Quick answer
Fear of intimacy, on the 1991 Descutner and Thelen scale, is anxiety about close dating relationships. It tracks with loneliness and less disclosure. Bartholomew and Horowitz showed that avoidance itself splits into fearful and dismissing patterns.
Fear of intimacy, in the research that named and measured it, is anxiety about exchanging personally significant thoughts and feelings inside a valued close relationship. It is not a mood you catch, it is not a moral failure, and it is not a diagnosis in the diagnostic manuals. Two papers do most of the useful work. One built a scale and showed that higher scores travel with loneliness, less self-disclosure, and briefer relationships. The other showed that "avoidance" is at least two patterns, with different views of the self and of other people. Treating them as one problem produces advice that fits neither.
The useful work is in the measurement paper and in the two-pattern model of avoidance.
What the Fear-of-Intimacy Scale actually measures
In 1991 Carol Descutner and Mark Thelen published the Fear-of-Intimacy Scale in Psychological Assessment. Across two studies they described it as a measure of anxiety about close dating relationships. Item analysis left a 35-item scale with high internal consistency and test-retest reliability. Construct checks mattered more than the item count. Scores correlated positively with a loneliness measure. They correlated negatively with self-disclosure, with social intimacy, and with social desirability, and those relations held when social desirability was partialed out. People with higher scores reported briefer relationships. Scores also lined up with therapists' ratings of clients' fear of intimacy. The authors' abstract is careful about use: the scale looked promising for clinical assessment and for research. It did not claim to discover a disease.
That last point is the one popular essays skip. A high score is a description of anxiety and inhibition around closeness, corroborated in their data by relationship length and by clinicians who knew the person. It is not avoidant personality disorder, not social anxiety disorder, and not evidence that someone is incapable of love. If a website tells you the scale "proves" you have a disorder, it is adding a claim the 1991 paper does not make.
The correlation pattern is the practical core. Fear of intimacy, as this instrument defines it, sits next to saying less about yourself and next to feeling lonely. The protection and the cost arrive together. Less disclosure predicts less of the contact that would correct the fear, and the loneliness measure moves in the other direction. The scale does not show that disclosure is always safe. It shows that the people who most fear closeness are, in these samples, already disclosing less and already more lonely.
Two kinds of distance, not one
A few months later in 1991, Kim Bartholomew and Leonard Horowitz tested a four-category model of adult attachment in the Journal of Personality and Social Psychology. They crossed a positive or negative model of the self with a positive or negative model of other people.
Secure is the cell most people mean by "comfortable with closeness": worthy of love, and expecting others to be responsive. Preoccupied is the opposite trouble from avoidance: a negative model of the self plus a positive model of others, so acceptance has to keep being won from outside. The two cells that both avoid intimacy are the ones that get blurred together in ordinary speech.
Fearful avoidance, in their words, combines a sense of unworthiness with an expectation that others will be untrustworthy and rejecting. Avoiding close involvement protects the person against the rejection they anticipate. Dismissing avoidance combines a sense of being worthy of love with a negative view of other people. These people, Bartholomew and Horowitz write, protect themselves against disappointment "by avoiding close relationships and maintaining a sense of independence and invulnerability." Both styles avoid intimacy. They do not avoid it for the same reason, and they do not sit at the same level of dependence. Fearful people still need others' acceptance to hold a positive self-regard, and they stay back to limit the disappointment. Dismissing people are less dependent on that acceptance. Their distance is backed by self-sufficiency, not by an open plea for reassurance.
The paper's own results keep the distinction from being a metaphor. Interview ratings of secure and fearful were negatively correlated, and so were preoccupied and dismissing, which is what the model predicts for opposite cells. A self-concept analysis found the effect on the self dimension: secure and dismissing participants, the two groups with a positive model of the self, scored higher on self-esteem, self-acceptance, and lower subjective distress than the fearful and preoccupied groups. Sociability tracked the model of others. Secure and preoccupied participants scored higher on sociability than fearful and dismissing participants. Friends' reports agreed that the dismissing group looked especially low on sociability. On interpersonal problems, dismissing participants' profile sat on the cold side of the circumplex, and friends tended to describe them as more introverted than they described themselves.
So "he is afraid of intimacy" can mean at least two observable pictures. One person wants closeness, expects to be rejected, and hangs back. Another person rates the self as fine, rates other people as not worth the dependence, and looks cold or socially absent to friends. A partner who answers both pictures with the same speech ("just let me in") is answering a theory, not a person.
Hazan and Shaver's 1987 paper, which Bartholomew and Horowitz discuss, had already treated romantic love as an attachment process and sorted adults into secure, avoidant, and anxious-ambivalent groups. Bartholomew and Horowitz's point is that the single avoidant category hid the split. Interview methods in the Adult Attachment tradition tended to see avoidant adults as people who denied distress and downplayed attachment needs. Self-report methods tended to see people who felt distress when they got close. Those are closer to dismissing and fearful, respectively, than they are to one trait called "avoidant."
What distance looks like without a quote sheet
In a fearful pattern, the distance often shows up as a wish followed by a cancellation. The person wants the conversation, then hears their own disclosure as dangerous, then retreats and calls the retreat realism. Because the model of the other is negative, a kind response can be filed as temporary or as a setup. The loneliness correlation on the Fear-of-Intimacy Scale fits this cell more naturally than it fits pure indifference: the closeness still matters, which is why its absence hurts.
In a dismissing pattern, the distance is more likely to be explained as taste. Need looks like weakness. Other people's feelings look excessive. Friends, in Bartholomew and Horowitz's data, saw less sociability than the self-report emphasized. A partner who keeps offering deeper talk and keeps being told that everything is fine is not always meeting a hidden yearning of the fearful type. Sometimes they are meeting a working model in which intimacy itself is the threat to independence.
Neither picture is "commitment phobia" as a single switch, and neither is the same thing as stonewalling in a fight. Conflict withdrawal in marital observation research is about what happens when a discussion floods. The patterns in this article are about whether a valued other is allowed to matter in the first place. Someone can be dismissing about closeness and still argue heatedly. Someone can be fearful and still stay in the room, silent, because leaving would also be a disclosure.
Changing the pattern the model actually describes
The papers say which lever belongs to which pattern.
If the pattern is fearful, the self-model and the other-model are both negative. Reassurance poured on a person who expects rejection often fails, because the expectation is doing the protecting Descutner and Thelen's scale associates with inhibited disclosure. The experiment that matches the evidence is small, specific, and falsifiable: say one personally significant thing to one valued person, on purpose, and record what they actually do, not what the model predicted. The 1991 scale's validity data say that higher fear travels with less disclosure and with briefer relationships. A single disclosed fact will not rewrite a model. A repeated mismatch between "they will reject this" and what the other person does is the kind of evidence the model is built to notice. If every trial confirms the fear, that is data too. The task then is safer people, not braver exposure to someone who does punish closeness.
If the pattern is dismissing, self-esteem lectures miss the cell. Bartholomew and Horowitz placed this style with a positive self-model. The protective move is independence and a downgrading of other people. The matching experiment is not "believe you are lovable." It is letting a chosen person matter on a schedule you picked: one standing conversation, one favor you do not score, one week when you do not translate their need as an attack on your freedom. Friends in their study already saw the sociability gap. The work is to take that outside view seriously without collapsing into the fearful cell's self-blame.
When the anxiety is rigid, when relationships keep ending at the same threshold, or when a therapist's outside rating would clearly diverge from the story you tell, the 1991 paper's own suggestion applies. The scale was related to therapists' ratings and was offered for clinical assessment. A clinician can tell fearful distance from dismissing distance, and can tell both from trauma, coercion, or a partner who is not safe to approach. This article cannot do that from the page.
If you are the partner of someone who keeps distance, the useful question is which cell you are looking at. Pursuit that demands instant openness terrifies a fearful model and confirms a dismissing model's belief that closeness means being overtaken. Clarity about your own limit ("I want a relationship where some personal things get said; I will not chase a maybe for a year") is fairer than a campaign to cure them. Their pattern is information. It is not a debt they owe your timeline.
Sources
- Descutner, C. J., & Thelen, M. H. (1991). Development and validation of a Fear-of-Intimacy Scale. Psychological Assessment, 3(2), 218–225. https://doi.org/10.1037/1040-3590.3.2.218
- Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61(2), 226–244. https://pubmed.ncbi.nlm.nih.gov/1920064/
- Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524. https://doi.org/10.1037/0022-3514.52.3.511
Frequently Asked Questions
- Is fear of intimacy a psychiatric diagnosis?
- No. Descutner and Thelen's 1991 Fear-of-Intimacy Scale measures anxiety about close dating relationships. Higher scores related to loneliness, lower self-disclosure, briefer relationships, and therapists' ratings. The paper does not turn the score into a disorder.
- How is fearful avoidance different from dismissing avoidance?
- In Bartholomew and Horowitz's 1991 model, fearful people hold a negative view of themselves and of others, and they keep distance to avoid expected rejection. Dismissing people hold a positive view of themselves and a negative view of others, and they protect independence by not needing anyone.
- Did friends see these styles the same way the person did?
- Friends' ratings reproduced the model, and friends described the dismissing group as especially low on sociability. Dismissing participants' interpersonal problems clustered on the cold side, with friends leaning toward seeing them as introverted.
- Why does telling someone to just open up often fail?
- A fearful pattern expects rejection, so pressure can feel like the danger the distance was built to prevent. A dismissing pattern is not primarily a self-esteem problem in this model, so reassurance that they are lovable misses the downgrading of other people.
- What is a first step that matches the research?
- For a fearful pattern, disclose one personally significant thing to one safe person and compare the result with the predicted rejection. For a dismissing pattern, let one chosen person matter on a schedule you picked. If closeness has been punished, the first step is a safer person, not more exposure.
- Is this the same as stonewalling during a fight?
- No. Conflict withdrawal is about shutting down inside an argument. Fearful and dismissing patterns are about whether a valued person is allowed to matter at all. Someone can argue heatedly and still keep a dismissing stance toward closeness.