The most repeated piece of pop-psych in the last decade goes something like this: "There's a difference between love and attachment. Love is healthy. Attachment is when you're addicted to someone, when you can't let go, when you stay because of fear rather than choice. Real love has no attachment in it." It's everywhere — TikTok carousels, dating-coach Instagrams, well-meaning friends at the bottom of the second bottle.
It's also wrong, in a fairly specific way. The framing comes from a misreading of John Bowlby's attachment theory crossed with a borrowed Buddhist vocabulary, and it has caused a lot of people to spend a lot of time worrying that what they actually have — the deep, settled, bone-level bond with their partner — is somehow the wrong kind of connection. It isn't.
This piece pulls the two concepts apart, looks at what Bowlby actually said, and explains why the "love vs attachment" framing as it lives online today is missing the thing the research is actually pointing at — which is much more useful than the pop version.
What Bowlby Actually Said
John Bowlby was a British psychoanalyst working at the Tavistock Clinic in London from the 1940s onwards. His three-volume Attachment and Loss (1969, 1973, 1980) made the case that humans are born with an evolved behavioural system whose function is to keep us in proximity to a small number of specific other people — first our primary caregivers, later our romantic partners and close friends. The system gets activated by threat, novelty, fatigue, illness; it gets deactivated by reliable, responsive contact with the attachment figure.
The crucial point: in Bowlby's framework, attachment isn't a pathology. It's the underlying architecture of long, secure bonds. Securely-attached children grow into adults who maintain close, durable, comfortable connections with the people who matter to them. That is what the system is for. The reason we have attachment-related distress (separation anxiety, grief at loss) is precisely because we have the capacity for attachment-related connection — they're two sides of the same evolved mechanism.
Cindy Hazan and Phillip Shaver's 1987 paper Romantic Love Conceptualized as an Attachment Process (Journal of Personality and Social Psychology) extended this directly to adult romantic relationships. Their argument: romantic love between adults uses the same attachment system that bonded children to their caregivers, just operating in a different context with a different attachment figure. Love and attachment, in this framework, are not opposites. Adult romantic love is an attachment relationship. (See attachment theory in dating.)
If you'd asked Bowlby or Hazan or Shaver in 1987 "is your bond with your partner love or attachment?" they would have looked confused. The question, in their framework, doesn't make sense. It's like asking "is what your liver does metabolism or chemistry?" Different levels of description for the same thing.
So Where Did the Pop-Psych Split Come From?
The "love vs attachment" framing as it appears online has two roots, both legitimate in their original context but mangled in transmission.
The first root is the Buddhist concept of upādāna, often translated as "attachment" or "clinging" — the grasping toward objects of desire that the tradition identifies as a source of suffering. In the original framework, this is a specific mental posture, not a category of relationship. The mistranslation happens when "clinging" gets folded into "attachment" in English, picks up the broader emotional connotations of that word, and then gets re-applied to relationship bonds. By the time it shows up on a dating-advice Instagram, "attachment" has come to mean something like "unhealthy emotional dependency" — a meaning Bowlby never used and Buddhism wasn't pointing at either.
The second root is genuine clinical observation. Therapists working with people in destructive relationships do see a pattern that's worth naming — a pattern where someone stays in a relationship that isn't working, can't seem to leave, returns after every breakup, and describes the bond as "I just can't let go." This is real. The clinical literature calls this protest behaviour (Bowlby's own term, actually), or, more recently, traumatic bonding in the Patrick Carnes sense. It's an activation pattern of the attachment system under conditions of intermittent reinforcement, similar to the way variable-ratio reward schedules produce particularly sticky behaviour. It is a real and worth-naming phenomenon.
The pop-psych framing takes this clinical pattern and calls it "attachment", as if all attachment were this pattern. That's the slippage. The result is that someone whose relationship is going perfectly well, who feels deeply bonded to their long-term partner, reads a TikTok and starts wondering whether what they have is "real love" or "just attachment." It usually is real love. The framing was always wrong. (See love vs infatuation.)
The Distinction That Actually Matters
There is a useful distinction in this neighbourhood. It's just not "love vs attachment." It's something more like: secure-functioning attachment versus insecure-functioning attachment. Both are forms of attachment. Both can feel like love. They produce very different relational outcomes.
Secure-functioning attachment, in the framework Stan Tatkin lays out (drawing on Mary Main, Bowlby, and the broader infant-research tradition), is the pattern where the bond between two people is reliable, mutual, regulated, and oriented toward each other's wellbeing. Either partner can express need without performance, receive comfort without keeping score, withdraw briefly without it being read as abandonment, return without elaborate repair. The system runs on relatively low background activation. The relationship feels safe. (See secure-functioning couples.)
Insecure-functioning attachment, by contrast, is what most people are gesturing at when they say "attachment, not love." It's the pattern where the bond is intense but unstable, where one or both partners are operating from high baseline activation, where the connection requires a lot of effort to maintain and breaks down repeatedly. The intensity often feels like love — and the body is producing many of the same chemistry (oxytocin, dopamine, the cortisol cycle of intermittent reinforcement) — but the structural pattern is corrosive.
"The question isn't love versus attachment. It's which kind of attachment, and what shape it's making your life take. Some attachment is the architecture of a steady life. Some attachment is the architecture of a slow erosion. Both feel like love."
The Five Tests That Actually Work
If you're genuinely trying to distinguish a healthy bond from a destructive one — which is the useful question hiding inside the "love vs attachment" confusion — these five framing questions cut through more cleanly than the standard pop-psych ones.
Test 1 — Does the relationship leave you bigger or smaller?
Arthur Aron's self-expansion model (1986, and an extensive empirical literature since) frames healthy long-term relationships as ones that expand the self — new interests, new capabilities, a wider range of possible action. Corrosive relationships do the opposite: they contract the self over time. Friends drift, hobbies fall away, the range of acceptable behaviour narrows. Both can feel intense. Only one is doing you net good.
Test 2 — Can you express need without performance?
Sue Johnson's EFT research, and the wider literature on what she calls "accessibility, responsiveness, engagement," finds that the durable bonds are the ones where both partners can ask for what they need — comfort, attention, distance — and have that request received without it becoming an event. The relationships labelled as "attachment, not love" in the pop sense often fail this test specifically: need can only be expressed through protest behaviour because direct expression isn't safe.
Test 3 — Is the intensity proportional to the depth?
New love is intense for everyone. The Helen Fisher brain-imaging research on early-stage attraction shows extremely high dopaminergic activation in the first 12–18 months across more or less all romantic relationships. What differs is what's underneath. In a healthy bond, the intensity dials down as deeper structures (shared history, integrated lives, settled trust) take over. In a destructive bond, the intensity stays at peak because nothing deeper is being built underneath — the activation is the relationship. (See when the honeymoon phase ends.)
Test 4 — How does it look from the outside?
The view from outside the relationship is often more accurate than the view from inside it. People in secure-functioning attachment relationships have friends and family who describe the relationship as "good for you" or "you seem really happy with them." People in insecure-functioning relationships have friends and family who are quietly worried, or who have stopped saying anything because they've learned that saying anything causes a fight. The outside view isn't always right. It's almost always worth taking seriously.
Test 5 — What would you do in their absence?
This is the cleanest diagnostic. Imagine your partner is away for a week. Are you fine — slightly missing them, but resourced, sleeping, eating, working, seeing friends? Or does the absence trigger acute distress, sleeplessness, a felt sense of being unable to function? The first is secure attachment; the second is insecure-anxious attachment activation. The first is what a long bond should look like in the everyday. The second is a signal that something in the bond is doing more regulation work than it should be doing alone. (See anxious attachment in dating.)
"I Love Them But I'm Not Sure I'm IN Love" — A Different Question
A separate version of the love-vs-attachment confusion shows up when someone says: "I love them. I'm attached to them. But I'm not sure I'm in love with them any more." This is usually pointing at a different distinction — the difference between passionate love and companionate love, which is a real distinction with a real research literature behind it.
Elaine Hatfield and Ellen Berscheid in the 1970s and 1980s distinguished passionate love (intense, often jealous, with high physiological activation) from companionate love (steady, trusting, deep, less aroused). The dominant pattern across long relationships is that passionate love fades from peak intensity within 18 months to 2 years, while companionate love can grow and deepen for decades. Most long-term couples experience both, at different times, in different mixtures. The fading of passionate-love peak intensity is not the relationship dying; it's the relationship maturing into something different.
The mistake is to read the fading of passion as "I must just be attached now, not in love." That's the wrong reading. It's more accurate to say: the passionate phase is over (which it always is, in every long relationship, however good), and what's underneath has become visible. If what's underneath is companionate love plus secure attachment plus shared life — that's not a downgrade from love. That's what love at month 24 and beyond actually is.
If what's underneath is companionate emptiness plus going through the motions plus structural mismatch — that's a different problem, and the diagnosis isn't "we became attached"; it's "we're not actually well-matched and the passion was covering for it."
Built for the long bond, not the high
LoveCertain matches on values, life stage, attachment style and communication style — and only shows matches above 70% compatibility. The peak-passion years fade in every relationship; what's left is what you matched on. £49 once. Full refund if no relationship in 90 days.
The Buddhist Question, Properly Translated
The Buddhist concept of upādāna does have something useful to say about relationships — just not what the pop version takes from it. The genuine teaching points at the suffering caused by clinging to specific outcomes, by demanding that a relationship be a particular thing forever, by gripping on to a partner as the source of one's wellbeing in a way that displaces all other sources.
Translated into relationship terms: the harmful pattern is not "attachment to your partner." The harmful pattern is using your partner as the entirety of your emotional infrastructure — no friends, no inner resources, no capacity to be alone — and then experiencing the relationship as fragile because everything hangs on it. The work is to widen the base, not to weaken the bond. A securely-attached person in a secure-functioning relationship has both the deep bond and the wider life. (See becoming securely attached as an adult.)
This reading is consistent with what Bowlby actually argued: attachment is not the problem; insecure attachment is the problem. The work is to develop secure attachment, not to dismantle attachment.
A note on "avoidant" framings
One way the "love is not attachment" framing gets weaponised is by avoidant-attachment partners who use it to justify keeping their partner at arm's length: "real love wouldn't need this much contact, that's just attachment." The Bartholomew four-quadrant model would call this the dismissing-avoidant pattern wearing a philosophical costume. The healthy intervention is the opposite — letting yourself want your partner properly. (See avoidant attachment.)
The Practical Implications
The reason any of this matters is that the language you use to describe what's between you and your partner shapes how you treat the relationship. If you've absorbed the idea that "deep attachment" is a flaw to be worked through, you'll spend energy trying to dismantle the wrong thing — the secure base that's actually doing the work of holding the relationship up.
If, instead, you can name what's there accurately — "we're securely attached, with companionate love that grew out of passionate love, and the bond is strong and reliable" — you can spend energy on the right things: maintenance, expression, repair after small ruptures, occasional renewal of novelty and play. The work of a long relationship is largely about keeping a secure-functioning attachment in good condition. That's a more interesting and more useful project than trying to figure out whether what you have is "really love." (See the weekly check-in template.)
If Your Bond Genuinely Isn't Working
The honest other side of this: sometimes what people call "I'm attached, not in love" is real, and what it's pointing at is that the relationship has structurally finished, the bond is the only thing keeping them present, and the relationship is now harmful to one or both partners. This is a real situation. The framework above gives you tools to tell the difference.
If the relationship fails most of the five tests above — if you're contracting not expanding, can't express need without performance, the intensity is brittle not deepening, the outside view is worried, and the absence test surfaces relief more than missing — what you have is something the Buddhist tradition was pointing at: a holding pattern that's outlived its usefulness. Naming it accurately is the first step. The second is the harder work of either repair or ending. The Esther Perel Mating in Captivity literature, and the Sue Johnson EFT framework on the other side of it, both treat this honestly. (See communication breakdown.)
The Certain Letter
Weekly relationship-science briefings. 4-minute read.
Where This Sits in the LoveCertain Approach
Our matching algorithm gives substantial weight to attachment compatibility — 20% of the overall score — because the research on what makes long relationships work consistently identifies attachment-pattern fit as a stronger predictor than passion intensity, shared interests, or even values overlap on the early-stage measures. We're explicitly trying to set up the structural conditions for secure-functioning bonds, because the secure-functioning bond is the long-term version of love. (See how matching works and compatibility science.)
For an accessible academic summary of the underlying attachment-theory literature, the Simply Psychology overview of attachment styles covers Bowlby/Ainsworth/Hazan/Shaver in plain language and links to primary sources.
The Honest Encouragement
If you're worried that what you have with your partner is "just attachment" — pause, and apply the five tests above. The odds are very high that what you have is secure attachment, that secure attachment is what mature love actually looks like, and that the pop-psych framing made you doubt the real thing. The bond is allowed to be deep. The bond is allowed to feel settled. The bond is allowed to feel like home in a way that doesn't have the early-months electricity. That's not the relationship downgrading. That's the relationship arriving.
If, on the other hand, you've applied the five tests honestly and the answers come back consistently in the wrong direction — that's worth taking seriously, and it's worth the harder conversation that follows. Either way, "is this love or attachment?" was the wrong question. "Is this secure-functioning?" is the right one. (See secure attachment.)
