The Adult Attachment Interview (AAI): Mary Main in a Strange Situation

Mary Main & Dan Siegel December-2010-UCLA(Updated 4-12-17) Only 55% of us had “secure attachment” as infants, according to research on 6,282 infant-parent pairs during 1977-1999. This would worry us all if we knew what it meant, because the other 45% suffer “insecure attachment.”  That means almost half of us have trouble with committed relationships. [1, 2]

This is because the attachment we had as kids continues all our lives, according to related research on over 10,500 adults during 1982-2009.  Plus, we pass on our “insecurity” to our children. [3]

These are the “quiet blockbuster” results of two extended, linked studies. First, using the “Strange Situation” procedure, Dr. Mary Ainsworth and her successor Dr. Mary Main studied infant attachment. (Dr. Main & Dr. Daniel J. Siegel, above.)

Second, Dr. Main used the Adult Attachment Interview (AAI) to study the infant’s adult parents. Results showed that almost half the adults were not securely attached, either, and this corresponded to their own infants’ attachment.

This 45% “insecure” figure starts to explain why we’ve got a 50% divorce rate. If (like me) you’ve tried dating after divorce, it won’t surprise you that science shows almost half the adults out there can’t manage a secure, committed, relationship. Ouch, you’ve experienced it.

And if 45% of us were “insecurely attached” in 1999, what’s the rate in 2017? In the almost 20 years since, we’ve become an “e-society” with email, cell phones, texting and computers further trashing our ability to relate in person. Psychotherapists interviewed for this blog said that a round number of “about 50%” insecure attachment is conservative.  Some say our insecure rate is higher.

In fact, the 2009 study of over 10,500 Adult Attachment Interviews said that secure attachment had fallen by another 17.1%, since the 1999 secure estimate of 55% above. That would mean 46% were secure in 2009 and the insecure rate was up to 54%–more than half the population.  [3]

Here’s more “about 50%” data. The Adverse Childhood Experiences (ACE) Study showed that 66% of 17,337 middle class adults had one or more types of childhood trauma, and 42% had two or more types. In less privileged populations these numbers are over 70%.  A U.S. average of all economic groups would show 50% or more suffer two or more types of ACE trauma.

That means, for example, they experienced both childhood physical and sexual abuse, or both childhood emotional abuse and neglect.

The ACE Study lists 10 such abuses, including traumas that happen to newborns (physical and emotional neglect). Such trauma puts children into “fight-flight,” a chronic state proven to shut down the organism’s capacity for feelings of attachment and love. [4]

Half of us have some degree of attachment trauma and don’t even know it.  Let’s get informed; then we can heal. If we didn’t get securely attached as kids, we can develop “earned secure attachment.”  “It’s possible to change attachment patterns,” as Dr. Main’s colleagues say. [5]

 Strange Situation  Experiment

bowlby-johnAttachment Theory isn’t new; it just gets too little air time. British psychiatrist John Bowlby (left) developed it in the 1950s while working on the post-war orphan crisis. [6]  Bowlby believed that all infants would seek to stay close to parents, since such “attachment” promotes survival. In 1952, he published a study of toddlers’ responses to separation from parents. It showed that “when toddlers were placed in unfamiliar surroundings that provided no stable caregivers, they underwent three… stages of response to separation: protest, despair, and finally detachment,” writes Dr. Main. [7]

Mary Ainsworth ca 1990Dr. Mary Ainsworth studied with Bowlby in London 1950-54, then researched his concept of “proximity-seeking behavior” in infant-mother pairs in Kampala, Uganda, published as “Infancy in Uganda” (1967).  Then she found “astonishing similarities” in Baltimore, MD pairs. [7]

Ainsworth created the Strange Situation in the early 1970s, as a science experiment at Johns Hopkins in Baltimore to document this infant behavior. “Ainsworth structured the Strange Situation to include three of Bowlby’s ‘natural clues to danger’… to arouse babies to seek proximity” to the parent, Main says. Researchers watch and video-tape through one-way glass, as infant-mother pairs react to apparent danger.  First a baby responds to a strange lab room; then to two entrances of a strange person; then to  two different separations from its mother (details in footnote 8). [8]

Babies were expected to stay close to parents as Bowlby thought.  Babies “that Ainsworth termed ‘secure,’ play and explore happily prior to separation; show signs of missing the parent during separation, such as crying and calling; seek proximity immediately upon the parent’s return; then return to play and exploration, ‘secure’ once again in the parent’s presence,” reports Main. [8]

But 30% of babies did not act secure–they avoided mom.  They showed no preference between mom and the stranger.  “While a majority of infants behaved as expected and were termed secure, to Ainsworth’s amazement six showed little or no distress at being left alone in the unfamiliar environment, then avoided and ignored the mother on her return,” Main reports (emphasis added). [7]

Ainsworth decided to categorize these babies separately as “avoidant” of mother. Now she had two types: (A) Insecure Avoidant and (B) Secure. She concluded that moms of avoidant babies didn’t respond or have the sensitivity to understand the babies’ real need, so infants felt “insecure.”  [8]

Still later Ainsworth saw that of the insecure babies, some had yet a third reaction: actually, they were “ambivalent” about mom. They were very distressed when mom left, but on her return, they alternated between avoiding and frantic clinging–plus, they never calmed down. Research showed that ambivalent attachment results from moms who are sometimes available, sometimes not, so babies learn they can’t depend on people.

So “surprisingly, Ainsworth found that infant responses to separation and reunion fell into three distinct, coherently organized patterns of attachment,” and added a third category: (C) Insecure Ambivalent, Main reports. [8, 9]

By 1977 Ainsworth had developed an “American standard distribution” for infants of “about” (A) Insecure Avoidant 20%, (B) Secure 70%; and (C) Insecure Ambivalent 10%. By 1988, Strange Situation research using Ainsworth’s three categories had been done with 2,000 infant-parent pairs in 32 studies  in 8 countries.  Some countries varied, but global results averaged the same. [9]

Scary Parents

Mary Main BerkeleyIn 1973 Mary Main became Ainsworth’s grad student at Johns Hopkins in Baltimore, working on the Strange Situation experiments from their start. After her doctorate Main moved to Berkeley, to see if Ainsworth’s Kampala and Baltimore findings would replicate. [8]  In 1977 Main did a Strange Situation study of 189 Bay Area infant-parent pairs which did replicate Ainsworth’s results. [8, 10]

But by 1979 Dr. Main was making her own discoveries—out of concern about the parents.

“In none of Ainsworth’s original observations was the possibility considered that some mothers… could also be frightening,” Main notes.

“For my dissertation at Johns Hopkins, I watched 50 children in the Strange Situation… Using Ainsworth’s three-part classification (secure, avoidant and ambivalent), I found at least five infants could not be classified…

“I had visited most of the mothers in their homes,” Main wrote, “and I knew that at least three of the five mothers of the un-classifiable infants had behaved most peculiarly with their offspring. One—frighteningly, to me—had treated her toddler as an animal.” [8]

Ainsworth was concerned, too; in fact, she’d put some babies in her Secure set only because they didn’t fit her other two sets, Main writes. [8]

By 1985, Main decided to “extend attachment theory to include the import of infant exposure to anomalous fear-arousing parental behaviors (Main & Hesse, 1990)… in conjunction with the discovery of a fourth Strange Situation attachment category, ‘insecure-disorganized/disoriented’ (or ‘D’) (Main & Solomon, 1990)…

“We proposed that fear of the parent could account for many instances of disorganized behavior, since the infant’s natural haven of safety will have simultaneously and paradoxically become the source of its alarm (Hesse & Main, 1999, 2000),” Main reports. [8 p.257; 11]

Main and Ainsworth agreed that some 15% of babies actually form a new, fourth group: (D) Insecure Disorganized. Their mothers were so frightening that the babies couldn’t develop any consistent response. These infants “exhibited a diverse array of inexplicable or overtly conflicted behaviors in the parent’s presence” including “disorganization, disorientation, and confusion,” Main wrote. This included crying loudly then suddenly freezing; ignoring the parent to rock on hands and knees; moving away; raising hand to mouth in fear; or swiping at the parent’s face. [8, 11]

This finding that 15% of average U.S. babies are so insecure they’re almost incoherent was so shocking, it was checked for years. [2, 10, 11]  But the number held. Worse, among children of American adolescent mothers the rate is over 31%, and is over 25% in many Third World nations. [12]

To remove the 15% of disorganized babies from Ainsworth’s original estimate of 70% secure, would reduce the secure set to 55%.  If so, Main would have four categories: avoidant 20%, secure 55%, ambivalent 10%, and disorganized, 15%.  [1, 11]

What was up with parents that 45% of their kids couldn’t manage secure attachment?

That was how researchers read Main’s documents for the 20 years from 1990 to 2010. [ 2, 10, 11]

That 55% Figure

I first heard this “55% secure” figure in 2010 from therapist Dr. Henry Cloud, New York Times best-selling author of “Boundaries.” He’s also the author of “Safe People,” a book whose point is that many folks we meet are able to relate, but almost as many are not. Asked for numerical odds at a lecture, he said that 55% are securely attached but 45% are insecure.  Later I asked therapists at a UCLA therapists’ conference, who also said that professionals commonly speak of 55% secure.

Checking online, I found an authoritative report in “Pediatric Child Health” by Dr. Diane Benoit, MD. “Infants with secure attachment greet and/or approach the caregiver and… are able to return to play, which occurs in 55% of the general population,” she writes.

“Infants with insecure-avoidant attachment fail to greet and/or approach, appear oblivious to their caregiver’s return… in 23% of the general population. Infants with insecure-resistant [ambivalent] attachment are extremely distressed by separations and cannot be soothed at reunions, displaying much distress and angry resistance” and can’t return to play, “in 8%.” Benoit earlier reports that “in normal, middle class families, about 15% of infants develop disorganized attachment.” Her first three categories above add to 86%, leaving 14% disorganized.

“Of the four patterns of attachment, secure, avoidant, resistant and disorganized, disorganized attachment in infancy and early childhood is recognized as a powerful predictor for serious psychopathology and maladjustment in children. Children with disorganized attachment are more vulnerable to stress, have problems with regulation and control of negative emotions, display oppositional, hostile, aggressive behaviours…” Dr. Benoit says (emphasis added). [1]

Similar numbers were reported by van IJzendoorn, one of Benoit’s sources, in 1999. His global total of 6,282 infants in 10 countries was 51.5% secure, 17.4% avoidant, 10.6% ambivalent, and 20.5% disorganized. (Data from developing nations perhaps produced the large disorganized rate.)  Statistics for 2,104 North American infants were cited as 62% secure, 15% avoidant, 9% ambivalent, and 15% disorganized. [2, p.233, p. 229]

“Studies show that between 35% and 45% of all children in the US experience some kind of attachment issue,” reported, a respected therapists’ website [13]

“About 55 percent to 65 percent of children fall into the ‘secure’ attachment category,” wrote attachment scholar Sean Brotherson in 2005, “while about 10 percent to 15 percent show an ‘insecure-resistant/ambivalent’ pattern, 20 percent to 25 percent show an “insecure-avoidant” pattern and 15 percent to 20 percent an “insecure-disorganized.” [14]

“Disorganized” or Not?

Recently, however, there has been enough misuse of the term “disorganized” by courts and social services, that Dr. Main’s network seems to have resumed citing only the three 1978 Ainsworth categories.

“Main, Hesse, and Hesse (2011, p.441) have criticized the ‘widespread’ and ‘dangerous’ presumption that infants can be divided into four categories,” wrote History of Psychology magazine recently. “ ‘The reification of our work from its context… has lead readers to treat D as a category equivalent in kind to ABC, rather than recognizing it… runs orthogonal to the basic Ainsworth patterns’,” Dr. Judith Solomon is quoted to say. [15]

That would put securely-attached back up to Ainsworth’s 1978 figure of 70%  – despite many studies showing our attachment rate has fallen. [3]

Yet in numerous writings over 20 years, Main, Solomon and colleagues called the disorganized group a “new fourth category.” They also repeatedly wrote that when only three categories were in use, most disorganized infants had been “forced” into the “secure” category. This happened only because they didn’t fit the other two categories, but in fact these infants were not secure, they often wrote.

That’s why the use of four categories is so “widespread,” as in the quote from Dr. Benoit above about the “four patterns of attachment, secure, avoidant, resistant and disorganized.”

Perhaps others have abused the term “disorganized” to mis-label children as “hostile, aggressive, coercive” or to otherwise harm kids.  Perhaps law suits have resulted from this. That’s awful.

Yet while I respect and revere Dr. Main’s group, the notion that 70% of flesh and blood Americans have suddenly become “secure” just by using different words, or due to lawsuits, makes no sense. It’s certainly not my experience nor the experience of any therapist with whom I’ve spoken.

It also leaves most of the 50% of us with emotional pain from attachment trauma alone, without recognition or a voice, and with no idea that we even need help, let alone any idea how to get healing. That’s why I’m speaking up.

Examples of the original wording by Dr. Main’s group feature:

“By 1985, our laboratory had developed a number of new methods for assessing attachment. Besides Ainsworth’s original tripartite analysis of the Strange Situation procedure… we emphasize… disorganized attachment, devised by us as a new category of infant Strange Situation behavior…” [8, p.257]

“The next discovery… was based on the simultaneous breakthrough reported by Main and Solomon (1986, 1990) that a fourth Strange Situatio classification—disorganized/disoriented—could now be recognized. Infants were placed in this fourth category (see Lyons-Ruth & Jacobvitz, Chapter 28, and Solomon & George, Chapter 18, this volume) when they failed to maintain the behavioral organization characteristic of those classified as secure, avoidant, or ambivalent/ resistant… by 1990 infants were termed disorganized/ disoriented in the Strange Situation when, for example, they approached the parent with head averted, put hand to mouth in a gesture indicative of apprehension…or rose to approach the parent, then fell prone to the floor…” [10, p.572]

“[T]he limits of the traditional Ainsworth et al. (1978) coding system became apparent because many children with an established background of abuse or neglect nevertheless had to be forced into the secure category… Disorganization of attachment is usually considered a type of insecure attachment…” 2, p.226]

“Our chapter begins with a review of previous studies reporting difficulties in “forcing” each infant in a given sample into one of the three major categories… However, the majority of these…disorganized-disoriented infants would have been identified as secure… had we forced them into the standard classification system.” [11, p.97-8]

Dr. Ainsworth herself, co-authoring a volume with Dr. Main in 1990, writes: “… it is acceptable to consider adding…the fourth D category that Main identified as disorganized… Now that so much careful work has been done to explore the ramifications of this new category, it may be accepted as a valuable extension of the Strange Situation classificatory system.” [16]

Whatever one thinks about categories, Dr. Benoit’s closing advise on achieving secure attachment says it all:

“During the first six months of life, promptly picking up a baby who is crying is associated with four major outcomes by the end of the first year of life. First, the baby cries less. Second, the baby has learned to self-soothe. Third, if the baby needs the caregiver to soothe him/her, the baby will respond more promptly. And finally, the caregiver who responded promptly and warmly most of the time (not all the time; nobody can respond ideally all of the time) to the baby’s cries, will have created secure, organized attachment with all of the associated benefits.” [1]

The Adult Attachment Interview (AAI)

Mary Main '09 Bowlby-Ainsworth award(Inge Bretherton,Everett Waters)Moving on to adults, by 1982 Main had seen enough disorganized babies to conclude that there must be a lot of scary parents.

Now she needed a way to document the behavior of the parents. [17] (Right: Dr. Main (center) receives the 2009 Bowlby-Ainsworth Attachment Award.)

That’s how the AAI came about.  It was created in 1982 to discern the level of secure, loving attachment the parents had during their own childhoods with their babies’ grandparents. Parents were asked “both to described their attachment-related childhood experiences – especially their early relations with parents – and to evaluate the influence of these experiences on their development and current functioning.”

Parents were studied not just on facts they gave, but on how coherent a narrative they could produce quickly. That’s easy for folks who had a secure childhood, but difficult for folks who did not. The AAI questions are designed to “surprise the unconscious” to yield information about the “state of mind with regard to attachment” that might not otherwise show up. Each AAI was taped and transcribed verbatim. Then transcripts were classified by specific patterns by independent trained specialists. [18, 10]

In 1982 Main also created the “Berkeley Longitudinal Study” to take 67 Bay Area infant-parent pairs from her 1977 Strange Situation and study them for a generation. In 1977 the babies were 12-18 months old.  Main and her team re-studied these pairs when the kids reached age 6 in 1982, and studied the kids again when they reached 19 in 1995.  [8, 10]  In 1982 Main and her team gave three different tests to the Bay Area pairs:

— 1. The parents were given the AAI (without their children).  [8, 10]

— 2. The six year olds were again studied with their parents in the Strange Situation (as they’d been studied in 1977 as infants).

— 2. The “sixes” were also individually (without parents) given a new Separation Anxiety Test (SAT): they were shown pictures of children being separated from their parents, then asked how they felt. This was taped, transcribed and sorted by Main’s student Nancy Kaplan into Ainsworth’s three sets: secure, avoidant and ambivalent. [19, 8]

Dr Main’s initial 1982 AAI results were so important that it’s a tragedy that your doctor never learned this in medical school; your therapist (and mine) never heard of it; you’ve never heard of it; and so you have to read about it here, since the media doesn’t report it.

Amazing Results You’ve Never Heard

Mary Main, Erik Hesse '09 Bolwby-Ainsworth AwardFirst: Main noticed that the initial 1982 adult AAIs fell into categories very similar to those into which infants had fallen years earlier in the Strange Situation.  As Main had only Ainsworth’s three infant categories in 1982, she at first classified adults into Secure (matching infant “Secure”);  Dismissing (matching infant “Avoidant”); and Preoccupied (matching infant “Ambivalent”).  (Dr. Main and Dr. Hesse, above.)

Main was astonished to find that the more adults who took the AAI, the more the adult data (in three categories) statistically converged on Ainsworth’s “standard distribution” for infants: 20%, 70%, 10%.  [10, p.552]

Second: the levels of adult attachment in the adults’ 1982 AAIs corresponded specifically to the attachment to that adult which their own infant exhibited in the 1977 Strange Situation.  “Secure-autonomous (“valuing of attachment relationships and experiences) was associated with infant Strange Situation security,” reports Dr. Main’s co-author Dr. Erik Hesse, and this was true of all categories.

Thus, a marked relation between a parent’s hour-long discussion of his or her own attachment history, and the offspring’s Strange Situation behavior 5 years previously, had been uncovered.” [10, p.552]

Third: the match of each parent’s own attachment with their baby’s attachment five years back was at an unheard-of level. It correlated 75% of the time, in a field where a 20% correlation is highly significant.  [8, 10]

Fourth: the six year olds’ responses to parents in the Strange Situation in 1982 correlated strongly to their 1977 responses as infants — five years earlier. These kids at six responded to parents just as they had as infants, in the same percents. [7]

Fifth: the six year olds’ 1982 solo responses to the Separation Anxiety Test also produced the same results and percentages as their response to parents in the 1982 Strange Situation.

Sixth: In 1986, Main published her discovery of the fourth infant category of disorganized, and quickly found that if adults were sorted into four analogous categories, again results were the same.

“The next discovery regarding the AAI was based on the simultaneous breakthrough reported by Main and Solomon (1986, 1990) that a fourth Strange Situation (infant) classification—disorganized/disoriented—could now be recognized,” Hesse reports.By 1990, it had been shown that unresolved AAI status in a parent was predictive of disorganized attachment in the infant…  91% of unresolved mothers had infants who had been judged disorganized with them in the Strange Situation 5 years earlier…

“Thus there was now an AAI category corresponding to and predictive of each of the four Strange Situation categories…  In 1996… in a combined (meta-anlytic) sample of 584 non-clinical mothers… a four-way analysis showed… 16% dismissing, 55% secure-autonomous, 9% preoccupied, and 19% unresolved,” (emphasis added). In 2008, a global sample of AAIs for 1,012 average mothers showed the same results. [10]

These numbers for adults are astonishingly similar to the statistics for infants cited by Benoit, van IJzendoorn and others reviewed above: 23% avoidant, 55% secure, 8% ambivalent, and 14% disorganized.

“Just over 50 percent are secure, around 20 percent are anxious, 25 percent are avoidant, and the remaining 3 to 5 percent fall into a fourth, less common disorganized category,” Dr. Amir Levine, MD states in the popular 2010 book “Attached,” speaking of adults. [20]

Finally: When the 1977 infants reached age 19 in 1995, they too were given the Adult Attachment Interview.  Again results correlated strongly: the 19 year olds’ responses in the 1995 AAI correlated precisely to their infant behavior in the 1977 Strange Situation, their 1982 behavior at six, and to their parents’ 1982 responses to the AAI. [8]

Bottom line?  The Strange Situation predicts an infant’s emotional behavior for life, and the AAI proves it.  [8, 10]

“Researchers worldwide have replicated the relation originally uncovered in the Bay Area study between a parent’s status in the Adult Attachment Interview and an infant’s Strange Situation response to that same parent… The same average parent-to-child, secure/insecure match of 75% holds even when the interview is conducted before birth of the first child

“Describing the strength of this relation across studies conducted several years ago, van IJzendoorn (1995) calculated that it would take 1,087 further attempted replications, every one yielding insignificant results, to reduce the present relation between adult and infant attachment status to insignificance,” Main concluded in 2000 (emphasis added). [7, p.1091]

That’s not all.

Our attachment rate has fallen even further since 1999-2000, around the time that electronic devices began to increasingly replace human interaction.

“Well-developed human beings can self-regulate their emotional state by being with other humans. But what about people who regulate their emotional state with objects?” i.e. electronics, warned neuroscientist Dr. Stephen Porges in 2014.

“We’re in a world now being literally pushed on us, by people who are challenged in their own social and emotional regulation, and we’re calling this ‘social networking.’ We’re using computers, we’re texting — we’re stripping the human interaction from all interactions… We’re allowing the world to be organized upon the principles of individuals who have difficulty regulating emotionally in the presence of other human beings.” [21]

Harsh economic reality, meanwhile, has turned parents who can still manage to give attentive, attuned care to infants into an endangered species.

“For decades to protect the mother-infant bonding, there were families around it, there weren’t single individuals raising babies, there wasn’t early day care,” warned UCLA’s Dr. Allan Schore in a 2012 interview.

“The early day protections have been lost,” he said, “so as a result… there is an increase in psychiatric disorders in this country and the reason is that we have not protected this early phase of infant life, the early mother bonding…”

Schore, the top attachment scholar known as the “American Bowlby,” cited the 2009 study of 10,000 adult AAIs using three attachment categories which showed that from 1999 to 2009, secure attachment fell by 17.1%.   [3]

Using four attachment categories, if 55% were secure in 1999, then after a 17% drop, only 46% were secure in 2009.

That means 54% of the population was insecure as of 2009, if disorganized infants have not simply disappeared.

This insecure percentage has risen since 2009, and will continue to rise, as long as we raise infants in a society flooded with enough electronic devices and bad economic priorities to render their parents incapable of attachment.

A first step would be to address the information and policy vacuum regarding this “silent epidemic.” [22]

Kathy’s news blogs expand on her book “DON’T TRY THIS AT HOME: The Silent Epidemic of Attachment Disorder—How I accidentally regressed myself back to infancy and healed it all.” Watch for the continuing series each Friday, as she explores her journey of recovery by learning the hard way about Attachment Disorder in adults, adult Attachment Theory, and the Adult Attachment Interview.


Benoit, Diane, MD, FRCPC (2004) “Infant-parent attachment: Definition, types, antecedents, measurement and outcome,” Paediatr Child Health, Oct 2004; 9(8) p. 541–545, retrieved April 12, 2017 from:  “Infants with secure attachment greet and/or approach the caregiver and… are able to return to play, which occurs in 55% of the general population… Infants with insecure-avoidant attachment fail to greet and/or approach, appear oblivious to their caregiver’s return… in 23% of the general population. Infants with insecure-resistant [ambivalent] attachment are extremely distressed by separations and cannot be soothed at reunions, displaying much distress and angry resistance” and can’t return to play, “in 8%.” Benoit earlier reports that “in normal, middle class families, about 15% of  infants develop disorganized attachment.” Her first three categories above add to 86%, leaving 14% disorganized.

van IJzendoorn MH, Schuengel C, Bakermans-Kranenburg MJ (1999) “Disorganized attachment in early childhood: Meta-analysis of precursors, concomitants and sequelae.” Dev Psychopathol. 1999; 11:225–49.  “During  the past 10 years nearly 80 studies on disorganized attachment involving more than 6,000 infant-parent dyads [6,281 pairs total, Table 1, p.233] have been carried out….In  normal middle class families about 15% of the infants develop disorganized attachment behavior.  In other social contexts and clinical groups this percent may become twice or even three times higher (eg in cases of maltreatment) … Disorganization… is usually considered a type of insecure attachment.”

3 Bakermans-Kranenburg MJ, van IJzendoorn MH (2009) “The first 10,000 Adult Attachment Interviews,” Attach Hum Dev. 2009 May; 11(3): 223-63. Retrieved April 12, 2017 from
This study showed a 17.1% drop in the secure category from the authors’ previous 1999 study to 2009. It used adult analogs of only Ainsworth’s three 1978 categories: 20% dismissing, 70% secure, 10% preoccupied (=100%).  The 17.1% secure drop was reported as 23% dismissing, 58% secure, 19% preoccupied (=100%). It also showed “18% additionally coded for unresolved loss or other trauma.” Many of the 18% would be classed “disorganized” were all four categories used, but instead were “forced” into three categories. Since historically most disorganized subjects were “forced” into the secure category, the 58% secure number includes a significant number of disorganized.
–Interview with Dr. Allan N. Schore by David Roy Green, World Healing Electronic Network, July 2012, minute 39, retrieved April 12, 2017 from  Schore cites the secure attachment drop in the 2009 study above. “There is concern about this… that what was there for decades to protect the mother-infant bonding situation, there were families around it, there weren’t single individuals raising babies, there wasn’t early day care… the early day protections have been lost…so as a result of that, I’m now writing that there is an increase in psychiatric disorders in this country and the reasons is that we have not protected this early phase of infant life, the early mother bonding…”

Felitti VJ, MD; Anda RF, MD, et. al, 1998, “Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: The Adverse Childhood Experiences (ACE) Study,” American Journal of Preventive Medicine 1998;14:245–258.  Detailed article: For a popular text, see Karr-Morse, Robin, Wiley, Meredith,  “Scared Sick,”  Penguin Basic Books, 2012

5  Siegel, Daniel J. , MD,The Developing Mind: How Relationships and the Brain interact to shape who we are,” 1999 Guilford Press.  Earned secure attachment occurs when we began life with insecure attachment, but “are now securely attached… individuals whose experiences of childhood… produce insecure attachment (avoidant, ambivalent, or disorganized),” but  “…had a significant emotional relationship with a close friend, romantic partner, or therapist, which allowed them to develop… secure” attachment.
Dr. Mary Main, Dr. Erik Hesse, Dr. Daniel J. Siegel, Dr. Marion Solomon, “Adult Attachment Interview with Mary Main,” Lifespan Learning Institute, Los Angeles, Jul 9, 2010 video:   “It is possible to change attachment patterns, and we can help people as therapists to get the kind of new experiences that creates changes within people especially given the neuroplasticity of the brain,” says Dr. Solomon.
Siegel, Daniel J. , MD, “The Mindful Therapist: A Clinician’s Guide to Mindsight and Neural Integration,” W.W. Norton, 2010;   “Mindsight: The New Science of Personal Transformation,” Random House, 2010  “Mindfulness has been shown to be effective in healing insecure attachment. The purpose of both psychotherapy and mindfulness practice is to provide this internalized secure base. Attunement, whether it is internal in mindfulness, or interpersonal in attachment, is what leads to a sense of secure base.”
Siegel, 2007, Wallin, 2007, p. 5-6  “The regular exercise of mindful awareness seems to promote the same benefits–bodily and affective self-regulation, attuned communication with others, insight, empathy, and the like–that research has found to be associated with childhood histories of secure attachment,” Siegel writes. “ Mindfulness and secure attachment alike are capable of generating… the same invaluable psychological resource: an internalized secure base.”

Bowlby, John, “The Nature of a Child’s Tie to His Mother,” British Psychoanalytical Society, London, 1958; “Attachment and Loss,” New York, Basic Books, 1969

Main, Mary,  2000, “The Adult Attachment Interview: Fear, attention, safety and discourse processes;” also titled “The Organized Categories of Infant, Child, and Adult Attachment: Flexible vs. Inflexible Attention Under Attachment-Related Stress,” Jour of Amer Psychoanalytic Assoc, 48:1055-1095; 2000.  *p.1091: “The same average parent-to-child, secure/insecure match of 75% holds even when the interview is conducted before birth of the first child…”  Retrieved March 14, 2014 from

Main, Mary, 2005, with Hesse, Erik & Kaplan, Nancy, “Predictability of Attachment Behavior and Representational Processes at 1, 6, and 19 Years of Age – The Berkeley Longitudinal Study,” Chapter 10 of “Attachment from Infancy to Adulthood: The Major Longitudinal Studies,” edited by Klaus E. Grossmann, Karin Grossmann, and Everett Waters, pp. 245–304, New York: Guilford Press. Main refers to it as “Regensburg;” retrieved June 4, 2014 from Regensburg 2005 .pdf
–Main’s summary of the Strange Situation in this document:
“Ainsworth structured the Strange Situation procedure to include three of Bowlby’s ‘natural clues to danger’ in eight episodes:  1. Introduction to the room.  2. Mother and infant are left alone in a toy-filled environment whose unfamiliarity supplies the first natural clue to danger.  However, the mother’s presence is expected to provide the infant with security sufficient for exploration and/or play.  3. Providing a second clue to danger, a stranger joins the mother and infant.  4. The mother leaves the infant with the stranger, providing two combined clues to increased danger. 5. The mother returns, and the stranger departs…. Many infants initially seek proximity but then, reassured of their mothers’ nearness, resume play. 6.  The mother leaves, and the infant remains entirely alone in the unfamiliar setting. Infant distress can be strong at this point, and this episode is often terminated rapidly. 7. The stranger, rather than the mother, enters the room. 8. The mother returns… By now, most infants are expected to be crying, and actively not only seeking proximity to mothers, but also… indicating a strong desire to be held… Nonetheless, they are expected to settle and renew interest in exploration and play…
“Somewhat surprisingly, Ainsworth found that infant responses to separation and reunion in this procedure fell into three distinct, coherently organized patterns of attachment (“secure,” “insecure-avoidant,” and “insecure-ambivalent” (Ainsworth, Blehar, Waters, & Wall, 1978)… Given the length and complexity of this chapter, we suggest individuals divide reading to its three central parts (secure attachment, pp. 261–273; avoidant attachment, pp. 273–279; and disorganized attachment pp. 279–288).”

van Ijzendoorn, Marinus H.; Kroonenberg, Pieter M. (1988)  “Cross-Cultural Patterns of Attachment: A Meta-Analysis of the Strange Situation,” Child Development,Vol 59 No 1, Feb 1988, p.147–56; (1988 global replication of Ainsworth’s three categories; written before Main’s 4th category developed 1986-1990).   Retrieved July 2, 2016  from
Ainsworth, Mary D.S., Blehar, M.C., et al, “Patterns of attachment: A psychological study of the  Strange Situation,” Erlbaum, Hillsdale, NJ, 1978

10  Hesse, E., (2008) “The Adult Attachment Interview: Protocol, Method of Analysis, and Empirical Studies,” Chapter 25 of Cassidy, Jude &  Shaver, Phillip R. (Eds), “Handbook of Attachment: Theory, research, and clinical applications,” 2nd edition, 2008, p. 552-598, New York, Guilford Press, retrieved June 4, 2014 from
van IJzendoorn MH & Bakermans-Kranenburg,MJ (1996). “Attachment representations in mothers, fathers, adolescents and clinical groups,” Journal of Consulting and Clinical Psychology Vol 64(1), Feb 1996, 8-21.

11 Main, M., & Solomon, J. (1990). “Procedures for identifying infants as disorganized/disoriented during the Ainsworth strange situation,” in Greenberg, M. T., Cicchetti, D., & Cummings, M. (Eds.),. Attachment in the preschool years: Theory, research, and intervention (pp. 121-160), University of Chicago Press
Main M, & Hesse E. (1990). “Parents’ unresolved traumatic experiences are related to infant disorganized attachment status: Is frightened and/or frightening behavior the linking mechanism?” Chapter 5 in: Greenberg MT, Cicchetti D, Cummings EM, eds., Attachment in the Preschool Years, University of Chicago Press; 1990 p 161–82
Main M, & Solomon J (1986). “Discovery of an insecure disoriented attachment pattern: procedures, findings and implications for the classification of behavior,” in Brazelton T, Youngman M. Affective Development in Infancy. Norwood, NJ: Ablex; 1986. p. 95–124.
p.96 “In this chapter we describe commonalities observed in the behavior of 55 12- to 20-month-old  [55 twelve- to twenty-month-old]  infants whose strange situation behavior could not be classified using the traditional (A,B,C) classification system, infants who would nonetheless have been assigned (“forced”) into one of these three classifications.
p.97: Our chapter begins with a review of previous studies reporting difficulties in “forcing” each infant  in a given sample into one of the three major categories.
p.98:  “However, the majority of these “unclassified” (now, disorganized-disoriented) infants would have been identified as secure (group B) with the parent in the Strange Situation, had we forced them into the standard classification system.”

12  “Among children of American adolescent mothers, the rate is over 31% (Broussard 1995). Disorganized attachment is also common among the Dogon of Mali (~25%, True et al 2001), infants living on the outskirts of Cape Town, South Africa (~26%, Tomlinson et al 2005) and undernourished children in Chile (Waters and Valenzuela 1999),” reports [1] above.

13 “Attachment Issues,” unsigned, GoodTherapy.Org, July 3, 2015, retrieved April 3, 2017 from ]

14 “Understanding Attachment in Young Children,” Brotherson, Sean, Bright Beginnings #6, Oct. 2005, North Dakota State University, Fargo, ND, retrieved April 3, 2017 from

15 “The Emergence of the Disorganized/Disoriented (D) Attachment Classification, 1979–1982,” by Robbie Duschinsky, Hist Psychol. 2015 Feb; 18(1): 32–46, retrieved 4-10-17 from  Cited within which: “Attachment theory and research: Overview with suggested applications to child custody,” Main M., Hesse E., & Hesse S., Family Court Review, 49, 426–463, 2011 ]

16 Ainsworth, Mary, PhD,  “Epilogue: Some Considerations Regarding Theory and Assessment Relevent to Attachments beyond Infancy,” in “Attachment in the Preschool Years,” ed. M.T. Greenberg, D. Ciccheti & E.M. Cummings, Chicago University Press, 1990, p463-488; quote p.480.

17  George, Carol, Kaplan, Nancy, & Main, Mary, “Adult Attachment Interview,” Unpublished MS, Department of Psychology, University of California at Berkeley, third ed. 1996.  Original 74-page MS dated 1984, 1985, 1996.  Described in [10] Hesse 2008.  Retrieved March 20, 2014 from, now removed.  See also Main 1985; Main & Goldwyn 1984

18  Main, Mary B., “Adult Attachment Interview Protocol,” 11 pgs, 20 questions, no date or publisher. Dr. Main requires intensive training for use of the AAI.  The questions per se, marked “Do not reproduce this material without permission of the author,” were retrieved most recently April 12, 2017 from
UCLA’s Lifespan Learning Institute in Los Angeles holds AAI workshops and has an extensive CD lectures on the AAI at

19  Kaplan, Nancy, (1987), Separation Anxiety Test (SAT): “Individual differences in six-year-olds’ thoughts about separation: Predicted to actual experiences of separation,”  Unpublished doctoral dissertation, University of California, Berkeley.
Kaplan, N. (2003, April), “The development of attachment in the Bay Area study: One year, six years, nineteen years of age.” Paper at the biennial meeting of the Society for Research in Child Development, Tampa, FL.

20 “Attached: The New Science of Adult Attachment,” Dr. Amir Levine, MD & Rachel Heller, MA, Tarcher 2010; useful reviews at

21 Porges, Stephen, PhD, p. 15 of “Polyvagal Theory,” National Institute for the Clinical Application of Behavioral Medicine (NICABM), April 2012, retrieved July 21, 2014 from

22 Brous, Kathy, “The Silent Epidemic of Attachment Disorder,”


Medical Disclaimer: This website is for general information purposes only. It is simply my own research. Individuals should always see their health care provider or licensed psychotherapist before doing anything which they believe to be suggested or indicated herein. Any application of the material on this website is at the reader’s discretion and is the reader’s sole responsibility.

Comments are encouraged with the usual exceptions; rants, political speeches, off-color language, etc. are unlikely to post. Starting 8-22-16, software will limit comments to 1030 characters (2 long paragraphs) a while, until we get new software to take longer comments again.

89,492 total views, 75 views today

Share Button

32 responses to “The Adult Attachment Interview (AAI): Mary Main in a Strange Situation

  1. Alexandra Zanon

    Thank you for such an informative article. Currently, I am a student returning to school to become a marriage and family therapist. I am very interested in the work of Bowlby & Ainsworth and the theory of attachment. When doing research for my literature review I was amazed at the work of Dr. Mary Main. Your article was very helpful in understanding how the attachment theory and the work of Mary Main have contributed to how adverse childhood experiences impact future relationships and parenting. Currently I work as a parenting specialist and see how this important work and research can help improve outcomes for my families by understanding the underlying causes and the importance of addressing unresolved childhood trauma.

  2. Pingback: Adult Attachment Classification

  3. Pingback: Adult Attachment Interview

  4. Pingback: Adult Attachment Indicator

  5. Pingback: Adult Attachment Lab

  6. Pingback: Adult Attachment Disorder Nj

  7. After 20 years of trying to figure out why my husband can’t attach, this finally makes sense. Thank you for this great resource. Please send list of therapists in Orange County.

    • If we marry one, “we usually is one” (I was). Unless your hub wants therapy, the only way to get him there is for you to say “No” to bad treatment & go to therapy yourself. We can’t push others; we can only “model” mental health. I let my ex distance me for 27 years. I needed to say “No” and go to therapy for allowing bad treatment, before anything could change.

  8. Pingback: To Hide or to Cling |

  9. Thank you for this terrific info! This is very important. It links with (jargon-free) research I did and a resource website I have which you and others may find useful:
    Here’s a 2-minute read article about my research:

  10. Pingback: Articles on attachment theory | Joshua238's Blog

  11. Pingback: adult attachment interview - BOKEP SHUSUPUTIH


  13. Hello Kathy, Just a quick note to say how much I enjoy your writing and appreciate all your sleuthing on the subject of attachment and its history. – Marenka Cerny, MFT

  14. This excellent website really has all the information I wanted
    concerning this subject and didn’t know who to ask.

  15. Pingback: Mary Main in a Strange Situation: The Adult Att...

  16. Mary Ainsworth passed away. I believe Mary Main is at Berkeley. UCLA’s Lifespan Learning Institute in LA holds AAI workshops and has CDs on it:
    It would be nice to assess our own AAI score, but it’s dicey if people start assessing each other. It took me months to figure out how they discovered all this, as their writings are so obscure. Why? The more I study their documents, the more it seems they don’t want the average person using the AAI. Perhaps they are concerned lay persons might abuse the AAI to judge or classify each other? They insist that professional therapists get special training to use the AAI. Yes, all therapists should be getting training and using it! Yes, I’m upset most of them never heard of it. But us un-trained folks using it on each other? I wonder. Also check out Hesse, E., (2008) “The Adult Attachment Interview: Protocol, Method of Analysis, and Empirical Studies,” 2nd edition, 2008, p. 552-598 Guilford Press. try here: andMain, Mary B., “Adult Attachment Interview Protocol,” 11 pgs, gives 20 AAI questions, no date or publisher. The AAI questions per se, marked “Do not reproduce this material without permission of the author,” are here:

  17. One Love’s phone AP helps us (or our loved ones) assess our risk of being harmed in domestic violence:
    I wonder if we could decrease spanking and increase awareness of positive parenting with an AP to self assess your own AAI score. Parents seek support from family, friends, colleagues and the internet. They all could do an AAI to see if their loved one is at risk for problems attaching to their child. Try to download the One Love AP. It’s amazing! The Ravens gave $400K to support it. Whom can I contact? Can I reach Mary Ainsworth?

  18. Pingback: directory - Earn 50% per listing every month |

  19. I would like to alert your readers to my books – one in particular on infant development: Kenny D. T., “Bringing up Baby: The Psychoanalytic Infant Comes of Age,” (Karnac, 2012)
    It gives a detailed scholarly account of Attachment Theory:

  20. I am a practicing psychotherapist trained at one of only two CACREP accredited universities. I was trained on Bowlby, Ainsworth and the Strange Situation; so yes if you happened to be my client, your therapist was indeed trained.

  21. Joseph Maizlish, MFT.

    Among the 45% lacking secure attachment are many who have learned to cope & carry on a good relationship by, for example, mutual recognition of special needs. Thank you for tracking the history of Ainsworth’s categorization and Main’s further development. I hope you get to present this in written and other forms to many lay and professionals.
    I hear echoing in my mind Dan Siegel’s remark that “The most robust indicator of balanced functioning of the young child’s brain is the coherence of the self-narrative of the principal caretaker.” Siegel’s remark was about the “SELF-narrative,” not the attachment narrative – but those must go together. Good relationships within a coherent-self caretaker and within a child and between them makes likely good relationships with others much more likely. From that we also get what Main called the “coherent narrative” regarding the parent’s attachment to their parents, as you note.
    What a high correlation between the parent’s attachment to their own parents, and their kids’ attachment to the parents! So these factors are shown to be moving through a lifetime and across generations.

    • Ever since I discovered this material, I’ve been doing my utmost to get that “earned secure attachment.” It’s so difficult re-wiring our brains; hope I’ve made a little progress in 5 years…Thank you for noting that this was a tough story to piece together. I get queries every day from all over the country saying therapists like you and Sue are hard to find.

  22. Thank so much, I’m eager to have your article! I’ve done 5 or 6 blogs on Dan Siegel and one on Allan Schore at:
    Have a look through my blogs at
    I was at Dan’s UCLA conference in March 2013 where the new campaign Relationships First was created (I signed the email list but never heard from them.) I’d love it if you could put an update on Relationships First in your blog for my site, so we could promote it and give readers an update!

  23. Hi Kathy, I would love to help! Dr. Dan Siegel and Dr. Allan Schore are rock stars in this arena. I want to recommend Dr. Stan Tatkin, who studied with Dr. Schore, and his Psychobiological Approach to Couples Therapy (PACT) informed by attachment studies and neuroscience. Fantastic stuff, we have to get couples/parents feeling secure so they can provide a secure attachment experience to their children. They have a new campaign called Relationships First which Dr. Siegel and Dr. Tatkin are founders (along with Harville Hendrix and other heavyweights) with the intention of highlighting the importance of relationship (attachment) for healthy living. I’d love to submit an article with my insights regarding finding a good attachment-based therapist. I will be in touch, and thanks again for your site.

  24. “However, her results were also so important it’s outrageous that your family doctor never learned about this in medical school; your therapist (and mine) never heard of this…”
    Love your site, love it! I understand your frustration with the lack of knowledge some in the therapeutic community display about attachment and trauma. But, NOT ALL OF US! Ainsworth’s Strange Situation is taught in masters level counseling psychology courses. Any psychotherapist should at least know about it. I’m a psychotherapist interested in attachment and a trauma informed approach and I work with many others in the field who are likewise. I fear readers may eschew therapy and the empathetic, healing connection it offers because of some bad apples. We know it’s the empathic connection between the therapist and the client that informs healing, not the cognitive based exercise they take home. Therapy can and has been successful for many no matter what the approach, as long as there is a therapeutic alliance and empathic connection. There are some really good therapists! Talented, heart centered, empathic, well educated therapists and social workers and doctors who connect at a heart level. True, our culture has been organized around a model that emphasizes left brain answers to right brain problems for a long time. The good news is we are all going through a paradigm shift together. Thank you again for your site, it is informative and much needed.

    • I am delighted to hear that Ainsworth’s Strange Situation is taught in masters level psychology courses and practicing psychotherapists should know of it.I had to suffer three therapists who were ignorant of these principles, and damaged me to the point that I quit therapy, before I did the research myself and learned this. See my book chapters starting with “I Oughtta Have My Head…” So many have had this experience.
      BUT OF COURSE no one can heal without a good attachment-based therapist! Thus my book title; it’s far to dangerous to “do it yourself.”
      Could you write a piece on how to find a good attachment-based therapist? I get queries on it from all over the world.

  25. Pingback: Developmental Trauma: What you Can’t See CAN Hurt You | "Don't Try This at Home"

  26. Pingback: The Hole in Me | "Don't Try This at Home"

  27. Pingback: The Hole in Half of Us | "Don't Try This at Home"

Leave a Reply to Gheri Cancel reply

Your email address will not be published. Required fields are marked *

 characters available