Schein's Levels of CultureWhat the Model Sees, and What It Cannot

After finishing his doctorate in 1952, Edgar Schein served as a psychologist in the US Army and was sent to interview American prisoners of war returning from Korea. They had been subjected to sustained attempts at indoctrination by their Chinese captors, and Schein’s task was to understand what had been done to them. The concept he developed from that work was coercive persuasion: belief change produced by controlling the whole environment in which a person’s beliefs are formed.

Schein went on to become one of the best-known theorists of organisational culture, and he never pretended the two subjects were unrelated. Late in his life he observed that coercive persuasion applies equally to the golden handcuffs. People who are economically committed to an institution can be socialised into its culture by much the same logic, and once they are, there is little reason left to be a dissident.

That is an uncomfortable origin for the model of organisational culture that management courses now teach as standard. It is also the most honest place to start, because Schein’s three levels of culture are the clearest account available of how culture hides, and the managerial reading of them has left out the part Schein learned first: that culture is also how groups hold people in place.

What Schein’s Three Levels of Culture Describe

Schein divided culture into three levels. At the surface are artifacts, the things a visitor can observe on arrival: the building and the office layout, the dress code, the way meetings run, the documents an organisation publishes about itself. Artifacts are easy to see and hard to read. A visitor can watch everyone stand for the morning meeting without knowing what standing means.

Below them are espoused values, the reasons people give for what they do. We put customers first. Our people are our greatest asset. These statements are usually sincere, and they are usually an imperfect guide to how the organisation behaves.

At the bottom are basic underlying assumptions, beliefs so settled that nobody thinks to state them: about what counts as evidence, how time works, what people are like, what success looks like. In his 1983 MIT working paper, Organizational Culture: A Dynamic Model, Schein defined culture as the pattern of basic assumptions a group has invented, discovered or developed in coping with its problems of external adaptation and internal integration, which have worked well enough to be considered valid, and are therefore taught to new members as the correct way to perceive, think and feel in relation to those problems.

The model’s lasting value lies in the distance between the second level and the third. Every organisation says one thing and does another, and Schein explains the gap without accusing anyone of lying. The espoused value is real. It simply sits on top of an assumption that pulls behaviour somewhere else. An organisation can believe sincerely in collaboration while assuming, at a depth nobody examines, that individual heroics are what get promoted. Behaviour follows the assumption, and the value survives as a hope printed on the wall.

The definition carries another detail that is easy to pass over. Assumptions become culture because they worked well enough to be considered valid. Culture is a record of problems a group once solved, and so it is conservative by construction: it keeps answering questions the organisation was asked years ago. When the environment changes, the assumptions carry on working as assumptions long after they have stopped working as solutions, which is part of why organisations behave less like machines that can be reprogrammed and more like the complex adaptive systems they are. A shared assumption is also a relief. It removes a question nobody has to answer again, and that relief is hard to give up even when the answer has gone stale.

That insight explains a great deal: why culture programmes so often stall once the posters are up, why new starters learn the culture from what happens rather than from what they are told, and why, as The Fractured Lens and Culture argues, the culture an organisation describes and the culture it lives are so rarely the same thing.

Why Organisations Read Schein as a Manual for Culture Change

The difficulty begins with what organisations did with the model. Schein argued that founders and early leaders supply most of a group’s initial cultural solutions, and that leadership matters most when a group’s habitual responses stop working. From there it was a short step to treating culture as something leaders create and manage, a lever like any other.

That step was taken at scale from the 1980s onwards. Culture became a management object, with its own audits and consultants. Each version rested on the same hope: that if the assumptions at the bottom of the iceberg could be surfaced, they could be swapped for better ones.

The most common modern form of that hope is the values framework. An organisation writes down its values, which in Schein’s terms sit at the second level, and then begins assessing people against them. Living the values becomes a line in the annual review, and sometimes a reason for exit. Something odd has happened in that move. Statements that Schein showed to be an unreliable guide to behaviour are now being used to judge behaviour, and the judging is done by the people whose assumptions the statements were meant to express. As The Review explores, the ritual asks for honesty in a setting built to grade it. Employees learn to perform the second level for an audience that rarely examines its own third.

Hugh Willmott saw a darker shape in this. His 1993 paper in the Journal of Management Studies, Strength Is Ignorance; Slavery Is Freedom, took its title from Orwell and compared the corporate culture movement to the Party’s methods of control in Nineteen Eighty-Four. The movement promised autonomy, because a strong culture would let people manage themselves without supervision. Willmott pointed out that the self doing the managing had first to be shaped by the culture, so the autonomy on offer was autonomy to want what the organisation already wanted. He called the claim that such autonomy could flourish inside a single, enclosed set of values a form of doublethink.

Gideon Kunda’s ethnography of an American technology company, published in 1992 as Engineering Culture, showed the same thing from inside. The firm he called Tech worked hard at shaping its engineers’ commitment, through leaders’ speeches and a stream of internal publications. It professed autonomy while rewarding only a narrow version of it. The effort it extracted was extraordinary, and so was the toll on people’s lives outside work. Kunda named this normative control, meaning control exercised through what people come to value rather than through rules about what they do.

Set Kunda’s account beside Schein’s work in Korea and the continuity is plain. Coercive persuasion worked by controlling the environment in which beliefs formed. Normative control works by the same logic in a nicer building, with a salary attached. Schein saw the link himself. The culture industry, which borrowed his levels and left his first research behind, mostly did not.

There is a further detail in the 1983 paper that the industry kept in a different form. Schein wrote that leaders must give members enough psychological safety to let go of old responses before a culture can change. The safety he meant was a condition for unlearning. In many organisations it has become a programme in its own right, with its own measures, and as The Illusion of Psychological Safety traces, a programme can become one more thing people learn to perform.

What Schein’s Model of Organisational Culture Cannot See

The model’s deepest limit sits inside its definition. Culture, for Schein, belongs to a group. The 1983 paper says so directly: there cannot be any culture unless there is a group which owns it. That sentence settles a question the model never asks out loud, which is who the group is, and who decides.

Joanne Martin’s Cultures in Organizations, published in 1992, offered three ways of answering it. An integration perspective looks for consensus and consistency, one culture with shared meanings. A differentiation perspective looks for subcultures, such as engineers and sales staff, or head office and the regions, each holding assumptions that may conflict with the others. A fragmentation perspective looks for ambiguity, where meanings are contested and keep shifting, and where the honest answer to what people believe depends on who you ask and when.

Schein’s model, as it is usually taught, lives in the first of these. He allowed for subcultures, and the 1983 paper notes that stable subgroups develop their own. But the shape of the model pulls towards coherence: three levels, one culture, a bottom layer everyone shares.

The organisation-wide culture survey shows what happens when the first perspective is imposed on the other two. Responses from the warehouse and the boardroom, the new graduate and the thirty-year veteran, are combined into a single score for each value and reported back as the culture. By Schein’s own test, that averaged culture is owned by no group at all. It describes an organisation that nobody inside it experiences, and it is then used to plan interventions in the cultures people actually live in.

Hybrid work exposed the same blind spot from another direction. When offices emptied, many leaders worried aloud that the culture was being lost, and in one sense they were right: the artifacts they had relied on to see it, the full floor and the buzz of the kitchen, had gone. Some employees experienced something different. Their working relationships had moved into chat threads and smaller calls, with new norms about cameras and response times, and for some the culture had become easier to live in. What some leaders had lost was the view of the culture from the centre, and that view is easily mistaken for the culture itself by the people who hold the centre.

Most organisations people actually work in are layered rather than coherent. Several cultures sit on top of each other, and much of what people hold in common is uncertainty about which stated values will survive the next restructure, or whether the new chief executive means what she says. The bottom of the iceberg, for a great many employees, is a hedge. They keep a version of themselves ready for the culture as announced and another for the culture as it turns out to be. That double readiness is invisible to a model built to find what is shared, and it is one of the most widespread forms of unknowing in organisational life.

Ownership is also a question of power. When leaders describe the culture, they tend to describe their own assumptions and present them as the group’s. People whose assumptions differ tend to appear in the model as a culture problem rather than as another culture, a resistance to be managed, as Why Change Doesn’t Fail Because People Resist It explores. The language of culture lets a disagreement about how things should be done be recast as a deficiency in the people disagreeing.

Mergers make this unusually visible. When two organisations combine, the language of culture clash arrives quickly, and the clash is rarely resolved by negotiation between equals. One set of assumptions becomes the culture and the other becomes legacy. People from the acquired organisation discover that habits they never had to defend, about how decisions are made or how much is said in meetings, have become evidence that they are not yet fully on board. Neither culture was more correct. One belonged to the side with the authority to name things.

Look again at the word Schein chose: the solutions are taught to new members as the correct way to perceive, think and feel. Read as description, that is accurate. Read as an experience, it describes the position of every newcomer, someone learning which perceptions are correct and which feelings are allowed. Schein studied this too, under the name organisational socialisation. The newcomer is where culture is most visible, because the newcomer still notices it.

The noticing does not last. In the first months a new employee sees the gap between levels with great clarity: the meeting where the stated value lost to the unstated one, the colleague who was praised for something the induction pack had warned against. Within a year most of those observations have stopped registering, absorbed rather than resolved, and the newcomer now teaches the next one which perceptions are correct. That loss of noticing is the culture succeeding, and it is also the moment an organisation loses one of the few people who could have told it what it looks like from outside.

Culture fit is where the word correct does its most consequential work. As a hiring criterion it asks whether a candidate already shares the assumptions at the bottom of the iceberg. It sounds like a question about compatibility. In practice it decides belonging before anyone has walked through the door, and it does so on grounds that, by Schein’s own account, nobody involved can fully state, because the assumptions being matched are the ones people take for granted. A culture that selects for its own assumptions will find them confirmed everywhere it looks.

Reading Organisational Culture Through The Fractured Lens

Schein’s levels remain the most useful map of how culture hides. The Fractured Lens reads the same terrain with different attention: to who drew the map, and what it costs the people standing at the edges of it.

That attention passes through the levels rather than replacing them. The meanings Schein mapped so well, the stories that connect an artifact to an assumption, are never neutral. Someone’s meanings became the assumptions everyone else is taught, and what looks from the centre like a shared understanding can look from the edge like a condition of membership, or like a sensemaking settlement that the people with least say were never invited to make.

Schein himself travelled some distance from the executive reading of his work. His later book Humble Inquiry argued that leaders, the people expected to set direction and declare values, are the ones who most need to learn to ask rather than tell. It is a striking final emphasis from the theorist whose model had taught a generation of executives that culture was theirs to shape. It is also a return to his beginning, in a way: to someone sitting across a table from a person whose beliefs had been reshaped by a group, trying to understand how it had felt.

His model asks every organisation what it takes for granted. Asked of the model itself, the answer may be a group, and a vantage point from which that group can be seen whole. Few people inside organisations ever stand there. They see their own corner, read the culture from the inside of it, and wonder, mostly in private, whether the people at the centre are describing the same place.