History
Adult attachment: from four categories to two dimensions
The vocabulary of adult attachment — secure, anxious, avoidant — reached general circulation as a set of types. The measurement literature abandoned types more than twenty-five years ago, and the popular version has not caught up.
The field's founding study was published in 1987 by Cindy Hazan and Phillip Shaver in the Journal of Personality and Social Psychology. Their proposal was that the patterns John Bowlby and Mary Ainsworth had described in infants and caregivers might also describe adult romantic bonds. To test it, they printed a questionnaire in a Denver newspaper, the Rocky Mountain News, and analysed the replies that arrived by post.
The instrument asked respondents to select which of three paragraphs best described them. The paragraphs corresponded to the three infant classifications, rewritten for adult relationships. Distribution of responses across the three roughly matched the distribution reported in the infant literature, which the authors treated as preliminary support for the analogy.
A newspaper questionnaire produces a self-selected sample: people who read that paper, noticed the item, and chose to reply. Nothing about the design supports generalisation to a wider population. The measure was also a single forced choice between three paragraphs, which cannot separate the several distinct ideas each paragraph contains.
Hazan and Shaver presented the work as exploratory, and the paper reads that way. The strong reading arrived later, as the three labels detached from the study that produced them.
Measurement improved quickly. Kim Bartholomew and Leonard Horowitz proposed a four-category scheme in 1991, generated by crossing two underlying axes rather than listing types. In 1998 Kelly Brennan, Catherine Clark and Phillip Shaver took a different route: they assembled the attachment items then in use, administered them together, and factor-analysed the pool. Two dimensions emerged, which they named anxiety — concern about abandonment and about a partner's availability — and avoidance, meaning discomfort with closeness and dependence. The resulting Experiences in Close Relationships scale, and its later revision by Chris Fraley and colleagues, became the standard instrument.
The same year, Fraley and Niels Waller published a taxometric analysis addressing the question directly: do the data contain discrete groups, or a continuous distribution? Taxometric procedures are designed to distinguish the two. Their results indicated that adult attachment is dimensional. Individuals differ in degree along the two axes; they do not sort into natural kinds with boundaries between them.
This is a substantive difference, not a technical footnote. Under a categorical model, someone belongs to a type, and the type is a fact about them. Under a dimensional model, a category label is a summary of a position on continua — informative in aggregate, arbitrary at the boundary, and unstable for anyone scoring near the middle. Assigning discrete labels to continuous scores also discards information and reduces statistical power, which is the practical reason the research literature stopped doing it.
Stability has been examined separately. Longitudinal work reports moderate correlations over periods of years, with meaningful movement in both directions, and scores that shift with relationship circumstances. The dimensions behave less like fixed traits than early descriptions suggested.
The categorical vocabulary persists because four words are easier to circulate than two continuous scores. That is a fact about language, not about the evidence. Where the type labels appear, they are best read as shorthand for regions of a two-dimensional space that the measurement literature has described in that form since 1998.
Sources
- Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology.
- Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology.
- Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview.
- Fraley, R. C., & Waller, N. G. (1998). Adult attachment patterns: A test of the typological model.
This article summarises published research. It is not professional advice, does not assess any individual situation, and makes no claim about outcomes. See the disclaimer.