Wednesday, August 15, 2012

Parsing the "Attachment Cycle": The Fox Terrier of Attachment Therapy

        Parsing the “Attachment Cycle”: The Fox Terrier of Attachment Therapy
                                   Jean Mercer

 Practitioners of Attachment Therapy often reference the development of attachment according to a pattern they call the “attachment cycle”. This paper examines the possible sources of this concept, notes that the “attachment cycle” is not congruent with current conventional views of attachment but seems to be derived from a mixture of unconventional and older conventional theories, and suggests that commonly observable aspects of early development contradict the “attachment cycle”. The potential dangers of belief in the posited “attachment cycle” are discussed.   

According to Steven Jay Gould, when Darwin was describing the “dawn horse” Eohippus he described it as the size of a fox terrier. Of course, in his day, most people likely to read his work rode to hounds and knew what a fox terrier (as opposed to a fox hound) was. They knew what size Darwin meant. But, as time went on and fewer and fewer people had the resources or the beliefs that would encourage fox hunting, fox terriers became less common, and relatively few people knew how big they were and how big Darwin meant that Eohippus was. Authors discussing Eohippus nevertheless copied Darwin and each other, and continued to compare the ancient animal to a fox terrier, communicating little to their readers. The same thing can and does happen as authors writing about  mental health issues copy each other--  and it happens most easily when it’s a diagram that’s copied.

When an author provides a simple, decorative diagram that illustrates a claimed connection between events, it’s hard for readers not to assume that the existence of the diagram confirms the existence of the events and of connections between them. Maslow’s all-too-famous hierarchy of needs exemplifies this phenomenon—thousands of people can recognize and even reproduce the pyramid diagram, even though they may have no clear idea of the claims being made or the presence or absence of empirical support for them.
In discussions of vernacular, unconventional psychological principles--  for example, the belief system behind “Attachment Therapy” (Holding Therapy, Rage-reduction Therapy, Z-therapy, and other names as well)—proponents often employ a diagram showing events in an “attachment cycle”. This diagram can be seen in the works of Foster Cline and of Vera Fahlberg, and currently at, at, at, and so on for pages of Google. Diagrams of the “attachment cycle” show repeated incidents in which a baby feels a need, a caregiver responds appropriately, the baby is gratified--  and after many repetitions the baby becomes attached to the caregiver. A disturbed “attachment cycle” is represented like this:

                                                  Baby has a need
                              (possibly including smiles and social reinforcement)
                               /                                                                \                             /
Trust does not develop, rage develops instead                     Baby cries
                              \                                                             /
                                 \                                                        /
                                         Caregiver does not respond
For some authors, the “attachment cycle” ends here. Meet the baby’s needs and the baby becomes attached (defined here as developing trust)--  fail to meet them, and he or she does not become attached (defined here as responding with rage). For Cline and some others, however, this diagram shows only a “first year attachment cycle”. Attachment is not complete, they say,  until the child learns to accept and indeed love the limitations and boundaries created by a powerful caregiver in the course of a “second year attachment cycle”. We can consider the first-year cycle by itself and then return to examine the notional second-year cycle.
It’s clear that these repeated sequences, posited by Cline and Fahlberg as part of a first-year attachment cycle, do occur. Babies cry when they need something, and they are either helped or not helped by their caregivers. Toddlers  also certainly test boundaries, and are either corrected or not corrected by their caregivers. Because babies cry when they actually do need something, a baby who is frequently ignored or treated inappropriately may not survive. In addition, to look forward to the second year, a toddler whose boundary-testing is not responded to appropriately may be injured or killed as a result of risk-taking--  these possibilities are plain. What’s not so clear is whether either of these scenarios has anything to do with attachment in any direct manner--  yet Cline, Fahlberg, and various followers have insisted that the demonstrably repeated sequences are the causes of attachment. 
Like the Maslow diagram, this depiction of a “cycle” seems intended to operate as its own proof. Many  readers who see it seem to be convinced that that where there’s a diagram, there’s reality. Cline and his colleagues have made no effort to adduce data to support their claims, but they have attributed some parts of  the “cycle” idea to other authors. Fahlberg (1991),    for instance,  references Rene’ Spitz’s book The first year of life (1965),  but she also shows diagrams of a related “arousal-relaxation cycle” whose source is unmentioned.
                             The “First-Year Attachment Cycle”
Is Spitz’s Work the Source of the “Attachment Cycle”?
What are the actual sources of the “attachment cycle” idea? Fahlberg’s reference to Spitz suggests that his work is a good place to begin the search. Like many other authors, Spitz was interested in the long-term effects of the infant’s repeated experiences with adequate or inadequate caregiving. As he pointed out with respect to feeding, “the two parts of the experience, the hunger screaming and the gratification which follows it, become linked in the child’s memory. … This development should be understood in the terms of Ferenczi’s… propositions on the stage of infantile omnipotence. Hunger screaming, followed by gratification, forms the basis for the feeling of omnipotence, which according to Ferenczi is an early stage of the sense of reality… In this achievement of enlisting the mother’s help for his needs through screaming, the human being experiences for the first time the post hoc ergo propter hoc in connection with his own action” (Spitz, 1965, p.153). But Spitz associated these sequences of experience with memory, with perception, and with the understanding of causality. The word “attachment” did not appear in the index of the book at all. Spitz’s work does not seem to be the source of the “attachment cycle”.
How Does Erikson’s Concept of Trust Relate to the “Attachment Cycle”?
What about other repeated events or cycles as proposed by influential authors? Fahlberg, Cline, and other authors occasionally refer to Erik Erikson’s concept of basic trust (Erikson, 1950/1963). A sense of trust is thought to develop in the course of repeated experiences of good care, but trust is not exactly the same as attachment. Erikson pointed out the need for an appropriate balance of trust and mistrust, but conventional attachment theory emphasizes the advantages of secure attachment over other attachment statuses, both with respect to later social relationships and with respect to support for childhood exploration and learning. Erikson speaks of a life cycle, of course, but he does not refer to a cycle of repeated experiences when he says this--  instead, the life cycle consists of a number of qualitatively different stages of social and emotional life, occurring in a predictable order.
A Behavioristic Approach
Neither Cline nor Fahlberg gives specific mention of the possibility that operant conditioning could play a role in attachment. To think along these lines involves examining the possible process of reinforcement for attentiveness to the caregiver--  and, in addition, for the caregiver’s attentiveness to the baby. J.L. Gewirtz (1969) proposed a theory of mutual effects of parent and child on each other, with social reinforcement for the child from that care and affectionate attention of the adult, and for the adult from the pleasure shown by the infant. In a series of systematic observations of mothers and babies (a strategy all too rare among theorists addressing this topic), Gewirtz was able to show mutual reinforcement and gradual change of mother’s and child’s behavior toward each other. In fact, the “first-year attachment cycle” is quite parallel to Gewirtz’s suggestion, as both involve spontaneous behavior related to infant needs, adult responses, and ensuing learning. However, like all operant conditioning approaches, Gewirtz’s theory  would allow for attachment behavior to be maintained after it is once established, even though the adult response became less and less frequent; the “attachment cycle” theory suggests instead that failure to respond produces rage and interferes with the development of emotional attachment.
Can the “Attachment Cycle” Be Recognized Under Some Other Name?
Fahlberg herself provides an additional  diagram that may provide a clue to the sources of the “attachment cycle”. This shows the “arousal-relaxation cycle” mentioned earlier. It involves arousal of energy and attention by a physical need, a reaction to that need (like crying), followed by either appropriate care and subsequent relaxation when the need is gratified or by continued distress and a failure to learn that caregivers will help. This cycle resembles, in name and otherwise, Wilhelm Reich’s “four-beat” motivational cycle. Reich ( 1980; originally published 1945) posited that in all motivation there is some form of mechanical tension, followed by an increased electrical charge, an electrical discharge, and mechanical relaxation. This cycle, Reich thought, had a biological foundation and could be seen in events ranging from orgasm to mitosis.  Reich, who died in prison after conviction for selling fraudulent medical devices,  believed that transfer of an unknown energy called “orgone” was at work in all these phenomena. I would suggest that Reich’s motivational cycle was the source of the “ first-year attachment cycle” presented by Fahlberg and by Cline, as well as by many imitators (e.g., Golding, 2008 ).
Neither Fahlberg nor Cline attributed the arousal-relaxation cycle or the “attachment cycle” they derived from it to Reich. This is not surprising, because authors dependent on unconventional beliefs often fail to cite their sources. It’s notable, though, that Robert Zaslow, Cline’s mentor (and perhaps Fahlberg’s also?) frequently referred to Reich’s theories as sources (Zaslow & Menta, 1975) and proposed a “soul cycle” that is comparable. Because the “attachment cycle” is not to be found in any other discussions of early emotional development, it may well be that this “fox terrier” has been brought back again and again without sourcing, to the point that few readers know what it is or where it came from.
                                        The “Second-Year Attachment Cycle”
Fahlberg had little or nothing to say about a “second-year attachment cycle”, but Cline
 ( 1992) emphasized the importance of this period, during which, he claimed, caregiver limit-setting was essential to the further development of attachment. He backed up this statement by a reference to Bowlby’s remark that caregivers who become attachment figures are usually stronger and more authoritative than the children, an idea that Cline tweaked into the claim that the adult’s strength and power are actually the causes of aspects of attachment.

It’s readily observable that parents all over the world do set limits on their young children’s actions, usually starting at about the end of the first year, when expert crawling and the beginning of independent walking make “mischief” more possible. Toddlers are pressed to become toilet-trained, to use spoken words rather than screaming or grabbing for what they want, and to stay away from dangerous or breakable things. There may be poor outcomes of behavioral development both for children who receive little limit-setting and for those subjected to many rules and much punishment for infractions. Careful guidance during the toddler period helps establish self-regulation and self-control.
Does limit-setting during this period have anything to do with attachment? Evaluations of attachment like the Strange Situation (Ainsworth, Blehar, Waters, & Wall, 1978) focus on the beginning of the toddler period at about 12 months of age, before much limit-setting is usually in place. This suggests that attachment in Bowlby’s sense is already established before the second year is well underway. Bowlby himself referred to this period of development as one in which parent-child relationships began to modulate from the intensity of earlier stranger and separation anxiety to include negotiation and compromise between the parties--  a situation that seems to contradict Cline’s claim.
Cline, Fahlberg, and others have also appealed to the basic trust/ basic mistrust concept proposed by Erikson as the important developmental issue of the first year of life. But what does Erikson suggest about the second year? He proposed that the second year or so of life involves a new focus on a sense of autonomy, or confidence in one’s own decision-making about how to do things, as opposed to a sense of shame and doubt that emphasizes other people’s opinions and worth. Recognizing that toddlers need to be socialized and learn to follow family rules, Erikson nevertheless suggested that the task of socialization can be done through consistent gentle guidance, helping the individual to have confidence in his ability to “do things right”, or as an alternative can overwhelm the toddler’s abilities and leave a readiness to feel himself in the wrong. Cline’s stress on the power and authority of the caregiver would seem to work against a sense of autonomy, which Erikson argued children needed to develop even though adults are, realistically, stronger and more knowledgeable than toddlers are.
Like other object-relations theorists, Spitz considered frustration to be a factor in the infant’s progress toward understanding reality. Limit-setting, even of the mildest form, is frustrating to young children--  so does this mean that Spitz’s views support the Cline “second-year attachment cycle” belief? No, because (again like other object-relations theorists), Spitz believed that frustration played an important role in development during the first year, when it would not often be associated with limit-setting.
There seems to be no support from theorists of personality development for the idea that limit-setting  in the toddler period contributes to the growth of attachment per se,  although it is clear that experiences of limit-setting influence other aspects of personality and behavior. Bowlby (1944) considered that some juvenile delinquents were influenced by poor attachment histories, but did not connect obedience specifically with attachment. Why, then, did Cline attempt a rapprochement between limit-setting and attachment? He has not explained this, but we can speculate that the goal is to complete Cline’s posited association between attachment and obedience. Cline and various colleagues have claimed repeatedly that problems of attachment are indicated by the child’s disobedience and undesirable behavior, and that treatment by means of  Attachment Therapy will make the child “respectful, responsible, and fun to be around” and ensure that he or she does things “Mom and Dad’s way”. Cline’s commercial parent-education program, “Love & Logic”, is primarily concerned with obedience and compliance with adult wishes. The concern with obedience, combined with the assumption that all behavioral problems ultimately result from attachment issues, appears to have brought compliance under the attachment umbrella and thus created the “second-year attachment cycle” belief.
                           How Can We Know What Causes Attachment?
Attachment behavior is readily observed, but attachment itself (as an internal state and motivating factor) must be inferred. As a result, no one can see attachment happening—although Gewirtz (1969) reported what appears to be a relevant set of processes. Until attachment has happened and attachment behavior is present, we can only guess at the internal process. This means that the problem of associating later attachment with one or more causal factors is a serious one.
Various causes of attachment have been suggested, some more and some less likely. Authors like Nancy Verrier ( 1993) propose that babies are already emotionally attached to their mothers at birth, but infants’ behavior does not support this view. Sigmund Freud attributed attachment to the experience of being fed by a particular person. John Bowlby emphasized the effect of pleasurable social and emotional interaction with a caregiver, and proposed that infants in the second year of life are biologically primed to learn a connection with a familiar adult.  As has been noted already, Cline and other Attachment Therapists hold that repeated experiences of satisfied needs create attachment to the adult who provided the satisfaction.

Almost all attachment phenomena share a particular problem that stands in the way of accurate analysis. This is the fact that most often, the person (or people) who plays with and shows affection to a baby is the same person (or people) who feeds, cleans, rocks, and comforts that child.  The confounding of these factors means that in the ordinary course of events we are not going to be able to separate the effects of feeding, of repeated satisfying cycles, or of social interaction.
Nevertheless, some observations do allow us to think about causes of attachment. For example, in many families, one parent is much more involved than the other in infant care. We would expect that parent to be the only important attachment figure, if the “first-year attachment cycle” applies; yet we see babies display great interest in a parent who does little physical care but who when present is playful and socially engaging. Sadly, we also see that children form attachments to caregivers who are neglectful and even abusive, and grieve when separated from the adults, suggesting that satisfying care is not very important to attachment, or that if it is important, very little of it is required to do the job. Finally, reports on the traditional kibbutz child-rearing methods (see Bettelheim,1969) suggest that infants cared for with little interaction with adults, but physically healthy, become engaged with and attached to nearby “crib-mates” and are distressed if those babies are separated from them for some reason; other infants cannot have participated in feeding or care, but may have been either socially interactive or simply familiar in the sense that they were almost always there. These observations imply that repeated experiences of the “attachment cycle” are not actually causes of attachment.

Several other developmental phenomena appear to argue against the “attachment cycle” as an explanation of emotional attachment. The first of these is the  prefiguring of attachment behavior by a period of infant “wariness”, in which the 6- or 7-month-old, who used to be highly sociable, begins to regard new people with a serious air and warms up slowly to those who behave pleasantly and engagingly. At about the same time, many babies begin to show fear of sounds and events that are familiar but perhaps startling, like the sound of the garbage truck or a vacuum cleaner being turned on. This change occurs before and predicts the obviously attachment-related behaviors of stranger and separation anxiety, but it is difficult to see how a gradual development by way of a repeated “attachment cycle” could be connected with the development of fearfulness and its mirror image, a strong preference for the familiar.
A second relevant event is the often dramatic sudden emergence of attachment behaviors at about 8 to 12 months of age. This developmental milestone would seem to indicate a reorganization of behavior and emotion of the kind posited by dynamic systems theory. In Bowlby’s attachment theory, the abrupt change (sometimes evident over a day or two) involves a combination of maturation and of experience rather than repeated experience alone. Simple repetition and reinforcement might be expected to show a conventional learning curve, but not the sudden increase in a behavior’s frequency shown in observations of infant and toddler behavior.
An additional phenomenon that appears incongruent with the “attachment cycle” theory is that of transitional objects. The toddler’s attachment to objects such as “blankies” and teddy bears has been described by Winnicott (1953) and is familiar to anyone who has cared for very young human beings. The display of attachment to the transitional object is parallel to attachment to caregivers, and is notably not a substitute for attachment to a familiar person. The toddler who is devoted to a transitional object wants the specific object as well as the familiar caregiver and may not be able to be calmed by the caregiver alone. Yet the transitional object cannot have participated in an “attachment cycle”, can provide little in the way of need satisfaction other than a sense of familiarity, and certainly does not have as many ways of satisfying needs as a human caregiver can provide. In addition, transitional objects do not appear to function until after attachment to caregivers begins to be displayed, and although the object may have been offered to the baby prior to attachment, it has usually been one of many, all providing equal experience, but only one eventually being selected as the needed blanket or bear. The transitional object becomes important only after it has become familiar and the baby has reached the necessary maturational stage, and the same may well be true of the human attachment figure.
Finally, the toddler characteristic of neophobia may be relevant to behaviors we consider attachment-related. Neophobia, the young child’s aversion to unfamiliar foods, objects, and situations, can be considered as the mirror image of the child’s intense preferences for familiar people and places as well as for a specific transitional object. The child’s fear of strangers and of separation can also be classed as forms of neophobia. The “attachment cycle” theory would seem to imply that rejection of a person depends on experience of that person’s failure to satisfy signaled needs, just as it suggests that attachment/preference results from experiences of satisfaction mediated by an adult. Yet unfamiliar foods, places, and people have by definition had no opportunity either to satisfy or to fail to satisfy infant needs. Examining attachment behavior in the context of other age-related behavior in this way leads to the conclusion that an “attachment cycle” would have to be congruent with several aspects of infant and toddler development in order to be a reasonable explanation of attachment--  but it is not.
                        Is It a Problem to Assume There is an “Attachment Cycle”?
Does it matter whether people think there is a “first-year attachment cycle”, with or without an added second-year cycle? In some ways, no, it does not matter--  and the belief may even have a positive outcome. If parents are concerned about attachment as an important aspect of development, and if they believe that they can foster attachment by sensitive, responsive treatment of their infant, those beliefs are likely to produce excellent child care and good development. It’s a much better situation than if, say, someone advised parents to whip an infant who refused food (see Pearl & Pearl, 1994).
In other ways, though, the belief in the “attachment cycle” can be quite harmful. One of the associated problems is the assumption fostered by Attachment Therapists that child disobedience indicates a failure of attachment, rather than a variety of other causes, many of which involve poor parenting. A second, and even more potentially harmful, belief is that both first and second-year “attachment cycles” can be recapitulated as treatment for older children. Acting out of this assumption involves efforts to display adult power and authority as well as arbitrary presentation of sweets, hugs, and eye contact as means of creating attachment (Thomas, 2000).  The second belief is based on the assumption that it is possible, by re-enacting some common early childhood experiences, to cause individuals to “regress” and resume the characteristics of their early lives, then to cause them to develop back to their present chronological age’s characteristics and to undergo experiences that will correct whatever problems had initially occurred. These principles and practices are without  empirical support and are potentially harmful, with a record of demonstrated child deaths and injuries (Mercer, Sarner, & Rosa,2003).
The “attachment cycle” theory proposed by Fahlberg, Cline, and other authors is derived from a salmagundi of other theories, in most cases uncited by authors who have discussed the concept. The theory, as repeatedly diagrammed by advocates of Attachment Therapy, is incongruent with Bowlby’s attachment theory and only partially congruent with some other suggestions about attachment. In addition, it fails to predict some easily observable phenomena of late infancy and the toddler period whose association with attachment is evident. Like Darwin’s fox terrier, the “attachment cycle” material appears to have been repeated unquestioningly by authors and Internet sites who have accepted it because they have seen it before--   a sort of “attachment cycle cycle”. This repetition, and the harmful implications of the idea for parents and practitioners, are the only reasons for a serious analysis of the “attachment cycle” like the one given here. Were it not for the potential harm connected with the belief system, there would be no point in examining it, as it is not and never has been a part of any conventional approach to the understanding of attachment.

Ainsworth, M.D.S., Blehar, M., Waters, E., & Wall, S. (1978). Patterns of attachment. Hillsdale, NJ: Erlbaum.
Bettelheim, B. (1969). The children of the dream.New York: Macmillan.
Bowlby, J. (1944). Fourty-four juvenile thieves: Their characters and home life. International Journal of Psychoanalysis, 25, 19-52, 107-127.
Bowlby, J. (1982). Attachment. New York: Basic.
Cline, F. (1992). Hope for high risk and rage filled children.  Evergreen, CO: EC Publications.
Erikson, E.H. (1950/1963). Childhood and society. New York: W.W. Norton.
Fahlberg, V. (1991). A child’s journey through placement. Indianapolis: Perspectives Press.
Gewirtz, J.L. (1969). Potency of a social reinforce as a function of satiation and recovery. Developmental Psychology, 1, 2-13.
Golding, K. (2008). Nurturing attachments. London: Jessica Kingsely.
Mercer, J., Sarner, L., & Rosa, L. (2003). Attachment Therapy on trial. Westport, CT: Praeger.
Pearl, M., & Pearl, D.(1994). To train up a child. Pleasantville, TN: No Greater Joy Ministries.
Reich, W. (1945). Character analysis. New York: Farrar, Strauss, Giroux.
Spitz, R. (1965). The first year of life. New York: International Universities Press.
Thomas, N. (2000). Parenting children with attachment disorders. In T.Levy (Ed.), Handbook of attachment interventions (pp. 67-111). San Diego: Academic.
Verrier, N. (1993). The primal wound. Lafayette, CA: Author.
Winnicott, D. (1953). Transitional objects and transitional phenomena. In D. Winnicott (Ed.), Collected papers: Through pediatrics to psychoanalysis. Middlesex, UK: Penguin.
Zaslow , R.W., & Menta, M. (1975). The psychology of the Z-process. San Jose, CA: San Jose State University Press.


  1. nice work
    I did not read Fahlbergs book however she introduced the cycle during a lecture in university.
    My understanding was a bit different from yours. Especially in the regard that that Fahlberg did not present the cycle as a attachment explaining theory but rather as one of many processes.
    I would agree with you that this cycle does not form attachment. However it does contribiute to the development of trust and consistence (or the opposite) which is associated with secure attachment.

    I could expand that further if interest exists but I think this comment is sufficient to offer a different perspective.

    1. Certainly if a young child is not fed and cared for regularly, there will be both physical and mental consequences. However, there would be a big difference between attachment in a family who neglected a child because of addiction or psychopathology, and attachment when refugee status or severe poverty meant that no one got much care, although the adults were attentive and affectionate toward the child.

      My concern about the focus on this "cycle" as the basis of attachment is that "attachment therapists" believe that they can create attachment by repeating earlier experiences-- so children are not allowed to help themselves to food, or even babies to hold their own bottles, all in the name of creating attachment.

    2. I don't like the prase "create attachment" ...
      In my opinion babies "create attachment" to somebody and caregiver (whoever they might be) create attachment towards the child.
      What I want to point out is that caregiver can not create attachments in another person. One might create opertunities to do so.

      The second point is that attachment is already there... It is not a secure attachment (otherwise we wouldn't aim to "create attachment")

      To your example: One way of looking at it is that the baby had a lack of certain learning experiences, hence couldn't form those connections in the brain. By giving the baby the opertunity that someone cares about them. It is not about getting basic needs met, (if those were not met the baby would not be alive) it is more about care and trust.

      There are of course different ways of doing that.
      (I hope) caregiver finds a way that suits the individual they are working with: Depending of age and development it might be appropiate to hold the bottle OR it might be more appropiate to praise the child on how well it can do that on his/her own already.

      In the place I work we also don't allow children to help themselfes to food. However we have children with a variety of special needs and it is difficult to compare that to a children's home for instance.
      However we do it not for attachment reasons, but to establish boundries, support social meals and consistency (if everybody helps themselves to food when they feel like it we don't have meals together)

      Also one needs to be dependent to be independent:
      In short that means if we let them help themselves to food they might do the same in a shop. (again I don't work in a normal children's home)

      However if those things happen only for attachment reasons I would be as concearned as you are!!!

      one needs to be very careful not to force this experiences as that would just create more anxiety rather than attachment.