Attachment styles explained

Four styles are commonly described in adults. Each names a habitual way of handling two things at once: the need for closeness and the fear of losing it. The definitions below describe movements, not characters.

• Secure: closeness is comfortable and so is independence. Asking for help costs little, and a disagreement is not experienced as a threat to the bond.
• Anxious, also called preoccupied: the need for closeness runs high and doubt about the other person's availability keeps returning. Signs of distance are spotted early and weigh heavily.
• Avoidant, also called dismissing: independence comes first and depending on someone feels uncomfortable. Emotions tend to be handled alone, and distance is what soothes.
• Disorganised, also called fearful: both movements coexist, seeking closeness and guarding against it. It is the most delicate of the four descriptions, and the one most often misused.

These four markers read better as quadrants of a plane than as boxes. The two axes are attachment anxiety, meaning the fear that the other person will move away, and avoidance of closeness, meaning discomfort with depending on someone.

From watching one-year-olds. In 1970, Mary Ainsworth and Silvia Bell published in Child Development a procedure in which a toddler is separated from their caregiver and then reunited, in an unfamiliar room. It is the reunion, more than the separation, that brings out three distinct ways of handling the moment.

A fourth description was added later for conflicted or disoriented behaviour that fitted none of the first three. A consensus statement signed by more than forty researchers, published in 2017 in Attachment & Human Development, warns explicitly against how it is used: this classification is not a validated individual-level marker, it cannot establish anything about a child's history, and its sensitivity and specificity are insufficient for that purpose.

The move to adults dates from 1987, with Cindy Hazan and Phillip Shaver's article in the Journal of Personality and Social Psychology, which proposed reading romantic relationships as attachment bonds. The four-category model itself was proposed in 1991 by Kim Bartholomew and Leonard Horowitz, built on two axes: the model of the self and the model of others.

One last step explains why tendencies is a better word than types. In 2000, Fraley, Waller and Brennan reanalysed self-report measures of adult attachment and showed that they describe two continuous dimensions better than a set of categories. In other words, the border between two neighbouring styles does not exist in the data: it is a reading convention.

Yes, and longitudinal work says something more precise than that. The review by Fraley and Roisman, published in 2019 in Current Opinion in Psychology, draws four lessons from those follow-up studies, and none of them points towards a fixed destiny.

• Adult attachment does appear to originate in early caregiving experiences, but those associations are weak and inconsistent across measurement domains.
• Styles appear more malleable in childhood and adolescence than in adulthood.
• Early experiences do not determine adult outcomes.
• Much remains to be understood, in particular relationship-specific patterns rather than one style per person.

The practical consequence runs both ways. A result obtained at twenty does not describe how someone will function at forty. Equally, change should not be expected to be fast or spontaneous: the malleability observed is greater before adulthood, which makes the work slower afterwards, not impossible.

The main useful purpose of this vocabulary is descriptive: it helps name a sequence, not designate a culprit. The point where it turns harmful is easy to spot, and it is the point where it is used to explain the other person's behaviour on their behalf.

1. Describe a specific recent situation rather than a style: what was said, what you felt, what you did next.
2. Look at the sequence rather than the starting point. A request for closeness and a need for space feed one another, and working out who started leads nowhere.
3. State an observable need, not a verdict on the person: “I need us to say when we will be home” can be discussed, “you are avoidant” cannot.
4. If the subject keeps returning and weighs on you, a professional trained in couple or individual therapy works on exactly this.