a committee or independent panel to evaluate whether or not they fit the criteria for CEM (for example). In doing so, they adopted a standard of proof based on the "preponderance of the evidence," which was felt to be the most reasonable safeguard for family members as well as children. As in civil cases, this standard requires only that the investigator determine that it is more likely than not the criterion was met (taking onto account the credibility of reporters in making such a decision).
It is illuminating to compare their criteria for poor parenting and emotional maltreatment (see Appendices 2 through 4), especially in light of the question posed by this paper: can we distinguish between poor parenting and CEM? What these criteria reveal is that the distinction is based primarily on parental acts rather than child outcomes. For parenting problems (Appendix 2), Heyman and Slep's criteria stem from the literature noted herein that highlights the two major determinants of parenting style. Caregiver difficulties involve one or more types of under-involvement or over-involvement, as well as marked difficulties in at least one aspect of parenting (e.g., failure to monitor a child; inappropriate discipline, etc). Below is an example of clinically significant poor parenting that does not fit criteria for child emotional maltreatment (from Smith Slep, personal communication, January 29, 2010):
Father is tired of his early teenage daughter's excessive focus, in his opinion, on appearance and attracting boys. One evening, to set a limit, he sits her down in the kitchen and cuts her long hair to shoulder length. The daughter is very upset and reports depressive feelings and difficulty concentrating and concerns about social rejection. [Despite the apparent impact, the act does not meet criteria for emotional abuse; however it could be seen as over-involvement and power assertive discipline, so it meets criteria for poor parenting].
In contrast, CEM criteria A (verbal or symbolic act or acts) involve one or more of the forms of emotional maltreatment described in the literature, such as berating, threatening, coercing a child, etc. (Appendix 4). In this manner, an investigator can record the act(s) reported, observed, or described in the clinical interview and determine if they qualify for CEM or poor parenting (as well as physical abuse or another form of maltreatment). To meet full criteria for poor parenting or for CEM, there would have to be some evidence of significant impact on the child, such as fear, distress, or the reasonable potential for significant disruption or disturbance in the child's development. Two examples of how these criteria are applied are shown below (from Slep & Heyman, 2006; p. 221):
Several benefits and improvements have emerged from a 41-site field dissemination of this categorical approach to CEM (Heyman, Collins, Smith Slep, & Knickerbocker, in press). First, very high reliability for CEM (90% overall agreement between "experts" and field investigators as to the determination of CEM or not), and very strong sensitivity and specificity was found when using this approach.Footnote 3 Follow-up questionnaires with social workers involved in the field trial indicated that these specific criteria were reasonably easy to learn and apply, and felt the process was fair to alleged offenders and victims. Remarkably, rates of one-year substantiated re-offense among offenders whose initial case was substantiated were cut in half (from 14% to 7%) under the new system, suggesting that clear criteria, coupled with perceived fairness of the system and formally being judged to have crossed the line into maltreatment, may have a preventative effect (Snarr, Heyman, Slep, & Malik, 2009).