Thank, Brian Allen, for Saying "Down With Attachment Disorders"!

A leading clinical child psychologist, Dr. Brian Allen of the Center for the Protection of Children, Penn State Hershey Children’s Hospital, has stood up to make an important statement. His recent article, still in early online form in the new journal Evidence-based Practice in Child and Adolescent Mental Health, is entitled “A RADical idea: A call to eliminate ‘attachment disorder’ and ‘attachment therapy’ from the clinical lexicon”.

Noting that those two terms are well entrenched in graduate education, parenting information, adoption work, and so on, Allen points out that “neither of these two concepts is empirically sound as commonly practiced”. In other words, the assumptions frequently made about the nature of attachment, even by well-trained and legitimate practitioners, are often unsupported by systematic evidence.  

One source of confusion about “attachment” as a clinical concept is that  attachment behavior and attachment theory are two very different things. Attachment behavior is observable and happens in the great majority of children between about 9 months and about three years of age. It involves responding to threats or discomfort by getting and staying close to familiar caregivers. Although most children do this, the manner and intensity of their behavior depends in part on individual  differences and on the caregiver’s response. Children who have experienced many changes of caregiver, or who have had insensitive or unresponsive  caregivers, will still show some attachment behavior in most cases, but (as Mary Dozier has shown) their behavior may be so subtle and understated that caregivers have trouble noticing it--  for example, they may just glance up at a caregiver who leaves the room, rather than bursting into tears and following him.

Attachment theory, like all theories, attempts to put observable attachment behavior into a framework that makes sense and helps us predict what kind of behavior may occur later. In John Bowlby’s first formulation of attachment theory, he proposed a framework that included the development of an internal working model (IWM) of social relationship. The IWM would begin with what a baby learned in the early relationship with a caregiver, including the valuable comfort of being able to get close to the caregiver in distressing circumstances.

The theory of the IWM, and its changes as a result of maturation and experience, provided an explanatory framework for the ways attachment behavior changes with age. As Allen points out, a basic fact about attachment behavior is that it alters as a result of a variety of experiences and goes on altering for many years. Later attachment behavior is not entirely determined  by early events, and the later events that affect this behavior do not always seem to have anything to do directly with attachment (e.g., mothers’ confidence). It is a great mistake to try to cherry-pick portions of attachment theory , such as the importance of early sensitive and responsive caregiving, but to neglect the message of this theory about the long-term changing and reshaping of attachment behavior as the result of a range of experiences. Unfortunately, terms like “attachment disorder” and “attachment therapy” have become shorthand for “early caregiving effects”, and their use often implies that the user is ignoring the larger picture of attachment theory.

With respect to the term Reactive Attachment Disorder, once thought of as encompassing two separate and somewhat opposed forms, Allen points out that one of those forms is now described as Disinhibited Social Engagement Disorder  and divorced in terminology  and otherwise from the attachment concept. The remaining use of the term Reactive Attachment Disorder is applied to a very small proportion even of  children with histories of severely neglect. What’s more, this disorder appears no longer to be present after children have been in family settings for a while, so presumably these children were not locked into the effects of their earliest caregiving experiences, but like anyone else are able to continue the development of their models of social relationships on the basis of maturation and new experiences. This means that the word “attachment” in the current name of this diagnosis has little meaning--  but it particularly lacks the very narrow meaning  so often ascribed to it.

Allen proposes that we stop using terms that incorrectly attribute behavioral problems to attachment difficulties, and that we stop saying we are treating attachment when in fact we are working with specific problems of parents and children. Not all relationship problems are attachment problems, and many behavior problems are not relationship problems  in any case. When we use terms that confuse the issues both of causes and of treatments, we make ourselves more likely to make serious mistakes.

What do we do then? How do we change the familiar names that make us think, wrongly, that we are actually talking about a specific aspect of attachment? Allen suggests a new emphasis on outlining the  problems a child or family is experiencing--  for example, aggressive behavior or callous/unemotional traits--  and adding to these what we know of a history of maltreatment. Then, treat the problems with well-established treatments, and, I would say, forget the popular belief that treating the problems successfully is wrong because  it ignores the “underlying” cause.