Haute Lumière · The reading

The stage you are leaving always feels like failure

Every model in the canon describes the same interval between one way of making sense and the next, and the interval is the part nobody names.

She has read this page before. What changed between the readings is not the page.

THE FEELING FIRST

It usually begins as a loss of interest in an answer that used to work. The account a person gives of their own life — why the work matters, why the household is shaped the way it is, why the ambition sits exactly where it sits — goes on being true and stops being interesting. Nothing has broken. The days still run, the invoices still go out, the people are still loved. What has gone is the sense that the explanation is explaining anything.

The first reading available for this is decline. A person reaches for exhaustion, for a dulled edge, for the idea that the appetite of the twenties does not survive contact with the forties. These readings are near at hand, they cost nothing, and they have the considerable advantage of being partly true, which is what makes them so hard to put down. Something has in fact been lost. The question is what kind of thing it was.

The developmental literature offers a second reading and it is stranger. It proposes that a person does not only accumulate knowledge and skill across a life but periodically changes the structure the knowledge is held in. The structure is not a belief. It is closer to the shape of the container: what counts as a reason, what counts as a self, what can be held at arm's length and what is still being stood inside of. When the container changes, everything in it is re-sorted, and for a while nothing is where it was left.

This is why the interval reads as loss rather than gain. The only instruments a person has for measuring their own capability are the ones the old structure issued. Measured by those instruments — decisiveness, certainty, the speed with which an opinion arrives — someone in the middle of a transition is genuinely performing worse. They are slower. They hedge. They find both sides of a thing they used to settle in one sentence and a glance. The instruments are not malfunctioning. They are measuring the wrong era.

The old structure cannot measure the new one. It can only report that something has gone missing, and on its own terms it is right.

A page like this is not the place to settle whether a particular person is tired, unwell, badly employed, badly matched, or developing. Those are separate questions with separate evidence and different people to ask. What a canon of stage models can do — the reason it is worth owning at all — is supply a description under which the interval is not a symptom. That description is not consolation. It is a different set of things to look at, and it changes what a person does next.

What follows is the argument itself: what a stage actually is, how it differs from an experience, why development runs unevenly, what the transition costs, where the models are weakest, how they get misused, and what a person can do with one that is more useful than being ranked by it. The claim is not that these models are true the way a measurement is true. The claim is that they are the best available description of something most people pass through several times and almost nobody has language for.

STAGE, DEFINED

A stage is not an achievement, a personality type, or an age. It is not how much a person knows, how kind they are, or how well they are doing. Those things vary enormously within any stage and are mostly better explained by other means. A stage, in the sense the models use, is the organising structure a person makes meaning with — and the whole difficulty of the subject is that this structure is not visible to the person using it, because it is the thing they are using to look.

The cleanest way in is the difference between what a person has and what a person is. A belief that is had can be examined, argued with, set down, taken up again. A belief that is inhabited cannot, because there is no vantage point outside it from which to see it as a belief at all. It does not feel like a position. It feels like the way things are. Everyone has a stock of both, and the proportion between them is close to what the word stage is pointing at.

Development, then, is a specific and repeatable move: something a person was looking through becomes something they can look at. A young child inside a rule cannot ask what the rule is for; the rule is simply how the world is arranged. Later the rule becomes an object in the world, and the question of its purpose becomes askable — which is not rebellion, it is a new organ. The adult versions are less obvious and more consequential. Being one's role, and having a role. Being one's opinions, and holding them.

This is why a stage cannot be reached by study. The vocabulary of any developmental frame is public and easy, and a person can acquire the sentences of a structure they are not living in about as fast as they can read them. Fluency in the language of nuance is not the same as the capacity to hold two true things that contradict each other and act anyway. The second one costs years. The first one costs an afternoon.

Anyone can borrow the sentences of a stage they have not reached. This is the most common event in the field, and the reason the field is so hard to measure.

Which gives the only reliable test. A structure shows itself under load, never in conversation. What a person does when they are contradicted in public, when the plan fails at the worst hour, when someone they need is disappointed in them — that is the structure reporting itself. In comfort, almost everybody performs their best hour. In pressure, everybody performs their actual one, and the gap between those two is the honest subject matter of this whole field.

It follows that a person cannot see their own current stage directly, and no amount of introspection will fix that, because introspection is being conducted by the very thing in question. What a person can see is the one they have left. This is not a limitation of the models. It is a structural feature of being a creature that thinks with its own frame, and it sets the terms for every practical use the rest of this reading will make.

STATE AND STAGE

A person comes back from a week of silence, or a night of grief, or an hour on a hillside that rearranged the furniture, and reports that everything is different now. Six weeks later the same person is having the same argument with the same sibling about the same unpaid thing, and the conclusion they draw is that they failed, or that the experience was not real. Both conclusions are wrong, and the confusion behind them is the most expensive one in the whole subject.

A state is an experience with a beginning and an end. A stage is a structure with no edges a person can feel. States are available at any stage and always have been: absorption, awe, terrible clarity, the collapse of the ordinary sense of being a separate item in a world. They are among the most valuable hours a life contains and nothing here diminishes them. But they pass, by definition, and what receives them is whatever structure was already in place.

Which is why two people can have the experience that appears, from the outside and even from their own descriptions, to be identical, and make completely different meanings of it. One takes it as evidence of being chosen. One takes it as evidence that the self is less solid than assumed. One folds it into an existing system of belief without disturbing a single item. The interpreting is not done by the experience. It is done by the structure the experience arrived into, and the structure does not change just because something remarkable happened inside it.

The disappointment that follows is therefore not a sign that nothing occurred. It is a sign that the experience was not built into anything. A state can be had. A structure has to be constructed, and it is constructed by repetition under conditions, over a span of time nobody finds flattering. The gap between the week and the six weeks is exactly the size of the work that was never described to the person on the way out of the retreat.

An experience arrives in an hour. A structure is built in a decade. Confusing the two is how a real hour gets thrown away as a failure.

None of which makes states useless to development — they contribute two things reliably. They demonstrate that a way of being is possible, which is not a small matter, because a great deal of what holds a structure in place is the unexamined certainty that there is nothing else. And they loosen. In the days after a strong state, a person's grip on their usual account tends to be lighter than normal, and things can be set down that were previously not even visible as things being held.

So the practical instruction is unglamorous and it is the same in every register. Treat the state as an opening rather than a result. In the loosened weeks, change something structural: a commitment, a practice held at a fixed hour, a relationship to a person who will not let the old account run unchallenged. The hour on the hillside is not the development. It is the window in which development is briefly cheap, and windows close.

Second week, mid-morning. The question she arrived with has quietly been replaced by a better one, and she has not noticed yet.

LINES, NOT LADDERS

The single most damaging simplification in this field is the one ladder. It produces the sentence that a person is at level five, and once that sentence exists everything downstream of it is wrong: the self-assessment, the ranking of colleagues, the private arithmetic of who is ahead. No serious reading of the evidence supports one ladder. What the evidence supports is many, running at different speeds in the same person, and almost never in step.

Call them lines. Cognitive complexity is one. The capacity to hold another person's interior as genuinely other is another. Emotional range, ethical reasoning, aesthetic discrimination, the body's own intelligence, the plain competence of handling money — each develops, each has recognisable earlier and later forms, and each can be strikingly advanced in a person whose neighbouring line has barely moved since school. This is not the exception. In the population of actual adults it is close to universal.

Everybody already knows the examples, which is a good sign that the account is describing something real. The strategist who can hold six moving variables at work and cannot run one argument at home without it becoming about worth. The person of extraordinary emotional literacy whose relationship to money is still entirely magical. The scholar of the subtlest possible arguments who cannot hear a plain complaint from someone who loves them. None of these is hypocrisy. It is unevenness, and unevenness is the normal shape of a human being.

Two further terms make the picture usable. The first is a centre of gravity: where a person mostly operates, across most lines, on a normal day. The second is a peak: the latest structure they can reach when conditions are good, which is usually one step past the centre and much less reliable. Development is largely the slow migration of the centre toward where the peak already occasionally goes, and it is measured in years, not in insights.

Nobody is at a stage. A person is spread across a dozen of them, and where they live is not where they can visit.

The practical value of dropping the ladder is immediate. A level is not actionable — there is nothing a person can do on Tuesday about being a four. A line is entirely actionable, because a line has content: this specific incapacity, in this specific domain, with a next step that can be named and practised. Choosing the line is the whole of the useful work, and choosing it well usually means choosing the one a person is most reluctant to look at rather than the one where progress would be most pleasant.

It also changes how a person reads other people, which is where the ladder does its real damage. Once unevenness is assumed, the question stops being who is ahead and becomes which line is carrying this situation. A room does not need everybody to be late-stage. It needs the line the decision depends on to be well developed in somebody the room will actually listen to, and it needs that person to be available at the hour the decision gets made.

THE TRANSITION COST

There are two kinds of change and the models are only about one of them. The first is change within a structure: acquiring, improving, correcting, revising. A person learns a language, changes their mind about a policy, gets better at their craft, adopts a new opinion about their father. All of this can be substantial and most of a good life is made of it. Through all of it the structure doing the revising is not itself revised. It simply has more and better furniture.

The second kind is change of the structure, and it is rare, expensive, and cannot be scheduled. It does not feel like learning. It feels like the floor of an argument giving way — not being persuaded of a different answer, but losing confidence in the question. A person in it often cannot say what has happened, which is why the accounts they give tend to be vague in a way that invites others to diagnose them, and the diagnoses are usually about mood because mood is the visible part.

The costs should be named plainly rather than softened, because people are entitled to know what they are choosing. A structure is held in place by relationships, and relationships built on the old one will strain. Some friendships were agreements about what the world is, and when one party stops agreeing the friendship discovers what else it was made of. A career can be organised around a competence that the new structure finds less interesting. A marriage can be organised around an arrangement that is no longer the arrangement either person is in.

This is the honest reason so many transitions stall, and stalling deserves respect rather than contempt. A person with three dependents, a mortgage, and no margin is making a defensible decision when they decline to dismantle the structure currently paying for all of it. Nobody is obliged to develop on anyone else's timetable, and a great deal of the coaching industry's damage comes from pushing people through an interval they had good reasons not to enter this year.

A structure is not held in place by conviction. It is held in place by everything that was built on top of it.

What a model contributes here is a shape for the middle. The old fluency is gone and the new one has not arrived, and the period between them is not a few weeks. It has a described form: an early relief at being free of something, a long flat stretch where nothing works and the person suspects they have simply broken themselves, and then a slow return of capability in a form that is not recognisable as the old one and is easy to miss on arrival.

Knowing the shape does not shorten it. What it does is prevent the specific error that makes it much longer, which is treating the flat stretch as a verdict and reconstructing the old structure in a hurry to escape it. Nearly everyone does this at least once. It works, in the sense that the floor comes back. And then the same interval opens again, three or four years later, in the same shape, and asks the same thing, usually at a worse moment.

WHAT MODELS MISS

Start with what holds, because it is easy to lose in the criticism. These models earn their keep by predicting the shape of an experience that people report without having read anything: the invisible frame becoming visible, the costly interval, the unevenness, the way an earlier way of making sense stops being available as a way to live rather than merely being disagreed with. That is a real pattern, reported across very different lives, and any account of adulthood that has no room for it is missing something obvious.

The first limit is that what gets measured is almost always language. A person is interviewed, or writes, or reasons aloud, and the structure is inferred from the complexity and the self-relation of what comes out. But verbal fluency is its own line, developed unevenly like every other. Someone who thinks in their hands, or in music, or in the practical arrangement of a room full of people, can be systematically under-read by an instrument that only listens. The instrument is not neutral about who it can see.

The second limit is that the tasks carry the assumptions of the people who built them. A frame that scores growing independence from the group as later development is saying something about what it values before it measures anything, and it will read a life organised around obligation as less developed rather than differently arranged. This is not a reason to discard the models. It is a reason to hold their scoring lightly and their descriptions more firmly, since the descriptions travel much better than the rankings do.

The third limit is the one people notice least. A model can establish that a sequence occurs — that this typically comes after that, and not the reverse — without establishing that later is better. Sequence is an empirical claim. Better is a value claim, and it has to be argued in the open rather than smuggled in on the back of a diagram whose arrow happens to point upward. A great deal of nonsense in this field is the result of reading a direction as a virtue.

Later is a claim about order. Better is a claim about worth. A diagram cannot get you from one to the other.

What survives all three criticisms is worth stating exactly, because it is what the rest of the reading rests on. Something like irreversibility of sight: once a person can see a frame as a frame, unseeing it is difficult and mostly a matter of pretending. Something like sequence: the capacity to hold a system as an object does not reliably arrive before the capacity to hold a rule as one. And the description of transition, which is the most useful and least contested part of the whole apparatus.

The honest position, then, is to use these models as descriptions rather than as instruments of rank. A map that says people in this passage often report the following is worth a great deal to a person in that passage. A map that says you are a four is doing a different job, usually a social one, and it should be asked what it is for and who benefits from the answer.

Nothing here is being resolved. The work at this hour is staying in the room with something that has not yet made sense.

THE HIERARCHY PROBLEM

Every field has a characteristic failure and this one's is contempt. It arrives quietly and it is always available, because the vocabulary is built for it. A person learns a set of stages, locates themselves flatteringly within it, and acquires a device for converting every disagreement into a diagnosis of whoever is disagreeing. The device cannot be refuted from inside, since objecting to it is exactly what the device predicts a person at an earlier stage would do.

It is worth being precise about why this is intellectually dishonest rather than merely rude. An unfalsifiable explanation that flatters its holder and discredits their opponents is not a theory in use; it is a status move wearing a theory's clothes. The test is simple and rarely applied: what would it look like if the other person were right and the frame were being misused. If no answer to that question is available, the frame has stopped doing cognitive work and started doing social work.

The confusion underneath is between capability and worth. It may well be that a later structure can hold more complexity, coordinate more perspectives, and act with more of the situation in view. None of that makes the person holding it more valuable, more entitled, or more correct about any particular matter in front of them. A model that quietly converts capability into worth has become a caste system with a reading list, and the people it flatters are the last ones positioned to notice.

Contempt for an earlier stage is the most reliable evidence of an unfinished one.

There is a second-order test that works well in practice. A person genuinely later in a sequence is usually more able to inhabit an earlier one, not less — they can speak in its terms, feel its force, and remember what was true about living there, because the earlier structure is included in what they now hold rather than discarded. The one who cannot bear the earlier stage is generally not past it. They are recently out of it and defending the exit, which is a different posture with a different smell.

The practical damage is not confined to manners. A leader who privately ranks a team by developmental stage will manage them worse, because the ranking supplies a ready explanation for every failure of communication that never once implicates the leader's own account of the situation. Feedback stops arriving. The most useful person in the room becomes the one who is most fluent in the leader's frame, which is a selection process for agreement rather than for sight.

The discipline that answers this is small and constant: use the model on yourself in the past tense and on nobody else in the present. Applied backward to one's own record it is a genuine instrument, because the evidence is available and the flattery is limited. Applied forward to a colleague it is a guess dressed as a measurement, made by someone who cannot see their own frame, about a person whose interior they do not have access to.

REGRESSION UNDER LOAD

Everyone has watched themselves do it. A person who has spent years becoming able to hold a difficult conversation without making it about their own worth is on hour nineteen of a bad day, and the whole apparatus is simply not there. What comes out of them is something they recognise from fifteen years ago, delivered with total conviction, and the recognition arrives about forty minutes too late to be useful. The question is what actually happened, and the usual answer, that they went backwards, is not quite right.

Capability is conditional on load, and later structures are the expensive ones to run. Holding several perspectives at once, staying in contact with a person who is angry without either collapsing or hardening, remaining uncertain in public while still acting — all of this costs energy, attention, and a margin of safety. When those are gone, the cheaper structure runs, because the cheaper structure is still there. It was never demolished. It was built over, and the earlier floors keep their lights on.

This means the question of which stage a person is at has no single answer, and the honest reply is always a distribution. There is the structure available on a good morning, rested, with money in the account and nobody disappointed in them. There is the structure available at two in the morning with a sick child, a deadline, and an unanswered message from someone whose regard they need. Both are the same person. Only one of them is going to be in the room when it matters most.

The later structure is the expensive one. When the margin goes, the cheap one runs, and it runs with total sincerity.

Which quietly rewrites what practice is for. The ambition to reach higher is mostly the wrong ambition, and it is the one the market sells because it is the one that can be sold. The better ambition is to widen the band of conditions under which the capability a person already has remains available to them. That is a question about sleep, money, quiet, the hour of the day a hard conversation is scheduled, and the number of people who can tell the truth to them without consequence.

It also makes a demand of institutions, and the demand is not soft. An organisation that keeps every person at the outer edge of their capacity will receive the least developed available version of each of them, regardless of what any of them would test at on a calm Tuesday. The meeting where a company makes its worst decisions is usually not staffed by its least developed people. It is staffed by its most depleted ones, and depletion is the variable nobody puts on the agenda.

For an individual the same logic yields an unglamorous but genuinely powerful move: identify the two or three conditions under which the expensive structure reliably goes offline, and treat those conditions as the object of work rather than treating the resulting behaviour as a character flaw. Most of what people try to fix through insight is not an insight problem. It is a margin problem wearing an insight problem's clothes, and margin responds to arrangement in a way that character does not.

USING THE MAP

Three uses hold up under pressure, and it is worth being strict about them, because almost everything else these models get used for is either social positioning or a way of postponing an ordinary decision. The three are: recognising a transition from inside it, choosing the next constraint to work on, and designing the conditions under which development is even possible. Each is concrete. None requires a person to know what stage they are at, which is fortunate, since they cannot.

The first use is the one that pays immediately. Being able to say that this flat, hedging, slow period has a described shape and a described end converts an interior verdict into an exterior situation. Nothing about the difficulty changes. What changes is the second layer — the self-assessment running underneath the difficulty, quietly concluding that the person has lost whatever it was they had. That second layer is where most of the suffering in a transition actually lives, and it is the layer a description can reach.

The second use requires a specific question, and it is not a comfortable one. What is currently being looked through rather than looked at. In practice: find the thing that, when questioned, produces not an argument but a flash of heat, or a change of subject, or the certainty that the question is silly. That flash is the most accurate instrument available to an adult about their own frame, and it is free. What follows it is not insight but practice, because the point is not to have the thought once.

The third use is about conditions, and it is the one most often skipped because it is administrative rather than interesting. Development appears to need a particular ratio: enough challenge that the current structure cannot quite cope, and enough support that failing to cope is survivable. Too much support and nothing moves, because the structure is never asked for anything it cannot already do. Too much challenge and the person defends, which looks like resistance and is actually an accurate assessment of risk.

Challenge without support produces defence. Support without challenge produces comfort. Development happens in the narrow place where both are true.

It is equally worth naming what these models are not for, since the misuses are so predictable that they can be listed in advance. They are not for sorting colleagues. They are not for deciding who to be with. They are not for hiring. They are not for explaining why a conflict is the other person's developmental problem, which is the single most common use in the wild and the one that most reliably indicates that the user has stopped thinking. Each of those uses substitutes a category for a person.

And a last practical note about pace. Nothing in this material moves on the timescale a reader would like. A frame that has organised a person's sense of themselves for twenty years does not become an object over a weekend, and the attempt to force it produces a convincing imitation rather than the thing. What can be done in a week is to notice the flash of heat, to arrange one condition better, and to stop treating a described interval as a private failure. That is a real week's work and it compounds.

READING THE CANON

The reason to read a canon rather than a model is that a single model is a single instrument with a single set of blind spots, and from inside it those blind spots are indistinguishable from the shape of reality. Read three that were built for different purposes and the agreements between them become visible, which is the nearest thing to triangulation this subject offers. Where several instruments with different assumptions converge, something is probably there. Where only one reports a feature, the feature may belong to the instrument.

The agreements are worth knowing in advance, because they tell a reader what to watch for. That there is a sequence rather than a random walk. That the direction runs from being embedded in something to being able to hold it as an object. That the passage between two structures is costly and has a middle in which the person functions worse. That development is uneven across domains within the same person. That the structure in use is not visible to the person using it. These recur across frames that agree about very little else.

The disagreements are equally worth knowing, since they are where a reader's judgement is actually required. How many stages there are, which is largely a question about how finely the instrument cuts. Whether there is a top, which is a question most frames answer badly. Whether the sequence is universal or an artefact of the setting that produced it. What exactly is being tracked — reasoning, the sense of self, the relation to other people, or several of these braided together and reported as one number.

The way to read them is against one's own record rather than against one's self-image, because the record is evidence and the self-image is a product of the very structure under examination. Take a specific decision from ten or fifteen years ago — a job taken, a person left, a line held — and reconstruct not what was decided but what could not be questioned at the time. What was simply the way things were. That list is a portrait of a structure, and it is available to a person now in a way the present one never is.

The past is the only place a person can see their own development directly, because the present frame is doing the looking.

Done honestly, the exercise produces two results, and the second one is the valuable one. The first is a sense of how far a person has travelled, which is pleasant and largely useless. The second is the realisation that there is, right now, a list of exactly that kind — things currently invisible as assumptions, which in ten years will be obvious and slightly embarrassing. That realisation is not humility as a posture. It is an accurate description of the situation, and it is the one thing that keeps a developmental frame from turning into contempt.

This volume sits on the Developmental Canon shelf, and the reading of it is free, in full, with nothing held back. The argument above is what it is about. What the volume does that a page cannot is set the models side by side with their apparatus intact — the case material, the reflection prompts, the visual summaries, and a bibliography kept whole rather than trimmed to fit — so that a reader can work with them rather than be told about them.

Free to read

Free to read, and free to hear. Every chapter of every book in this house, and every narration of it, is open to anybody. No account, no card, nothing to cancel.

What is in it

• The framework skeletonopen this in the house search

A stage is not what a person knows. It is what they cannot yet see themselves using.

Every model of growth is a map of losses that somebody survived and then described.

The one who cannot bear an earlier stage is usually standing just outside it, defending the exit.

What is invisible today will be obvious in ten years, and slightly embarrassing. That is the whole subject.

Keep looking

Every phrase on this page opens into the house search. The shelf holds The Developmental Canon and six other shelves, and the reading is free.