Why Mixed Signals Are So Hard to Walk Away From: Attachment Trauma and Intermittent Reinforcement

There are relationships that are clearly bad and surprisingly easy to leave, and there are relationships that barely qualify as relationships and somehow occupy half your brain. The second category can be particularly confusing because sometimes there is not even enough of a relationship to explain the amount of emotional energy it is consuming. The person is interested, except when they aren't. They initiate, disappear, come back warm, become vague again, say something intimate, fail to follow through, and occasionally give you just enough evidence that you start wondering whether this time something has actually changed.

If you have ever been caught in one of these dynamics, you may have noticed that inconsistency does not necessarily make you less interested. Sometimes it does the exact opposite. You think about the person more, analyze interactions more carefully, feel enormous relief when they return, and become increasingly preoccupied with trying to understand what is happening. This is often described online as “intermittent reinforcement,” and there is some useful psychology underneath that term, but the internet has also turned it into one of those concepts that now gets used to explain almost everything.

The more interesting question is not simply whether inconsistent attention can strengthen persistence. It can. The more useful question is why this particular kind of uncertainty, with this particular person, can become so difficult to leave alone.

Mixed signals keep the question open

An outright rejection hurts, but it gives the brain something concrete to work with. A person who says, “I don't want a relationship with you,” may leave you heartbroken, but there is considerably less information to analyze than there is when someone says they care about you, acts deeply connected for several days, disappears for a week, comes back with a reasonable explanation, becomes affectionate again, and then does essentially the same thing a month later.

Ambiguity creates a different problem because the brain keeps trying to solve it. You are not only grieving the relationship; you are continually evaluating whether there is actually a relationship to grieve. Every new interaction can reopen the question. Maybe they really were overwhelmed. Maybe that affectionate conversation meant more than you thought. Maybe they are scared. Maybe they are avoidant. Maybe you asked for too much. Maybe you did not ask for enough. Maybe you just need to give them space. There are enough plausible explanations that the analysis can continue almost indefinitely.

This is one reason I am cautious when people immediately interpret this kind of preoccupation as proof that they are anxiously attached. Attachment anxiety can absolutely intensify sensitivity to ambiguous relational cues, and research suggests that people higher in rejection sensitivity may be particularly alert to possible signs of rejection and have more difficulty disengaging once they perceive it. But sometimes a person feels uncertain because the situation is genuinely uncertain. You do not need an insecure attachment style to become confused by someone who is behaving confusingly.

The question is usually not, “Is this my attachment system or is this person inconsistent?” Frequently, it is some amount of both.

What intermittent reinforcement actually means

Intermittent reinforcement comes from learning theory. Very broadly, a behavior that is rewarded only some of the time can sometimes become quite persistent, even after the reward stops. This is referred to as the partial reinforcement extinction effect, and it has been studied for decades in learning research.

The classic examples usually involve behaviors that are reinforced unpredictably. You do something repeatedly because sometimes it produces the desired result, and because you do not know which attempt will work, the absence of a reward does not immediately tell you that the reward is gone for good. It may simply mean, not this time.

You can probably see why people apply this concept to relationships. If somebody is affectionate every single day and then stops completely, the change is fairly obvious. If somebody has always alternated between closeness and distance, a period of distance does not necessarily signal an ending because distance has previously been followed by reconnection. You already have evidence that waiting sometimes pays off.

Where I think we need to be a little careful is turning this into “mixed signals literally hijack your dopamine and make you addicted to the person.” That is a much bigger claim than the behavioral research allows us to make. Human attachment relationships are considerably more complicated than laboratory reinforcement schedules, and people are not slot machines with texting privileges.

Still, the basic principle is useful. Unpredictable positive experiences can help sustain behavior because the next rewarding interaction remains possible. If you have had ten disappointing interactions followed by one intensely connecting conversation, that conversation may not simply feel good; it can restore your belief that the version of the relationship you want still exists.

The best moments can become disproportionately important

This is where mixed-signal relationships can get particularly sticky. People do not necessarily evaluate relationships by mathematically averaging every interaction. A few emotionally intense experiences can become enormously important, especially when those experiences seem to reveal what the relationship could be.

You may have two weeks of vague communication followed by a four-hour conversation in which the other person is vulnerable, affectionate, funny, attentive, and completely present. For those four hours, the relationship feels easy. You see the version of them you have been waiting for, and because that version is real—they really were warm, connected, and present—it becomes difficult to dismiss.

The problem is not that you imagined the good interaction. The problem is deciding how much weight it deserves.

We tend to get ourselves into trouble when we evaluate a relationship according to its peak moments rather than its overall pattern. Someone may be capable of extraordinary intimacy and still be unable or unwilling to sustain the type of relationship you want. Someone can genuinely care about you and still be inconsistent. Someone can mean what they say while they are saying it and still repeatedly fail to follow through afterward.

Those things are frustrating because they do not fit neatly into “they care” or “they don't care.” Sometimes the answer is that they care and their behavior is still not enough for you.

Attachment can make uncertainty louder

Attachment theory is useful here, provided we do not turn it into astrology with childhood trauma. Adults vary in attachment anxiety and avoidance, and those differences can affect how people respond when relationships feel threatened or uncertain. People higher in attachment anxiety tend to be particularly sensitive to signs that connection may be at risk, while attachment avoidance is more often associated with strategies that reduce or suppress attachment needs under stress.

For someone with more attachment anxiety, inconsistency can create an almost perfect environment for hyperactivation. You become more attentive to small changes in contact, more motivated to restore connection, and more likely to think about the relationship precisely when connection feels uncertain. Research on relational ambivalence has also found that attachment anxiety is associated with simultaneous pulls toward closeness and distance, which helps explain why these relationships can involve wanting somebody intensely while also feeling angry, exhausted, or desperate to stop caring.

This does not mean every uncomfortable reaction is an attachment wound that needs to be healed out of existence. If somebody has established a pattern of disappearing after intimacy, promising things they do not follow through on, or becoming affectionate primarily when they sense you withdrawing, noticing that inconsistency is not pathology. Sometimes anxiety is providing useful information.

Where attachment history can complicate things is in how difficult it becomes to tolerate the uncertainty without trying to resolve it. The unanswered question itself can begin to feel unbearable. You may find yourself less focused on whether you actually like the relationship you are having and increasingly focused on trying to establish what the relationship means.

You can become attached to potential

One of the harder things to recognize is that sometimes we are grieving a future more than a present relationship. There may have been enough genuine connection to make the imagined future plausible, but not enough consistency for that future to actually exist.

Potential is powerful because it cannot really disappoint you in the same way reality can. Reality is the unanswered text, the canceled plan, the affectionate weekend followed by emotional distance, or the conversation about doing things differently that somehow never becomes different. Potential is what the relationship would be if the person were less overwhelmed, less afraid, less avoidant, more available, finally divorced, finished with this work crisis, farther along in therapy, or able to give you the version of themselves that occasionally appears.

The occasional really good interaction can keep potential alive because it seems like evidence. You are not fantasizing about a completely imaginary person; you have seen them be exactly what you want. The painful question is whether that person represents their actual relational capacity or simply something they can offer intermittently.

That distinction can take a surprisingly long time to accept.

Why walking away can make the longing temporarily worse

People sometimes assume that if a relationship is unhealthy or unsatisfying, creating distance will immediately bring relief. Occasionally it does. Other times distance makes the preoccupation temporarily louder because you have removed the behavior that previously reduced uncertainty.

If checking your phone occasionally resulted in a message, checking had a payoff. If initiating contact sometimes produced several days of closeness, reaching out had a payoff. If threatening to leave resulted in reassurance, pulling away had a payoff. Removing those behaviors means you also lose the possibility of the immediate relief they sometimes produced.

Then there is grief, which is probably less exciting than saying your dopamine receptors are being hijacked but usually deserves more attention. Walking away from an inconsistent relationship means giving up not only what you were receiving but also what you kept hoping you were eventually going to receive. You may be grieving the future conversation, the eventual realization, the person finally understanding what they are losing, or the fantasy that enough patience would eventually turn intermittent closeness into something dependable.

That grief can make returning feel very tempting because returning restores possibility. Unfortunately, possibility and compatibility are not the same thing.

Mixed signals are different from normal human inconsistency

Nobody is perfectly consistent. People get sick, busy, depressed, overwhelmed, distracted, frightened, or temporarily less communicative. A partner can need space without being avoidant, take longer to respond without rejecting you, and have periods of lower emotional availability without the relationship becoming fundamentally unstable.

The useful question is usually not whether inconsistency ever occurs. It is whether there is a recognizable pattern and whether the relationship can absorb conversations about that pattern.

A person who says, “I've been disappearing when I get overwhelmed, I can see how confusing that is, and I want to figure out something more workable with you,” is doing something very different from a person who repeatedly disappears and then tells you that your reaction to the disappearance is the real problem. Someone who cannot provide as much contact as you prefer may still be reliable about what they can provide. Two people can also have completely legitimate needs that simply do not fit very well together.

This is another place where attachment language can accidentally become unhelpful. If every difference gets labeled anxious or avoidant, we start treating incompatibility as a psychological puzzle that somebody needs to solve. Sometimes you need more contact than another person wants to offer. Nobody has to be disordered for that to hurt.

Pay attention to the pattern, not the explanation

When a relationship is confusing, people often spend enormous amounts of time trying to understand why the other person behaves the way they do. Maybe they are frightened by intimacy. Maybe they had a difficult childhood. Maybe their marriage damaged their ability to trust. Maybe work really is overwhelming. Maybe they are depressed. Any of those explanations can be true, and understanding someone compassionately can be valuable.

The problem is that explanation can quietly become a substitute for information. Knowing why someone struggles to respond consistently does not tell you whether they are able to respond consistently. Understanding why someone disappears does not make them present. Knowing why somebody cannot commit does not create a commitment.

I think this is where people can become almost too psychologically generous. They keep finding increasingly sophisticated explanations for behavior that is making them miserable. Eventually, the more useful question may be much simpler: Whatever the reason is, what is it actually like for me to be in this relationship?

That question can be harder because it shifts attention away from solving the other person and back toward your own experience.

Sometimes the hardest part is accepting that the connection was real

There is a temptation when trying to leave an inconsistent relationship to rewrite the entire thing as fake. We tell ourselves the person never cared, everything was manipulation, or we were stupid for believing any of it. Sometimes manipulation is present, of course, but often the reality is less satisfying because two things were true at once.

The connection may have been real, and the relationship may still have been inadequate. The person may have cared deeply and still been unable to offer enough. The intimate conversations may have been sincere, and the repeated disappearances may also have been real. You do not necessarily have to discredit the good parts in order to take the painful parts seriously.

In fact, accepting both often helps more than forcing yourself into a simpler story. If the connection meant nothing, then of course you ought to be able to stop caring immediately. If it did mean something, then there is actually something to grieve.

The goal is not to convince yourself that you never wanted the person. It is to become more accurate about the relationship you were actually having instead of continually evaluating it according to the best version of what it occasionally became.

The question may not be “Do they care about me?”

This is probably the question people ask most often in inconsistent relationships, and unfortunately it is not always the most useful one. People can care and still be unavailable, unreliable, ambivalent, frightened, selfish, overwhelmed, incompatible, or simply unable to participate in the kind of relationship you want.

A better set of questions is often less dramatic. Is there enough consistency here for me? When something goes wrong, can we talk about it? Do words and behavior generally line up over time? Am I spending most of my energy enjoying this relationship or trying to determine whether I have one? When I communicate a need, does something actually change, or do we simply have another very insightful conversation about why nothing changes?

That last one can be particularly brutal for psychologically sophisticated people because a good conversation can feel like progress. Sometimes it is. Sometimes two people understand the dynamic beautifully and then continue doing exactly the same thing.

At some point, understanding the pattern has to be accompanied by information from behavior.

Why this can be so difficult to let go of

If mixed signals have kept you attached to someone much longer than seems reasonable, it does not necessarily mean there is something uniquely dysfunctional about you. Ambiguity can sustain hope, intermittent positive experiences can support persistence, and attachment insecurity or rejection sensitivity can make uncertain relationships especially consuming. Add genuine affection, chemistry, history, loneliness, hope, and a few really wonderful experiences, and it is not particularly mysterious that walking away can be difficult.

The challenge is usually not learning to feel absolutely nothing. It is learning to evaluate the relationship using a larger sample of evidence. The beautiful conversation counts, and so does the week afterward. The apology counts, and so does what happens after the apology. The affection counts, and so does whether you can reliably access the person when the relationship becomes uncomfortable rather than exciting.

Mixed signals are powerful partly because every good moment invites you to reconsider the entire story. Eventually, it can be useful to reverse that process and ask whether one good moment actually changes the overall pattern.

Sometimes it does. Sometimes it really doesn't.