Two children, two houses, on the same street.
In the first house, nobody looks up. The mother is busy with her own difficulties, and the child's drawings, first words, and small triumphs land in a room where no one is watching. He learns early that if he wants anything he will have to arrange it himself.
In the second house, nobody looks away. The mother knows everything — what he ate, who he spoke to, what he is thinking, whether the friend is a good influence. His achievements are discussed at length, because they are also hers. He is rarely allowed out. He is adored without interruption.
Twenty-five years later, both of them have the same problem.
That is the finding, and it is the least intuitive claim in Greene's account of narcissism: in the backgrounds of almost all deep narcissists we find either abandonment or enmeshment.1
Not a spectrum with neglect at one end and smothering at the other, with health in the middle. Two opposite conditions producing one identical outcome — and the outcome is not low self-esteem, or difficulty trusting, or any of the things you would predict. It is that they have no self to retreat to, no foundation for self-esteem, and are completely dependent on the attention they can get from others to make them feel alive and worthy.1
The interesting question is why two opposite inputs converge, and the answer is genuinely elegant once you see it.
Go back to the process the two houses were both supposed to be running.
The key window is roughly ages two to five. The child is separating from the mother, and confronting two facts at once: it cannot get instant gratification, and it is alone and yet dependent on its parents for survival.2
That is not a comfortable position, and the child's solution is a piece of psychological engineering worth admiring. It identifies with the best qualities of its parents — their strength, their ability to soothe — and incorporates those qualities into itself.2
Read what is actually required for that transfer to work, because the requirements are what both houses failed.
The original has to be visible. You cannot copy a soothing function you have never seen operate on you.
The copy has to be permitted to run. The child has to be allowed to try self-soothing, badly, and to have the attempt go well enough that it is worth repeating.
And there has to be room for something separate to exist. The whole point of the operation is that the child ends up with its own version — not a shared one, not a borrowed one it has to keep going back for.
Greene names the parental conditions accordingly: if the parents encourage the child's first efforts at independence, validate its need to feel strong, and recognize its unique qualities, then the self-image takes root and can be built on.2
Three conditions. Now look at which one each house breaks.
The abandoning parent might be a deep narcissist themselves — too self-absorbed to acknowledge the child, to encourage its early efforts at independence.1
Greene's phrasing is worth attending to, because abandonment here does not require anyone leaving. The parent is present. The parent may be feeding, clothing, and housing the child competently. What is absent is acknowledgement, which is a much more specific thing.
The mechanism of failure: the child's first efforts at independence produce no response. It manages something and there is no one to register that it managed. And since the whole developmental operation depends on transcribing a soothing function off an adult who is performing it, a child who is never soothed has no original to copy from.
So the self-image never takes root — not because the child was told it was worthless, but because the process that builds one requires a witness and there was not one.
What that child learns instead is the wrong lesson, and it learns it correctly: attention is real but it comes from outside and it has to be extracted. Which is not a false belief. It is an accurate generalization from the available data, and it becomes a life.
The other route is stranger and much less recognized, and Greene describes it precisely.
The parents could be enmeshers — overinvolved in the child's life, suffocating it with attention, isolating it from others, and living through its advancement as a means to validate their own self-worth.
And then the operative clause: they give the child no room to establish a self.1
Four elements are named there, and each does specific damage.
Suffocating attention. The abandoned child got none. This one gets a flood — and the flood is just as fatal, because attention that never withdraws never leaves a gap for the internal regulator to be needed in. You do not build a thermostat in a room where someone is permanently adjusting the heat.
Isolation from others. This is the element people miss, and it is not incidental. Other children, other adults, other households are where a child discovers that its family's way is a way rather than the way — which is the raw material of having tastes and opinions of one's own. Cut that off and there is nothing to construct a distinct self out of.
Living through the child's advancement. Here is the sharpest part. The achievements are real and the praise is real, and the praise is not for the child. It is the parent's self-worth being validated through a proxy. Which means the child receives, thousands of times, the message that its accomplishments matter enormously — and that they belong to someone else.
No room. The three above produce this. There is no unobserved space, no separate opinion, no achievement that is only yours. A self is built in privacy out of idiosyncratic material, and this child has neither.
Compare what each child looks like at twelve, because this is where the diagnostic difficulty lives.
The abandoned child looks like a problem. Under-supervised, hungry for notice, visibly working the room. Teachers write things down. It reads, correctly, as a child who is not getting enough.
The enmeshed child looks like a success story. Well turned out, high-achieving, close to their family, in constant contact with a parent who is deeply invested. Every visible indicator is positive. The parent is admired for their devotion.
And Greene's account says both children arrive at twenty-five with the same missing organ.1
That asymmetry in visibility has a consequence: the enmeshed route produces the version of this that nobody intervenes in, because there is no point at which an outside observer sees anything wrong. There is no neglect to report. The parent's behaviour is, by every conventional measure, love.
Greene's account of the branching at adolescence applies to both, and it is worth having because it produces two very different adult presentations from the same deficit.
If the child is an extrovert, they can function reasonably well and even thrive. They become masters at attracting notice and monopolizing attention, and they appear vivacious and exciting. In a child, those look like signs of future social success. Underneath, they are becoming dangerously addicted to the hits of attention they stimulate to make them feel whole and worthy.3
If the child is an introvert, they retreat into a fantasy life, imagining a self quite superior to others. And because that self is too unrealistic ever to be validated by anyone, they also have moments of great doubt and self-loathing — Greene's phrase is that they are either a god or a worm. Lacking a coherent core, they could imagine themselves to be anyone, so the fantasies keep shifting as they try on new personalities.3
Same hole. Four possible presentations, depending on the route in and the temperament.
Greene supplies the developmental scaffolding elsewhere in the book, and it maps onto these two routes closely enough to be worth stating.
He uses Bowlby's four attachment schemas: free/autonomous, dismissing, enmeshed-ambivalent, and disorganized.4
The free/autonomous stamp comes from mothers who give their children freedom to discover themselves, remain continually sensitive to their needs, and also protect them. Note that all three are required, and that the first two are in tension — freedom and continual sensitivity — which is why this is difficult rather than obvious.
Dismissing mothers are often distant, sometimes hostile and rejecting. Such children are stamped with a feeling of abandonment and the idea that they must continually fend for themselves. That is Greene's first route.
Enmeshed-ambivalent mothers are not consistent with their attention — sometimes suffocating and overinvolved, other times retreating because of their own problems or anxieties. And they can make their children feel as if they have to take care of the person who should be taking care of them.4 That is the second route, with one detail Greene's narcissism chapter leaves out: the enmeshment is intermittent.
That detail matters more than it looks, and the handshake below takes it up.
The adult patterns follow. Children of the dismissing parent tend to avoid negative emotional situations and wall themselves off from feelings of dependency — they may find it harder to commit, or unconsciously push people away. Children of the enmeshed variety experience a great deal of anxiety in relationships and feel many conflicting emotions, remaining always ambivalent: they pursue people and then unconsciously retreat.4
Two opposite childhoods; two opposite-looking adult strategies; one shared inability to be alone with yourself comfortably.
Greene's own examples sort onto the two routes, and it is worth noting how thin his etiological evidence usually is.
Tolstoy: his mother died when he was two, and Greene says it left a giant hole in him that he could never fill, which he attempted to fill with numerous affairs. He behaved recklessly in his youth as if that could make him feel alive and whole, was continually disgusted with himself, and could not figure out who he was — pouring the uncertainty into his novels and assuming different roles in the characters he created.5 That is the abandonment route in its most literal form, and it also demonstrates the shifting-identities symptom: he was, professionally, a man who could be anyone.
Stalin: a father who beat him mercilessly and a rather cold and unloving mother.6 Two sentences, total, for one of the most consequential personalities of the twentieth century.
Jeanne de Belciel: in her earliest years she displayed an insatiable appetite for attention and wearied her parents, who finally got rid of her by sending her to a convent.7 Note the direction of Greene's causal claim here — the appetite came first, and the abandonment was a response to it. That is the reverse of the model this page describes, and he does not flag the reversal.
Which brings the evidence problem into focus rather than hiding it. Greene's etiologies are retrospective, brief, and constructed after the outcome is known. A cold mother is available in the biography of almost any historical figure if you go looking, and the inference from adult behaviour back to a childhood cause is not a finding — it is a narrative move.
You are trying to work out why a particular relationship keeps going the same way, and the usual explanations have stopped being useful.
Do not run this diagnostically on somebody else. Greene's material invites that and it is the wrong use — you cannot verify a childhood you were not present for, and the confident reconstruction of another adult's parents is one of the least attractive things this book can be used for.
Run it on yourself, and run it on a mechanism you can actually check in the present, not on a memory.
Ask the question that separates the two routes without requiring you to remember anything: when something good happens to you and nobody knows about it yet, how long can you hold it?
Take a real instance. Something went well last month. How long between the moment you knew and the moment you told someone?
If the answer is measured in minutes, and if the good thing did not feel entirely real until it was witnessed, that is worth noticing. Not as a verdict — plenty of people share good news quickly because they are happy — but as a data point about whether your own registration of an event is sufficient on its own.
Then the second question, which goes at the other route: when you achieve something, whose face do you see?
If there is a specific face, and if the achievement is not complete until that face responds, then some part of the accounting is still being done externally. Greene's enmeshment description — the parent living through the child's advancement1 — describes a system where achievements were jointly held, and joint holding does not end when the parent does.
Now the useful part, which is not insight. It is a small deliberate experiment.
Take the next good thing and hold it for forty-eight hours. Do not announce it. Do not mention it in passing so that someone asks. Just have it.
Watch what happens, specifically. There will be a pull — a moment where telling someone is the obvious thing and you have to actively not. Notice what the pull feels like, because that is the clearest look you will ever get at the size of the external dependency. And notice what happens on the other side: whether, by hour thirty, the thing has become more solidly yours or has started to feel unreal.
That is the whole test, and it is repeatable. A working internal regulator makes an unwitnessed good thing feel more yours over time. Its absence makes it evaporate.
And the corrective follows from the mechanism rather than from advice. If the answer is that it evaporates, the thing to build is not confidence. It is the habit of completing things privately — a practice, a project, a piece of work that nobody sees and that you finish anyway. That is the specific room neither the abandoning nor the enmeshing house permitted, and it is the only room the self gets built in.
The base is Kohut-lineage self psychology plus Bowlby's attachment schemas, and it carries the standing 🚩 REPLICATION RISK for this corpus. The specific claim — that abandonment and enmeshment converge on the same outcome — is a clinical-theoretical proposition, not a measured one, and Greene cites no study for it. [POPULAR SOURCE] throughout.
The load-bearing problem is that this framework cannot be wrong. Two opposite parenting failures produce the same result. Extroverts present one way and introverts the opposite way. The confidence is a cover for emptiness, and the self-loathing is also a symptom. There is no childhood and no adult presentation that the model excludes — and a model that predicts everything predicts nothing. That does not make it false; clinical frameworks often outrun their evidence and turn out to be onto something. But it does mean the page cannot be used as a diagnostic instrument, only as a description that might organize what you already know.
Second tension, and it is internal. Greene's Jeanne des Anges case has the appetite for attention arriving first and the parental rejection arriving as a response to it.7 His model requires the parents' failure to be causally prior. He does not notice the conflict. And the alternative reading — that temperament comes first and shapes the parenting it receives — is a serious position in developmental psychology and would substantially reduce the explanatory weight of the two routes.
Third, a practical warning the source does not give. The enmeshment category is unusually easy to weaponize against parents, particularly mothers, and the descriptors — devoted, involved, closely bonded, invested in the child's achievements — describe a great deal of ordinary and successful parenting. The distinguishing feature in Greene's text is specific and narrow: whether the child is left room to establish a self.1 Everything else on the list is present in plenty of households that produce entirely well-regulated adults. Used without that constraint, the category becomes an accusation available against any close family.
Open question: if both routes fail by preventing the same transfer, then the transfer — not the parenting style — is the actual variable. Can it happen later, and from someone other than a parent? Greene says movement up from below the halfway mark is very difficult but not that it is impossible, and the developmental literature on later corrective relationships suggests the window is less absolute than the two-to-five framing implies. Nothing in the source addresses it.
Inner Child and Child Psychology Hub holds this territory at higher resolution and with a different purpose, and the difference is worth stating because it determines what the material is good for.
That corpus is oriented to repair — it treats early patterning as something an adult can work with, and it is correspondingly careful about what is fixed and what is not. Greene's use is oriented to identification: he wants you to recognize the type across the table and act accordingly. The same developmental account, pointed at two different jobs.
And the two uses have different tolerance for error. A repair frame that overreaches costs a person some misdirected effort. An identification frame that overreaches costs somebody else a job, a relationship, or a reputation, on the basis of a childhood the diagnoser never saw. The vault should hold Greene's version as the weaker-warranted of the two, precisely because its consequences fall on third parties.
The convergence is real and worth keeping: both hold that what is transmitted in early attachment is not content but regulatory capacity — how a person handles their own states — and that this is why the effects show up decades later in situations bearing no resemblance to the original.
→ Seduction as Emotional Operant Conditioning
This is the mandatory psychology-to-behavioral-mechanics handshake and it turns on a single detail that Greene's narcissism chapter drops.
His enmeshment description is uniformly smothering. But his Bowlby material describes the enmeshed-ambivalent parent as inconsistent — sometimes suffocating and overinvolved, other times retreating because of their own problems or anxieties.4
That is not smothering. That is an intermittent reinforcement schedule, and the behavioral corpus is unambiguous about what intermittent schedules do: they produce the most persistent, most extinction-resistant behaviour available. Continuous reinforcement is easy to stop. Unpredictable reinforcement is not, because no single absence carries information — the next one might pay.
Apply that to the child and the adult pattern stops needing a separate explanation. Greene notes that children of the enmeshed variety pursue people and then unconsciously retreat,4 and treats it as ambivalence about closeness. Under the conditioning frame it is something more specific: a person trained on a variable schedule, running the pursuit behaviour that the schedule installed, in every subsequent relationship, indefinitely. The retreat is not ambivalence about the person. It is the behaviour of an organism that has learned the reward is unpredictable and that both approaching and withdrawing have historically preceded it.
The insight neither domain reaches alone, and it has real consequences: the abandoning parent and the enmeshing parent are not symmetrical, and the second is stickier. Consistent absence is a stable environment; a child learns from it and generalizes accurately. Inconsistent presence is a slot machine, and the behaviour it produces is designed by the schedule to survive long stretches of no return.
Which explains something the developmental literature struggles with: why adults from the neglect route often know and can name what happened to them, while adults from the enmeshment route frequently cannot — and defend the parent with real warmth. The neglect produced a belief, and beliefs can be examined. The intermittent schedule produced a conditioned behaviour, and conditioned behaviour is not stored as a proposition. There is nothing to disagree with.
→ Attachment Style and Reader Immersion
An unexpected pairing that does real work in both directions.
That page treats attachment patterning as something a narrative can engage — that readers bring a characteristic way of relating to closeness, and that fiction which understands it produces a stronger hold. It is a craft observation about audiences.
Put it beside Greene's two routes and something falls out about why certain artists are made by these childhoods rather than merely damaged by them.
Greene's own case supplies it. Tolstoy, mother dead at two, hole never filled — and Greene notes that he poured this uncertainty into his novels, assuming different roles in the characters he created.5 The absence of a fixed self is described as the pathology. It is also, precisely, the capacity to inhabit anyone, which is the central technical requirement of the novelist.
That is not a romantic claim about suffering producing art, which is usually false and always cheap. It is a specific structural claim: the missing organ and the professional skill are the same absence, evaluated by different criteria. A person with no stable self-image cannot rest, and can also enter a stranger's interior without resistance — because there is nothing of their own in the way.
The craft page's finding sharpens it further. If readers bring their own attachment patterning to the encounter, and if the writer's own patterning is what let them build the interior in the first place, then a novel is a transaction between two regulatory systems, neither party aware of it. Which suggests why some books land catastrophically hard on some readers and slide off others with no relation to quality.
And it puts a real cost on the finding. Greene's chapter treats the deep narcissist as someone to be avoided. The same chapter's most celebrated example wrote Anna Karenina, and made his wife's life a forty-eight-year ordeal. Both are true, and the vault should not resolve them: the capacity that produced the work is not separable from the deficit that produced the marriage.
Sharpest implication: the enmeshed child gets more attention than almost anyone, and arrives at adulthood unable to generate a single unit of it internally — because the transfer requires a gap, and constant attention never leaves one. Which means the parental behaviour that reads unambiguously as love is producing the same outcome as the behaviour that reads unambiguously as neglect, and only one of them will ever be noticed by anyone outside the house.
Generative questions: