Haute Lumière · The reading

The plan was good and the room was furious

When a decision lands badly it is rarely a failure of strategy; it is a failure of who was visible when the strategy was drawn.

She is reading the list of names a second time. The second reading is the one that finds who is not on it.

THE ROOM

A regional hospital system receives forty million dollars to redesign its community health services. Leadership assembles a planning committee of twelve people: the chief executive, the finance chief, the chief medical officer, three department heads, two board members, an external consultant, three senior administrators. They work for four months. What they produce is genuinely good — charts, projections, a plan with internal logic and real ambition inside it. By every measure the committee has available, the work is finished and the work is strong.

Then they present it to the community, and the room turns. The health workers who actually deliver services in the neighbourhoods the grant names were not consulted. The patient advocacy groups whose members will feel the redesign first were not invited. The local Indigenous health council, whose ancestral lands the hospital stands on, learned about the plan from a press release. The facilities and janitorial staff, whose working conditions the plan would alter substantially, heard about it in an all-hands meeting where questions were welcomed and time-limited.

Nothing in that sequence is a failure of strategy. The plan was well made. Nothing in it is a failure of intention either — the leadership team sincerely wanted better community health, and had gone to considerable trouble to get the money that would produce it. What failed was sight. The stakeholder map was drawn from inside a boardroom, and a boardroom can only see the kinds of power that boardrooms are built to register.

The people a stakeholder map fails to show are almost always the ones with the least institutional standing and the most at stake.

This is the ordinary condition of modern organisational life rather than an unusual accident. Decisions are made that move dense webs of people while only a fraction of the web is visible to the people deciding. The fraction that is visible is not random. It is precisely the fraction that shares the decision-makers' vocabulary, occupies their calendar, and appears on their org chart, which is to say the fraction least likely to be surprised by anything.

And it is worth naming what kind of problem this is, because the naming determines the remedy. It is not a communications problem, so more announcements will not touch it. It is not a goodwill problem, so sincerity will not touch it either. It is a perceptual problem — a limit on what a particular vantage point can register — and the only things that shift a perceptual limit are new dimensions of looking.

WHO COUNTS

For most of the twentieth century the dominant answer to who matters was short. Shareholder primacy, articulated most famously by Milton Friedman in 1970, held that the social responsibility of a business is to increase its profits. Everyone else — employees, customers, communities, the living world — mattered only through the bottom line, as a variable that could raise or lower it. The answer had the great advantage of being computable.

It was never merely a theory. It was an operating system that shaped how whole generations of leaders were trained to think. Business schools taught it. Legal frameworks encoded it. Compensation structures rewarded it and kept rewarding it long after the argument itself had grown embarrassing. When an idea is wired into pay, it does not need to be believed to be obeyed.

The stakeholder revolution that answered it arrived most prominently through R. Edward Freeman's Strategic Management: A Stakeholder Approach in 1984, and proposed something that was radical for its moment: that organisations carry obligations to every party who affects or is affected by what they do. Not shareholders alone, but employees, customers, suppliers, communities, regulators, the broader public. It gave language and legitimacy to what many practitioners had already sensed in their hands — that treating people as instruments of shareholder value was both ethically thin and strategically short.

The evidence accumulated behind it. Organisations attending to a broader field of stakeholder interest often outperformed those fixed exclusively on shareholder return — not always, not immediately, but consistently enough to put real pressure on the Friedman orthodoxy. That is an enormous step and it deserves to be honoured as one. Most of the good practice in the field descends directly from it.

An idea wired into compensation does not have to be believed to be obeyed.

What follows is not a refutation. It is a deepening, and the distinction matters. Conventional stakeholder theory established that organisations exist inside webs of mutual obligation, which is the foundational insight and remains true. The question this book takes up is what that theory still cannot see — and the honest answer is four things, each of which shows up in the Meridian Health room.

FOUR LIMITS

The first limit is motivational. Stakeholder theory in practice usually argues that attending to stakeholders is good for the bottom line. This happens to be true, and it is also a trap, because it keeps profit at the centre and turns stakeholder care into a technique for reaching it. The stakeholders remain a means to an end — a more sophisticated means than Friedman imagined, but a means. People notice when they are being attended to instrumentally, and they discount everything that follows.

The second limit is perceptual, and it is the one that produced the Meridian room. Stakeholder maps are drawn by people, and people map who is already visible, already vocal, already institutionally recognised. The community health workers were not omitted out of malice; they were omitted because the map was drawn by a vantage point that had never had cause to see them. Position determines field of view, and field of view determines the map, and the map is then treated as the territory.

The third limit is architectural. Much of what is called stakeholder engagement is, in practice, informing people of decisions already taken, or collecting input that may or may not alter anything. Genuine co-creation — where people help frame the questions rather than answer someone else's — remains rare, because most engagement processes are designed to manage concern rather than to be changed by it. The architecture announces its own intentions long before anyone reads the report.

The fourth limit is the one nobody names, and it is developmental. Conventional theory treats every stakeholder as though they were running the same meaning-making system, differing only in interest. But a community elder who makes meaning through devotion to tradition, a social entrepreneur working from strategic innovation, and a systems thinker perceiving through complexity are not holding different opinions about alignment. They are holding different definitions of what alignment is.

Ignoring developmental difference does not remove it from the room. It removes your ability to see it working.

Each of these four limits has the same shape: something real is operating in the room, and the instrument in use has no channel for it. The remedy is not more rigour applied to the existing instrument. Adding detail to a map that has the wrong number of dimensions produces a more confident wrong map. The remedy is more dimensions.

Nothing is being decided here, which is exactly why it is possible to notice what the last meeting decided without asking.

INVISIBLE CAPITAL

Start where conventional analysis starts, because it is not wrong, only partial. Understanding who holds institutional power and who has direct interest in the outcome is genuinely necessary work. The classic power-interest matrix still earns its place: high power and high interest are the key players who must be closely engaged; high power and low interest are actors who could be triggered into involvement; low power and high interest are passionate communities with little institutional leverage; low power and low interest are peripheral actors who may nonetheless carry consequences.

The correction is to name what the word power is measuring when it is placed on that axis. As conventionally assessed, it means institutional, economic and political capital — budget authority, formal position, legal standing, access to press. Those are real forms of power. They are not the only ones, and a matrix that admits only those will systematically underweight the three that matter most in exactly the situations where engagement decides the outcome.

Relational capital is the trust and connection held inside a community, built slowly and impossible to purchase. Wisdom capital is the deep working knowledge carried by elders, by Indigenous peoples, by long-tenured staff who know which procedure has failed twice before and why. Moral capital is the ethical standing that comes from lived experience of the issue under discussion — the authority a person has because the decision will land on their own life.

Run Meridian Health through this correction and the room becomes explicable. The community health workers had almost no institutional power, so the conventional map placed them in keep informed. Their relational capital was decades of trust with exactly the communities the grant was written for, and it was not replaceable by anyone in the boardroom. Their wisdom capital was intimate knowledge of what actually works on those streets. Their moral capital was the standing that comes from doing the hardest work for the least recognition, and it was precisely the standing the plan needed in order to be credible when announced.

Power measured only as budget and title will always misfile the people whose consent the plan depends on.

So the first dimension keeps the matrix and adds one question to it, asked out loud and recorded: whose power is invisible to the current way of seeing, and which forms of capital is this process failing to register. Asked honestly, it moves names across the grid, and the names it moves are usually the ones whose absence would have detonated in the room.

THE WEB

Most stakeholder maps draw people as separate bodies orbiting the organisation, planets around a sun. The picture is wrong in a way that matters, because it implies that each relationship can be managed independently of the others. In reality stakeholders sit inside dense webs of mutual influence, and the second dimension of mapping is the relationships rather than the parties. Pull one thread and the whole web moves, usually somewhere nobody was watching.

The first question this dimension asks is which stakeholders influence each other. A decision that satisfies the board and alienates frontline staff does not stay contained in staff morale. It arrives in patient care, then in community trust, then in reputation, and finally returns to the board as a declining metric that nobody connects to the original decision because eighteen months have passed. Tracing those causal chains before deciding is the whole difference between alignment and perpetual firefighting.

The second question is where the feedback loops run. Some stakeholder relationships reinforce: community trust rises, engagement rises, outcomes improve, trust rises further, and a small early investment compounds into something that looks like luck. Others balance: one department gains resource, another loses it, and the tension must be held rather than solved. Mapping which is which tells you where a modest intervention will produce a disproportionate effect, and where effort will simply be absorbed.

The third question is the one most often skipped, and it is about time. Quarterly earnings pressure runs on ninety days. Community trust runs on a generation. Ecological consequence runs on a civilisational scale. Regulatory cycles, union contracts, electoral terms and clinical outcomes all run on their own clocks. A great deal of what gets described as stakeholder conflict is not disagreement about the good at all; it is two parties applying different discount rates to the same future and each finding the other irrational.

Much of what looks like conflict is two clocks, running honestly, at different speeds.

Practitioners often stall here, because the web looks infinite and the mapping looks like a research project nobody has time for. It is not necessary to map everything. It is necessary to map enough to catch the dynamics the current process is structurally blind to, and a rough interdependence sketch made on a whiteboard in thirty minutes routinely surfaces what a polished stakeholder list of forty names does not. The value is in the lines, and the lines take minutes.

FOUR MINDS

The third dimension asks how different stakeholders are making meaning of this situation, and it is the one that makes this framework distinct from anything in conventional practice. Drawing on developmental psychology — Robert Kegan, Clare Graves, Susanne Cook-Greuter — it holds that people at different developmental centres of gravity do not merely hold different views about a change. They construct the change itself differently, and therefore find different things obvious, threatening and worth defending.

Take a hospital installing a new electronic health records system, which is a dull enough example to keep the point clean. Someone centred in Blue — order, structure, devotion to established practice — experiences the change as a threat to systems that have worked for decades. That is not irrational. It is respect for institutional wisdom and a legitimate fear that what was built carefully will be dismantled carelessly. Honouring it means making room for grief about what is ending, not only enthusiasm about what is starting.

Someone centred in Orange — achievement, innovation, strategic optimisation — sees an opportunity to reduce error and gain advantage, and their enthusiasm is real and useful. It can also blind them to the human cost of rapid change and to the concerns of people who experience transition differently. Someone centred in Green — inclusion, empathy, relational sensitivity — will attend first to the most vulnerable: patients with limited technical literacy, older staff struggling to adapt, communities whose data privacy is genuinely at risk. That attention is essential, and it can also slow something that needs to move.

Someone centred in Yellow — systems thinking, adaptive flexibility — may perceive how each of these is partially right and attempt an implementation that honours all of them. This integrative capacity is valuable and it carries its own characteristic failure, which is a quiet superiority that dismisses the urgency other people are actually feeling. Each position sees something real. Each position is blind to something the others hold, which is the only reason the others are necessary.

The working question is therefore never which perspective is most evolved. It is how to build processes capable of holding all of them at once, drawing on what each contributes while preventing any single one from running the room. That is developmental facilitation applied to organisational life, and it requires someone — an individual, or a capacity distributed through a team — who can read developmental difference without ranking it.

The moment you say that is just resistance, you have stopped mapping and started disposing of people.

One caution, and it is not decorative. This dimension is a tool for understanding, never a labelling system. If it is being used to categorise colleagues, to explain away objections, or to feel advanced relative to stakeholders at earlier stages, it has been turned into a weapon and it will do more damage than having no framework at all. Developmental awareness that does not increase humility has been misunderstood at the root.

THE EMPTY CHAIR

The fourth dimension is the most frequently skipped and the most consequential. It asks who is affected by this decision and absent from the conversation. Every stakeholder map has an edge, and beyond that edge sit affected parties who have become invisible — future generations who inherit what is decided today, ecosystems with no standing in human deliberation, communities in distant geographies whose labour and resources feed the system, people inside the organisation with no positional authority to be heard in a formal process.

The move that makes this dimension work is to treat absence as information rather than as an oversight to be corrected. When a map is missing voices, the correction is not to add names and continue. It is to ask why those parties were absent, which structural dynamics rendered them invisible, and what the decision would look like if their interests were genuinely at the centre rather than appended to it. The answers are uncomfortable, because they describe how the institution is built.

At Meridian Health, that question surfaced something nobody had intended and nobody could defend: the strategic planning process had never once included a representative of the communities with the worst health outcomes — the exact communities the forty million dollars was meant to serve. It was not malice. It was architectural blindness. A process designed by and for people with institutional standing cannot see those without it, and it will reproduce that blindness every cycle until the architecture changes.

Four practices make absence visible in ordinary rooms. The empty chair: place a real, unoccupied chair at the table for every significant decision, and rotate the responsibility for speaking from the position of whoever is most affected and least represented. It sounds like theatre for about four minutes, and then it changes what the group can perceive. Reverse mapping: instead of working outward from the organisation and asking who needs engaging, work inward from impact and ask who is most affected and how far they sit from the decision. The gap between most affected and closest to the decision is a precise measurement of an ethical blind spot.

Generational impact assessment: for significant decisions, ask explicitly how this lands on people in seven years, twenty-five years, a hundred. There is nothing mystical in it — it is long-horizon responsibility, practised for millennia in Indigenous governance traditions such as the Haudenosaunee Seventh Generation principle, which asks leaders to weigh decisions against seven generations to come. Modern organisations are only beginning to rediscover what is practical about it. And ecological inclusion: the natural systems that sustain human life are affected by nearly every organisational decision and appear on almost no stakeholder maps. What would a supply chain decision look like if the watershed had a seat. What would facility planning become if old-growth forest had representation.

Absence is not an accident in the room. It is a design, and a design can be read.

Attention held this long is a form of accounting. It costs something, and it costs less than the room that turns.

THE BODY KNOWS

Conventional stakeholder theory ignores the body entirely, and this is a strange omission given how much of the relevant information arrives that way. When you are in a room with people whose needs you genuinely hold in view, something measurable happens in your nervous system. Breathing deepens. Peripheral awareness widens. Tension, alignment and disconnection register through posture and field quality well before anyone says a sentence that could be quoted in minutes.

The inverse signal is quieter and more useful. When significant parties are absent or unseen, there is often a tightness, a flatness, a sense that something is missing that cannot immediately be named. Most experienced facilitators know the feeling — we are deciding something here, and something is not right. That sensation is not anxiety to be managed. It is frequently the most accurate available report that the stakeholder map is incomplete.

This is not mysticism, and it is worth being precise about why. Stephen Porges' work on the autonomic nervous system establishes that humans are exquisitely attuned to social safety and threat, largely below the threshold of conscious awareness. A genuinely inclusive process tends toward ventral vagal activation: openness, curiosity, real social engagement, the state in which disagreement is survivable. A process where voices are excluded or suppressed tends toward sympathetic activation — tension, defensiveness, manufactured urgency — or toward dorsal collapse, which shows up as disconnection, numbness and people going through the motions with perfect attendance.

Read that way, the somatic signal is not soft data. It is a fast, pre-verbal read of the relational field, delivered before the analysis finishes, and it is often right about something the analysis is not equipped to see. The appropriate response is neither to obey it nor to dismiss it, but to treat it as a prompt for a specific question: who is not here who should be, which perspective is not being held, what impact is going unseen.

When the room goes flat and nobody has said anything wrong, something true is being left out.

The capacity to read field dynamics is a skill rather than a gift, and it develops the way other skills develop — through repetition, attention and correction. Notice the state of your own breathing in the first two minutes of a decision meeting. Notice when it changes and what was said immediately before. Over a few months this becomes an instrument, and it is an instrument that reports on exactly the variable your spreadsheet cannot hold.

SIX FAILURES

No framework is immune to misuse, and this one fails in six recognisable ways. Performative inclusion comes first: the maps get made, the sessions get hosted, the report gets published, and there was never any intention to be changed by any of it. People detect the performance quickly, and it corrodes trust considerably faster than honest exclusion would have. If input will not be permitted to alter the decision, asking for it is the more expensive option.

Stakeholder fatigue is the second, and it is real. Organisations that consult endlessly without deciding, or without ever demonstrating that what they heard shaped anything, teach people that participation is unpaid work with no output. Those people disengage, sometimes permanently, and they are usually the ones whose participation mattered most. The remedy is responsive engagement: ask, listen, act, and report back specifically on how the input changed the outcome. Every loop closes or it should not have been opened.

Developmental superiority is the third, and it is the failure mode of sophisticated people. I can see the whole system and they cannot is a seductive thought, and its arrival is almost always a reliable sign that something the dismissed stakeholder can see is being missed. Fourth, the map is mistaken for the relationship. A stakeholder map is an instrument that tells you with whom to build trust; no quantity of analysis substitutes for the patient work of actually building it, and the two are frequently confused because one of them can be finished on a Thursday.

Fifth, mapping is treated as static. Stakeholder landscapes move continuously — actors emerge, power shifts, issues change shape — and a map drawn once and never revisited becomes a liability that is trusted precisely because it exists. Sixth, and most often overlooked: the mapper forgets to map themselves. Every person drawing a map is also a stakeholder, with interests, biases, a developmental centre of gravity and an institutional position that determines what can be seen from where they sit.

A stakeholder map without a line about what the mapper cannot see is a map of one person's blind spot, drawn confidently.

The honest version of the document therefore carries a section titled what I might be missing because of who I am and where I sit. This is not self-flagellation and it is not ritual humility. It is basic epistemic responsibility, and it is the single cheapest addition to the practice — a paragraph that costs ten minutes and routinely saves the meeting that would otherwise have gone the way Meridian Health's went.

THE ETHICS

Mapping people is an exercise of power, and this needs stating plainly before any of the practices above are taken into a real organisation. The act of deciding who matters, in what order, and through which channel, allocates attention and therefore allocates outcome. That power is not made safe by good intentions. It is made accountable by a small number of commitments that are easy to state and genuinely inconvenient to keep.

Consent and transparency come first. Stakeholders should know that they are being mapped and should have some say in how their perspectives are represented. Covert developmental analysis — assessing where someone is making meaning from, without their knowledge, in order to gain a strategic advantage over them in a negotiation — is a straightforward violation of the ethic this framework runs on. The test is simple: would this person be comfortable reading the document that describes them.

Data stewardship follows directly. Developmental and relational information is deeply personal, more so than most of what an organisation stores about people. How it is recorded, who can see it, how long it is kept and what happens to it when the project ends are all decisions that must be made deliberately rather than by default. A stakeholder map containing developmental assessments deserves the handling given to clinical records, because that is functionally what it is.

Cultural humility is the third commitment and the one most often skipped by competent people. What counts as appropriate engagement, respectful communication and genuine inclusion varies enormously between cultural contexts, and a single model of engagement applied universally will read as consultation in one room and as intrusion in the next. The approach adapts to the stakeholder ecosystem it enters; it does not arrive with a fixed protocol and interpret friction as resistance.

Would this person be comfortable reading the document that describes them. Everything else in stakeholder ethics is downstream of that question.

Power awareness holds the other three together. Naming who defines the categories, who convenes the sessions, who writes the summary and who decides what counts as a theme makes the exercise auditable by the people it concerns. None of this slows the work meaningfully. It changes what the work is capable of producing, because people participate differently in a process whose mechanics they can see.

SHARED DIRECTION

The word alignment gets used as though it meant agreement, and the confusion is expensive. Agreement is everyone holding the same position, and it is usually purchased by silencing whoever was going to complicate things. Alignment is something else and something better: shared direction, held widely enough that disagreement becomes useful rather than destructive. The book states it in one line, and it is the sentence the whole argument rests on.

Alignment is not the absence of disagreement. It is the presence of shared direction held with enough spaciousness that disagreement becomes generative rather than destructive.

Seen this way, a stakeholder map is not the destination. It tells you who stands in the landscape and what forms of power, interdependence, meaning-making and absence run through it. It does not tell you how to bring those people into shared direction, how to work the tensions that necessarily arise when parties with real differences try to move together, or how to hold direction as conditions change underneath it. That is the work that follows mapping, and it is ongoing rather than concludable.

What can be done this week is concrete and takes under an hour. Take a live decision. In one colour, list every stakeholder you can identify and place them on a power-interest grid, thoroughly. In a second colour, draw the lines between them — who influences whom, which loops reinforce, which balance, where the clocks run at different speeds. In a third, annotate each with a short note on how they may be making meaning of the situation: what they value, what they are protecting, what they fear, written with curiosity rather than diagnosis.

In a fourth colour, add the absent — those affected and not yet visible — and for each one note why they are absent, what structural dynamic made them invisible, and what including them would actually require. Then stand back and look at the whole thing. The most useful measurement on the page is the distance between most affected and most included. One action to narrow that distance, taken this week, is worth more than another quarter of analysis.

Keep the map. Return to it monthly and change it as your perception deepens and conditions move, because a living map is an instrument and a filed one is a souvenir. The full argument, with the case, the developmental material and the practices worked through in detail, is one chapter of 7,160 words on The Developmental Canon shelf. It is free to read and free to hear — read aloud in full — and nothing in it is held back behind anything.

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 Stakeholder Landscape — Mapping Who Mattersopen this in the house search

A map drawn from a boardroom will always be, first of all, a map of the boardroom.

Consultation you do not intend to be changed by is a performance, and people can smell a performance.

The body registers an empty chair some minutes before the agenda does.

Adding detail to a map with too few dimensions produces a more confident wrong map.

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.