It can be one of the most confusing experiences in a relationship.
Someone appears genuinely interested in you. They seek closeness, show affection, talk openly and perhaps seem intensely invested in the relationship.
Then, as the relationship becomes more serious, something changes.
They become harder to reach emotionally. Difficult conversations are avoided. Requests for reassurance seem to create irritation or withdrawal. After conflict, they may need considerable space. In some relationships, the person who once seemed enthusiastic about getting closer begins questioning the relationship itself.
This creates an understandable question:
If someone finds closeness so uncomfortable, why get into a relationship in the first place?
Attachment theory offers a more useful explanation than simply concluding that the person never cared.
The apparent contradiction is often not between wanting love and not wanting love.
It is between wanting connection and what happens internally when connection starts to involve vulnerability, dependence and emotional risk.
Avoidant attachment does not mean avoiding relationships
The term avoidant attachment can be misleading.
It can sound as though someone with an avoidant attachment pattern simply wants to avoid romantic relationships altogether.
That isn't what attachment avoidance means.
Modern adult attachment research commonly describes attachment using two broad dimensions: attachment anxiety and attachment avoidance.
Someone higher in attachment avoidance tends to be less comfortable depending on other people, revealing vulnerability or having another person depend heavily upon them. Greater psychological or emotional distance may feel safer and easier to regulate.
That does not necessarily remove the desire for companionship, attraction, sex, affection, partnership or love.
A person can therefore genuinely want a relationship while also having difficulty with some of the things a close relationship eventually requires.
And that distinction explains a great deal.
The beginning of a relationship may feel very different from established intimacy
Early romantic relationships contain excitement, attraction and possibility.
They may also contain relatively little real dependence.
Each partner still has considerable independence. There may not yet be shared finances, domestic responsibilities, expectations about the future or a long history of emotionally difficult conversations.
Someone can therefore feel extremely enthusiastic during the early stages of a relationship without experiencing the same attachment pressures that develop later.
As intimacy increases, the psychological stakes change.
The relationship starts asking questions such as:
Can I depend on this person?
Can they depend on me?
What happens if I disappoint them?
What happens if I need them?
What happens if they reject me after I have become vulnerable?
For someone uncomfortable with dependence and vulnerability, these questions can create tension even when the relationship itself is good.
This is one reason an initially intense relationship followed by withdrawal should not automatically be interpreted as deliberate deception.
Equally, it should not automatically be excused as "just their attachment style."
Both interpretations are too simple.
What attachment researchers mean by "deactivation"
Attachment theory provides a particularly useful concept for understanding withdrawal: deactivating strategies.
When an attachment situation feels threatening, people higher in avoidance may regulate that discomfort by reducing their engagement with attachment-related thoughts, emotions or behaviours.
Psychologically, the strategy is something like:
Closeness feels uncomfortable, so reduce closeness.
This can manifest in many ways.
A person might become absorbed in work or other activities, emphasise their independence, minimise their emotional needs, become uncomfortable discussing the relationship, focus increasingly on a partner's faults, withdraw following conflict or find it difficult to ask for support.
None of these behaviours proves that somebody has an avoidant attachment style. There are many reasons people withdraw from relationships.
But reducing emotional engagement when attachment-related stress increases is consistent with what researchers describe as an avoidant regulatory strategy. Research also suggests that outward emotional distance does not necessarily mean an absence of internal distress.
That distinction matters.
A person who appears remarkably calm while withdrawing may not necessarily be experiencing the situation as calmly as they appear.
"But they seemed so loving before"
This is where relationships involving avoidance can become particularly confusing.
Imagine that someone is warm, affectionate and highly engaged during the first months of a relationship.
Later they become distant.
The other partner naturally tries to reconcile these two versions of the same person:
Which one was real?
Attachment theory does not give us grounds to assume that the affection at the beginning was fake.
Feelings can be genuine while the person's capacity to tolerate everything those feelings eventually require is more limited.
Intimacy when a relationship feels exciting and relatively unconstrained is different from intimacy when somebody needs reassurance after an argument, wants to discuss commitment or asks you to remain emotionally present while both of you are upset.
The real test of relationship security is therefore not simply how intensely someone connects.
It is also how consistently they remain capable of connection when closeness becomes difficult.
Intensity and security are not the same thing.
Avoidant attachment is not an inability to love
One particularly damaging interpretation of avoidant attachment is that people with avoidant tendencies simply cannot love as deeply as securely attached or anxiously attached people.
Attachment research does not support such a conclusion.
Attachment dimensions describe patterns involving intimacy, dependence, perceived availability, emotion regulation and responses to attachment-related threat. They are not measures of someone's "capacity for love."
Someone can care deeply for another person and still behave in ways that make a relationship extremely difficult.
Those two statements can both be true.
This also leads to an important distinction:
Understanding someone's behaviour is not the same as accepting it.
Knowing that withdrawal might be connected with an attachment strategy does not make ghosting, contempt, repeated stonewalling, infidelity, cruelty or chronic refusal to communicate harmless.
Attachment theory can help explain patterns.
It does not remove personal responsibility.
Dismissive-avoidant and fearful-avoidant patterns are not identical
Every person described as "avoidant" should not be treated as though they behave identically.
Adult attachment is better understood dimensionally, particularly through attachment anxiety and avoidance. The familiar labels are useful shorthand for different combinations of those dimensions.
Someone commonly described as dismissive-avoidant tends to be relatively high in avoidance without equally high attachment anxiety. Independence may be strongly emphasised and attachment needs may be minimised.
Someone commonly described as fearful-avoidant has stronger avoidance alongside substantial anxiety about relationships. This can create a more obvious approach-and-retreat pattern: wanting closeness while simultaneously feeling threatened by it.
Those distinctions help explain why two people who are both called "avoidant" can behave very differently.
It also explains why diagnosing a partner from a few behaviours is unreliable.
What happens when an anxious partner meets an avoidant partner?
Attachment becomes especially interesting when we stop looking at one person in isolation.
Relationships contain two attachment systems.
Someone higher in attachment anxiety may respond to uncertainty by seeking greater connection, reassurance and confirmation that the relationship is safe.
Someone higher in attachment avoidance may respond to relational pressure by seeking more autonomy and emotional distance.
Put those responses together and a self-reinforcing pattern can develop:
One person feels distance and moves closer.
The other experiences the increased pursuit as pressure and moves further away.
The additional distance makes the first person even more anxious.
The increased anxiety produces more attempts to reconnect.
Neither person necessarily intended to create the cycle.
Yet each person's attempt to feel safer can inadvertently make the other feel less safe.
Research examining couples rather than individuals supports the importance of considering both partners' attachment orientations. A partner's anxiety and avoidance can interact with our own attachment tendencies and relate to trust, satisfaction and commitment within the relationship.
This is why discovering both partners' attachment patterns can sometimes be considerably more illuminating than identifying one person's style alone.
The real problem is often what happens after conflict
Taking space during an argument is not inherently unhealthy.
Sometimes temporarily stepping away allows both people to regulate themselves and return to a conversation more constructively.
The important word is return.
There is a major difference between:
"I'm overwhelmed. I need an hour, but let's talk again tonight."
and an indefinite emotional shutdown in which the original problem is never revisited.
Healthy relationships require repair.
That means being able to return after disconnection, understand what happened, acknowledge each person's experience, take responsibility where appropriate and work out what should happen differently next time.
Without repair, temporary peace can hide accumulating resentment.
This is also where attachment explanations have limits.
A person may understand perfectly well why their attachment system makes conflict difficult.
They still have to learn how to participate in conflict differently.
Can someone love you and still want to leave?
Yes.
And this may be one of the hardest parts of relationships to accept.
Love is an important component of a relationship, but it does not automatically provide emotional regulation, communication skills, compatibility, accountability or the ability to tolerate intimacy.
A person may have genuine feelings for their partner and still conclude—rationally or defensively—that distance feels safer.
Conversely, someone's decision to leave does not automatically demonstrate that avoidance caused the breakup.
Sometimes people leave because a relationship isn't right for them.
Attachment theory should never become a framework in which every rejection means:
"They loved me so much that their attachment system became frightened."
That explanation may feel comforting, but there is usually no way to know another person's internal motivation with such certainty.
Sometimes the most accurate answer is simply that we do not know.
Can a secure partner "fix" an avoidant partner?
This is another idea that deserves caution.
A responsive and emotionally secure relationship can certainly provide conditions that support security. Attachment relationships are dynamic, and people's experiences with partners matter.
But being secure does not mean becoming someone else's therapist.
Nor does it mean remaining indefinitely available while another person repeatedly withdraws, refuses to communicate or behaves destructively.
A secure partner can offer consistency.
They can communicate clearly.
They can avoid escalating a pursue-withdraw cycle.
They can respect a reasonable request for space.
They can establish healthy boundaries.
What they cannot do is perform another person's emotional work for them.
Greater security ultimately requires participation from the person whose patterns are causing difficulty: recognising those patterns, tolerating uncomfortable emotions, communicating more directly and practising different responses.
Sometimes a secure response to another person's withdrawal is patience.
Sometimes it is a boundary.
And sometimes it is leaving a relationship that can no longer meet your needs.
Attachment styles explain tendencies, not destinies
Perhaps the most important point is that an attachment style is not a character diagnosis.
People are not simply "anxious people", "avoidants" and "secures".
Attachment orientations vary in degree. They can also differ across relationships and change through experience.
Research consistently associates attachment insecurity with aspects of relationship functioning, wellbeing and responses to stress, but those findings describe statistical tendencies across groups—not an inevitable script for every individual.
Two people with similar attachment scores can therefore behave quite differently.
Self-awareness matters.
Communication matters.
Experience matters.
The partner matters.
And what someone actually does with their patterns matters enormously.
So why does someone with avoidant tendencies keep seeking relationships?
Usually because the supposed contradiction isn't really a contradiction.
Human beings can simultaneously desire connection and protect themselves from aspects of connection that feel dangerous.
A person may want companionship while finding dependence uncomfortable.
They may want intimacy but struggle with vulnerability.
They may fall in love and still retreat when the emotional stakes become high.
They may genuinely intend for a new relationship to be different while discovering that familiar protective strategies return when they feel threatened.
This does not make hurtful behaviour acceptable.
But it gives us a much more useful question than:
"Do avoidant people actually love?"
A better question is:
"What happens to this person's behaviour when closeness, vulnerability or conflict activates their attachment system?"
And there is an equally important question to ask yourself:
"What happens to mine?"
Because understanding attachment becomes most useful when it stops being a way of diagnosing the person across from us and starts becoming a way of recognising the pattern developing between us.
Your own tendencies towards reassurance, withdrawal, independence, closeness, conflict and repair all form part of that dynamic.
Discovering your attachment style—and, if you're in a relationship, understanding how your partner's pattern interacts with yours—can help make those otherwise confusing reactions easier to recognise.
Attachment styles describe tendencies rather than diagnoses. They can help you understand relationship patterns, but they should not be used to diagnose a partner, excuse harmful behaviour or predict with certainty why somebody behaves as they do.