716214

research-article2017
CCP0010.1177/1359104517716214Clinical Child Psychology and PsychiatryCrittenden

Test of Time
Clinical Child Psychology
and Psychiatry
Gifts from Mary Ainsworth 2017, Vol. 22(3) 436­–442
© The Author(s) 2017
and John Bowlby Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1359104517716214
https://doi.org/10.1177/1359104517716214
journals.sagepub.com/home/ccp

Patricia M Crittenden
Family Relations Institute,
Reggio Emilia, Italy & Miami, FL USA

Abstract
Attachment theory has developed over many decades - and continues to develop. Its roots lie
in several seminal publications of John Bowlby (the basis of attachment theory) and Mary D. S.
Ainsworth (the notion of individual differences in attachment). This paper identifies the prescient
contributions of these early publications and two processes (a long-term dialogue and reflection
on discrepancy) that underlay emergent theory. Because I was a student of Ainsworth when both
attachment theory and individual differences in attachment organization were becoming better
known, I offer some of my recollections of that period, suggesting how that period may have
affected current work in attachment.

Keywords
Ainsworth, Bowlby, attachment, DMM, disorganized attachment

I first met Mary Ainsworth in an awkward telephone conversation. It was 1977 and I had been
advised by an instructor to take one of her courses. I called to ask Ainsworth what the course
was about. ‘Attachment’. The conversation stopped. I didn’t know what attachment was, but
my teacher knew I had been doing attachment field work for 3 years. Within the year, I had
enrolled in a doctoral programme with Mary Ainsworth as my mentor, read Patterns of
Attachment (Ainsworth, Blehar, Waters, & Wall, 1978), and not only met John Bowlby, but
spent hours with him in a dark, closet-like room pouring over my videotapes of child protection
mothers and babies. The Infant CARE-Index (ICI), my master’s thesis, was taking shape. Now,
40 years later at the close of my career, I look back seeing how Ainsworth’s and Bowlby’s gifts
to me – and to all of us – have shaped my life’s work and three generations of research and clini-
cal applications.
In this Test of Time review, instead of picking one publication, I would like to pick a few from
Bowlby’s and Ainsworth’s work, specifically those that encapsulate their gifts to future generations
of clinicians and researchers.

Corresponding author:
Patricia M Crittenden, 9481 SW 147 St. Miami, FL 33176 USA.
Email: [email protected]
Crittenden 437

The roots of attachment theory

Early on, Bowlby put three crucial stakes in the ground; they supported the emerging field of
attachment studies and can give current work greater focus. In 1944 in the midst of war, Bowlby
published Forty-four Juvenile Thieves – in which he proposed that early and extended separation
of a young boy from his mother could generate risks that might culminate in criminal behaviour a
decade later. The idea was revolutionary. Adverse childhood events (now we call them ACEs)
shape development? The recollections of suffering adults are not fantasies? The paper made little
or no impact, possibly because the leap from infancy to late adolescence, without an intervening
pathway, was too much to be believed. But the stake of the reality of early dangerous events was
set in the ground. Others, myself included, would hammer out the details of the pathway, but that
would come slowly and take decades of effort.
Just 5 years later, Bowlby asserted that family processes were part of that pathway. The paper
was The Study and Reduction of Group Tension in the Family (Bowlby, 1949). Behind that unin-
spiring title lay a ground-breaking idea: children brought to child guidance clinics often signalled
problems within their families and the family’s problems needed to be addressed in order to help
the child. Although this is now widely understood within family systems work (which Bowlby also
helped to initiate; Byng-Hall, 2009), it got lost in most attachment research (but see Landini,
Crittenden, & Landi, 2016 for evidence connecting children’s dysfunction to parents’ attachment).
Instead, too often researchers treated pattern of attachment as an individual characteristic, not even
a dyadic attribute. (I note that many studies using the Strange Situation show only the infant – and
not the mother – in the video. How can one evaluate a dyadic strategy if one sees only the infant?
Ainsworth’s observers, of course, saw everything in the room.) Bowlby knew in 1949 what
Winnicott (1964/1987) said later: ‘There is no such thing as a baby, there is a baby and someone’.
The difference is that Bowlby knew who the ‘someone’ must be.
In post-war England, Bowlby observed first-hand the power of attachment figures to protect
children during dangerous circumstances. In Maternal Care and Mental Health (Bowlby, 1951),
Bowlby made clear that bombs were less dangerous to children’s psychological well-being than
separation from their parents with care given by strangers. The ‘someone’ must be the child’s
attachment figure; notably, however, current work does not limit this to mothers. This article was
the first notice of the new field of attachment studies, built on the stakes of early childhood danger,
family processes and children’s attachment to their mothers, along with uncertainty about the ben-
eficial effects of well-intentioned public policy and professional intervention. To me, these points
seem as challenging today as they were six decades ago. Do we leave children in their homes with
dangerous parents or endanger them by taking them from their parents and placing them with car-
ing strangers? What is the impact of divorce when the children are young? What about parental
violence and child sexual abuse? Parental mental illness? How does one balance dangers to physi-
cal and psychological safety?
Mary Ainsworth had met John Bowlby in London during the war. In the 1950s, she and her
friend John settled down to do the grunt work that transforms great ideas into sound theory that can
attract other researchers and potentially change the lives of children. John and Mary set about for-
mulating a theory that even now sustains lively debate, generates research internationally and
informs clinical practice.
Almost two decades passed quietly. One could be forgiven for forgetting about attachment.
Mary was in Uganda, observing ‘primitive’ African families, and then in Baltimore observing
‘modern’ American families. John was at the Tavistock Centre in London (in an office that no one
can identify today!) buried in the writing of what would become his trilogy, Attachment and Loss
(Bowlby, 1969/1983, 1973, 1980). Hardly anyone was interested in what this compassionate and
intellectually astute man had to say.
438 Clinical Child Psychology and Psychiatry 22(3)

The 1960s and 1970s saw the outcomes. In 1967, Mary published Infancy in Uganda
(Ainsworth, 1967). I choose to highlight it because of its anthropological observation of mother–
infant behaviour in its ecological context (isn’t that where reality is?) and its use of a non-western
population. (There are real people outside of the middle class in wealthy countries, aren’t there?)
Since then, attachment studies have expanded to include high risk children in wealthy countries,
but is life elsewhere ‘normal’? Mary said ‘yes’ – at the very beginning. Ainsworth humanized
attachment to include all humans right from the get-go and she did so with an anthropological field
study. The message: get out of the laboratory and go observe families where they live their lives.
In 1969, Attachment and Loss. Vol. I: Attachment (Bowlby, 1969/1983) was published. It was
the second announcement of a new field of inquiry, but silence ensued among researchers. Few
people picked up the banner. When I read Volume I, Volume III was only a plan, but my amazement
for what Bowlby had accomplished would sustain me until . . . well, it sustains me even now.
Bowlby outlined with careful detail the premises of attachment theory. More important, I think,
were his eager integration of new and challenging theories into a more comprehensive theory and
his respectful treatment of those who disagreed with him. After a half century in which psychoana-
lytic theory and behavioural theory had functioned as opposite and incompatible belief systems,
Bowlby said, ‘Let’s integrate’ and he integrated psychoanalytic theory with systemic approaches
and ethological and evolutionary theory. Indeed, let’s find the ‘truth’ of human development and
functioning and let’s do so by respecting all sources of knowledge. I loved the way Bowlby scooped
it all in and found a way to make an integrated understanding. I learned from Bowlby that one can
disagree respectfully. In a world where truth is never known with certainty and many are commit-
ted to ideas that are incompatible with one’s own, Bowlby demonstrated the process of respectful
and civil disagreement.

The roots of individual differences in patterns of attachment

While Bowlby reflected, integrated, and wrote in the Tavistock Centre, Ainsworth was gathering
data in Baltimore. Without scientific data about behaviour, attachment theory and studies were
doomed. No longer could clinical experience and reflective thought – that formed the basis of
Freud’s work – be accepted by the scientific community. Bowlby had emulated Freud, but Mary
understood the need for sound data that were submitted to statistical analysis. Patterns of
Attachment, Mary’s magnum opus, both made attachment theory and studies viable and also estab-
lished the notion of individual differences in attachment relationships. Individual differences in
patterns of attachment would generate thousands of studies over the next several decades.
As she had done in Uganda, Ainsworth had used longitudinal home observation, with home
visits every 3 weeks for the first year of life. In Baltimore, the observations were more formal and
were followed by a new standardized laboratory assessment, a ‘strange situation’ in which to
observe babies and mothers under controlled conditions. The Strange Situation, now the gold
standard of research on individual differences in attachment, was born and with it researchers who
lacked a year for home observation could summarize that important first year of life with just
21 minutes of effort.
The outcome of the Strange Situation? Patterns! Not numbers, not ratings. Patterns; it’s right
there in the title of the book. Patterns of Attachment provides a sound empirical basis for all the
work in attachment that followed. The patterns, of course, were the ABC attachment strategies, that
is, babies’ strategies for eliciting protection and comfort from their busy parents.
Whereas Bowlby defined ‘attachment’ as the universal innate propensity of humans to form
protective and comforting relationships, the hallmark of Ainsworth’s work is ‘individual differ-
ences in the quality of attachment relationships’. Bowlby’s contribution to the notion of individual
differences was to suggest using alphanumeric labels to avoid premature evaluation of the patterns.
Crittenden 439

The world took up the patterns, but forgot Bowlby’s advice. We labelled and evaluated – and pos-
sibly lost our way in the search for security.
The psychological undergirding for Ainsworth’s patterns came in 1979 when Mary returned
from meeting with Bowlby in London. She brought with her a draft of chapter 4 from Volume III:
Loss of Bowlby’s trilogy. Her excitement was palpable; we students were among the first to hear
how Bowlby was combining attachment strategies with information processing, specifically mem-
ory systems as differing sources of information (Tulving, 1979). The ideas blew me away. Within
3 years, my assessment tool, the Parents Interview, would mimic the structure of the Strange
Situation by increasing threat to the respondent over the course of the interview, but with the added
component of comparing parents’ representations in different memory systems (Crittenden, 1982).
Mary also sent the draft chapter to Mary Main and there it became the basis for the Adult Attachment
Interview . In time, further work on memory systems, combined with neurological work, would
replace Bowlby’s notion of ‘an internal working model’ with the more neurologically accurate idea
of ‘multiple dispositional representations’. But, again, the core of the idea lay there in Bowlby’s
early writing and it found expression in the work of Ainsworth’s students.
In one of her late papers, Attachments beyond Infancy (Ainsworth, 1989), Mary expanded pat-
terns of attachment beyond her infant patterns and asked questions about later ages and unsafe
conditions. She focused on the importance of maturation and development for changes in patterns
of attachment. That idea, too, has not been fully embraced by those who expect linearity of infant
patterns across the lifespan. Bowlby’s notion of pathways, that wind around obstructions, climb
over hills, join with other pathways and occasionally disappear in valleys, is needed if we are to
understand attachment across the lifespan. In Attachment beyond Infancy, Mary displayed curios-
ity and a humble understanding of what she herself accomplished. For her, it was less about prov-
ing some hypothesis or protecting some idea than about exploring development descriptively and
supporting her family of student explorers.

A way of working
I want to highlight two processes that produced this burgeoning field of study. One was a dialogue.
An on-going conversation maintained through letters and visits, over five decades and across three
continents. Mary and John wrote to each other, again and again (Landa & Duschinsky, 2013). They
talked about their discoveries, their observations and what their students and colleagues were
doing. More than anything, they tested new ideas in the safety of their attachment relationship.
Slowly their ideas shifted and transformed into an increasingly elegant and encompassing theory.
Attachment theory developed in a dialogue within an attachment relationship – and the relationship
sustained Mary and John during the decades when most of their colleagues were not supportive.
The other process was that of examining discrepancies. Repeatedly when I worked with Ainsworth,
data were analysed, a hypothesis was supported and then we pored over the error cases. Were they coded
incorrectly? (How boring!) Were they coded correctly, but with tools that failed to discern important
differences between these cases and others. Yes! This is where the nugget of new information should be
sought. Learning to find new meanings in discrepancies was a crucial part of Ainsworth’s gifts to me.
Like her letters to John Bowlby, there isn’t a publication to point to, but to omit this idea would omit the
basis for her brilliant publications. They were based on her exploration of discrepancies.

Where do we go from here?

Towards the end of his career, Bowlby grimly noted that researchers ‘know more and more about
less and less’. He was referring to all the ‘variable’ research carried out in laboratories using various
440 Clinical Child Psychology and Psychiatry 22(3)

assessments, but without knowing the ‘subjects’ of study well enough to find individualized mean-
ings in the results. (Over 30,000 such attachment studies exist according to Cassidy & Shaver,
2016.) Several people have pointed to the limitations of the research using the ABC + D model of
individual differences in attachment (Friedman & Boyle, 2008; Rutter et al., 1999; Thompson &
Raikes, 2003;Waters & Crowell, 1999). It seems we’ve lost the pathway. Maybe the way forward
calls for glancing back to the roots of the Bowlby–Ainsworth theory of attachment.
A courteous acknowledgment of alternative perspectives on individual differences would be
welcome. There are divergent ‘schools’ within current attachment theory, but researchers using the
dominant ABC + D model do not mention other perspectives. The outcome is that many third-
generation researchers seem unaware of the Dynamic-Maturational Model of Attachment and
Adaptation (DMM) (cf. Meins, 2017). Dialogues generate new ideas, whether the dialogue is
friendly or conflictual. Cross-fertilization leads to healthier species.
What does attachment work need now? New and clinically helpful ideas might come from a
shift from focusing on security to considering the familial context of adaptation. Exploration of
developmental pathways for individuals, within their families, might inform our understanding of
age-salient dangers and the robust quality of human adaptation. Bowlby wrote about clinical cases;
Ainsworth knew each of her Ganda and Baltimore families personally. Few researchers today carry
out extended home observations. Ainsworth’s work reminds clinicians to move beyond tick-boxes
to think about individual differences in developmental pathways across the lifespan.
Maybe we should return to slow, open-ended observational research – because it increases the
probability of finding something you didn’t expect and, therefore, can’t be found with existing
assessments or tick-boxes. Similarly, we need more appreciation of individuals’ strategic pattern-
ing in the face of threat; this could help researchers and clinicians who use the ABC + D model to
discover how threatened children and adults adapt. I am reminded that Ainsworth thought that she
herself used a Type A strategy (that, ironically, she initially considered ‘secure’ in her Baltimore
study – until she analysed the home observation data from the preceding year. Possibly Type A was
an adaptive strategy for a single woman who travelled widely?)
My points? All strategies are useful solutions to problems that children face. We each consider
our own strategy to be desirable. Case studies and family functional formulations can test these
ideas and inform group comparisons. Many of the case studies that are published use the DMM.
Possibly similar close examination of dissonant cases classified as ‘disorganized’, i.e., cases that
do not fit hypotheses, could lead to new and more differentiated understanding. As Crittenden and
Ainsworth (1989, pp. 442–443) said, the appearance of disorganization is likely to be transitory as
developing children generate more contextually attuned protective strategies. In the DMM, we
would now call this process ‘reorganization’.
Scientific technology has advanced greatly since Ainsworth carried out her observations (with-
out videotape) and Bowlby wrote his information processing chapter (without neuro-imaging).
New tools should lead to more precise observations and constructs; for example, semantic and
episodic memory were expanded to include imaged (perceptual) memory (Schacter & Tulving,
1994) and then connotative language and various somatic memory systems. Neither Bowlby nor
Ainsworth showed a penchant for hanging onto old ideas when better ones became available.
Finally, their work provides clear evidence that many problems begin very early in life and are
interpersonal processes from the beginning. Theirs is a model of adaptive competence, not of,
disorganization.
In sum, I think the message from Bowlby and Ainsworth is that we humans, from the first
moments of life and in all contexts, have the capacity to organize around the threats that we experi-
ence so as to survive and thrive. To do this, we construct context-adapted self-protective strategies
for use with our attachment figures. Clinicians need to support this natural tendency, keeping it alive
Crittenden 441

and, by functioning as transitional attachment figures, freeing it from past constraints. We do this
most effectively if we understand each person’s life context, know how their interpersonal dialectic
shaped adaptation, use that process ourselves to permit people who suffer from the limitations of
their past adaption to engage in a change process, and have valid assessments that assess the crucial
constructs. Those assessments should be based on interpersonal patterns of protective behaviour, not
solely on symptoms or checklists. Most important, we should apply these ideas in a transitional
attachment relationship with our patients or clients; with them, we could practice what we preach.
The result might save time, money and patients’ trust. Patients come to us with their suffering and
their hope. Attachment theory’s basis in adaptation to difficult circumstances can enable clinicians
to find and build on that hope. These, I think, are ideas that can meet the test of time.
I knew Ainsworth better than Bowlby; she was my mentor and colleague for the last two dec-
ades of her working life. To the end, she was generous and demanding. It’s a combination that
produced an exceptionally productive group of students: Mary Blehar, Inge Bretherton, Jude
Cassidy, Virginia Colin, Roger Kobak, Mary Main, Robert Marvin, Sally Wall, Barbara Wittig and
me. She rarely co-authored her students’ studies, preferring to highlight their independent effort,
but if my experience of her input on the ICI and the Preschool Assessment of Attachment (PAA) is
typical, she more than earned co-authorship on a wide body of the attachment research that fol-
lowed her own work. Her influence is felt clearly in the second and even third generations of
attachment researchers’ work. Indeed, Ainsworth defined the topic for the next three to four dec-
ades: individual differences in attachment relationships.
At heart, Mary wasn’t a theorist; instead, she was a psychological anthropologist; she wanted to
report on what was there – in an organized manner. Mary made observation and description respect-
able. Her greatest contribution, however, was not what she discovered, but rather what she knew
about how to make discoveries, that is, open-minded curiousity about discrepancies. In this, her
original training as a clinician shined through. Clinicians don’t know the answers; they know how
to help troubled people to find the answers. I’d like to think the DMM has made good use of both
Bowlby’s gifts of attachment theory and respectful competition between ideas and also Ainsworth’s
gifts of individual differences honed through careful observation of family life and respect for
discrepancy as a source of new information. I’d like to think the DMM points the way to improved
clinical work – as Bowlby had hoped (Bowlby, 1988). But that is for time to tell.

Acknowledgements
I would like to thank Robbie Duschinsky for his feedback on the history of attachment and the clarity of this
article. His knowledge of this history helps to support a dialogue between past and present attachment research
which is exceptionally productive.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publi-
cation of this article.

Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.

References
Ainsworth, M. D. S. (1967). Infancy in Uganda: Infant care and the growth of love. Baltimore, MD: The
Johns Hopkins University Press.
Ainsworth, M. S. (1989). Attachments beyond infancy. American psychologist, 44(4), 709.
442 Clinical Child Psychology and Psychiatry 22(3)

Ainsworth, M., Blehar, M., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of
the strange situation. Hillsdale, NJ: Erlbaum.
Bowlby, J. (1944). Forty-four juvenile thieves: Their characters and home-life. International Journal of
Psychoanalysis, 25, 107–128.
Bowlby, J. (1949). The study and reduction of group tensions in the family. Human Relations, 2, 123–128.
Bowlby, J. (1951). Maternal care and mental health. World Health Organization Monograph (Serial no. 2).
Geneva, Switzerland: World Health Organization.
Bowlby, J. (1973). Attachment and loss. Volume 2: Separation. New York, NY: Basic Books.
Bowlby, J. (1980). Attachment and loss. Volume 3: Loss, sadness and depression. New York, NY: Basic
Books.
Bowlby, J. (1983). Attachment and loss. Volume 1: Attachment. New York, NY: Basic Books (Original work
published 1969)
Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London, England: Routledge.
Byng-Hall, J. (2009). Foreword. In R. Dallos, & A. Vetere (Eds.), Systemic therapy and attachment narra-
tives: Applications in a range of clinical settings. London, England: Routledge.
Cassidy, J., & Shaver, P. R. (2016). Preface. In J. Cassidy, & P. R. Shaver (Eds.), Handbook of attachment
(3rd ed., p. x). New York, NY: Guilford.
Crittenden, P. M. (1982). Parents’ interview. Charlottesville: University of Virginia, (Available from the
author).
Crittenden, P. M., & Ainsworth, M. D. S. (1989). Child maltreatment and attachment theory. In D. Cicchetti,
& V. Carlson (Eds.), Handbook of child maltreatment (pp. 432–463). New York, NY: Cambridge
University Press.
Friedman, S. L., & Boyle, D. E. (2008). Attachment in US children experiencing nonmaternal care in the early
1990s. Attachment & Human Development, 10, 225–261.
Landa, S., & Duschinsky, R. (2013). Letters from Ainsworth: Contesting the ‘Organization’ of attachment.
Journal of the Canadian Academy of Child & Adolescent Psychiatry, 22, 172–177.
Landini, A., Crittenden, P., & Landi, G. (2016). The parents of child psychiatric patients. Annals of Psychiatry
& Mental Health, 4, 1087.
Meins, E. (2017, January). The predictive power of attachment. The Psychologist, pp. 20–24.
Rutter, M., Andersen-Wood, L., Beckett, C., Bredenkamp, D., Castle, J., Groothues, C., . . .O’Connor, T. G.
(1999). Quasi-autistic patterns following severe early global privation. Journal of Child Psychology and
Psychiatry, 40, 537–549.
Schacter, D. L.,, & Tulving, E. (Eds.). (1994). Memory systems. Cambridge: The MIT Press.
Thompson, R. A., & Raikes, H. A. (2003). Toward the next quarter-century: Conceptual and methodological
challenges for attachment theory. Development and Psychopathology, 15, 691–718.
Tulving, E. (1979). Memory research: What kind of progress? In L. G. Nilsson (Ed.), Perspectives on mem-
ory research: Essays in honor of Uppsala University’s 500th anniversary (pp. 19–34). Hillsdale, NJ:
Erlbaum.
Waters, E., & Crowell, J.A. (1999). Atypical attachment in atypical circumstances. Monographs of the Society
for Research in Child Development 64(3): 213–220.
Winnicott, D. W. (1987). The child, the family, and the outside world. Reading, MA: Perseus Publishing
(Original work published 1964)

Author biography
Patricia Crittenden completed her doctorate with Mary Ainsworth as her mentor. Her Master’s thesis, that
created the CARE-Index assessment, was developed with support from John Bowlby. After graduation, she
collaborated with Ainsworth on the development of the Preschool Assessment of Attachment (PAA) and the
expansion of the DMM model of individual differences in attachment strategy into the school years.