Org, people & execution
Schein's Three Levels of Culture
Schein's model of culture as three levels: visible artifacts, espoused beliefs and values, and the basic underlying assumptions that actually drive behaviour. Diagnosis means working down the stack, because change that stops at the surface changes nothing.
Also known as Schein's culture model, Three levels of organisational culture. First set out by Edgar H. Schein in 1985; the primary source is cited in full below.
- Format
- Structural model
- Level
- Team · Business unit · Corporate
- Best for
- Plan execution
- Decision stage
- Diagnose · Plan
- Difficulty
- Advanced
- Time to apply
- Two to six weeks of observation and group inquiry for a serious diagnosis; a single workshop yields level-two claims dressed as level-three findings.
Plate · The model
The components
Artifacts
The visible surface: architecture, language, dress, rituals, published documents, observable routines. Easy to see and hard to decipher, because the same artifact can express opposite assumptions in different organisations.
Signals of strength
Office layout and who sits where · Rituals, ceremonies and recurring stories · Dress, language and forms of address · Visible processes, org charts and published values · What newcomers notice and veterans no longer see
Espoused beliefs and values
What the organisation says it believes: strategies, goals, philosophies, value statements. These may align with deeper assumptions, or serve as aspirations, or function as rationalisations that predict what people will say while failing to predict what they will do.
Signals of strength
Mission and value statements · Leaders' explanations of decisions · Norms invoked in debate ('that is not how we treat clients') · Gaps between the stated value and the rewarded behaviour
Basic underlying assumptions
Taken-for-granted beliefs about people, time, truth and the organisation's relationship to its environment, formed when past solutions worked repeatedly until they dropped out of awareness. They drive behaviour, resist confrontation, and define what members find literally unthinkable.
Signals of strength
Behaviour under pressure that contradicts stated values · Topics that are undiscussable, and the undiscussability itself undiscussable · Anxiety or irritation when the assumption is challenged · Uniform 'obvious' answers to why things are done this way · Newcomers' repeated mistakes in the same direction
When it earns its keep
- A change programme keeps failing in the same way and you suspect the cause sits below process and structure, in what people take for granted.
- Espoused values and observed behaviour visibly diverge, and you need language to describe the gap without blaming individuals.
- You are integrating after an acquisition or a leadership succession and need to understand what the organisation actually believes before deciding what to keep.
- You are diagnosing why a strategy that reads well on paper meets quiet, patterned resistance in execution.
And when it doesn't
- You need to compare many organisations quantitatively. Schein's method is clinical and case-based; survey instruments such as the Competing Values Framework suit benchmarking better.
- You want a quick engagement read-out. Assumptions surface through observation and guided group inquiry over weeks; a pulse survey measures climate, which is the weather rather than the geology.
- Culture is being reached for as the explanation of first resort. Incentives, staffing and plain resource shortages explain many failures more cheaply, and should be ruled out first.
- You expect the model to prescribe a target culture. It is a diagnostic structure; it deliberately does not say which assumptions are good.
How to run it
Before starting, gather the inputs the analysis depends on:
- Sustained observation of artifacts: physical space, rituals, meeting behaviour, stories, who gets promoted and what gets laughed at.
- The organisation's espoused record: value statements, strategy documents, induction materials, leaders' public claims.
- Access to insiders across tenure and level, including recent joiners, whose surprises are primary diagnostic data.
- The founding and crisis history of the organisation, since Schein holds that assumptions form where past solutions worked repeatedly.
- A concrete presenting problem, because Schein warns against studying culture in the abstract.
- 1
Start from a real problem
Anchor the diagnosis to something the organisation is trying to do and cannot. Schein is insistent on this: culture deciphered in the abstract produces description without traction, and the culture only shows its relevant face under the pressure of an actual task.
- 2
Record artifacts without interpreting them
Walk the site, sit in the meetings, read the noticeboards, note the dress, the jokes, the seating, who speaks and who waits. Write down what a newcomer would see. Resist decoding at this stage; artifacts are easy to observe and dangerously easy to misread from your own assumptions.
- 3
Elicit the espoused beliefs and values
Gather what the organisation says about itself, from the values poster to what managers tell recruits. Treat these as claims rather than findings. They matter mostly for the next step, which is where the diagnosis actually happens.
- 4
Hunt the contradictions
Compare levels one and two. Where artifacts contradict espoused values, an underlying assumption is at work: the firm that espouses candour but whose meetings are silent, the safety-first plant with improvised workarounds. List every contradiction; each is a thread to pull.
- 5
Surface the basic assumptions with insiders
In small group sessions, use the contradictions and the organisation's history to name what is actually taken for granted. Ask what a newcomer must learn to stop making mistakes, what is undiscussable, which past successes taught which lessons. Assumptions ring true when insiders wince slightly and agree.
- 6
Assess the assumptions against the task
Sort the surfaced assumptions into those that help the presenting problem, those that hinder it, and those that are neutral. Plan change to work with the grain where possible: Schein's counsel is that you rarely change assumptions frontally, but you can redeploy existing assumptions to support new behaviour, and only sustained new success experiences rewrite the deeper level.
Reading the result
A layered cultural diagnosis: an artifact inventory, the espoused record, the contradictions between them, and a short list of named underlying assumptions assessed as helping or hindering the presenting problem, with implications for how change should be designed.
- The contradictions are the findings. Where all three levels align, culture is working as designed; where they diverge, the assumption underneath the contradiction is what any change effort will actually meet.
- Judge assumptions against the task rather than against virtue. An assumption that hinders digitisation may be the same one that built the firm's service reputation.
- Expect the deepest level to change last and only through lived success. A new value statement is an artifact, whatever it says.
A worked example
A family-owned shipping agency diagnoses its culture before digitising
A third-generation family-run port agency on the Humber handles vessel calls for dry-bulk and short-sea traffic: berthing, customs, crew changes, husbandry. The incoming managing director, the founder's granddaughter, wants to adopt a digital port-call platform and win container-feeder business. Two previous software rollouts quietly died. Before a third attempt, she commissions a proper cultural diagnosis anchored to that presenting problem.
- Artifacts
- Harbour-view office with the founder's portrait and a brass ship's clock. Boarding clerks in at 05:30 when a vessel is due, unasked. Paper call files in ring binders, one per vessel, annotated in pencil. Masters and agents greeted by first name; the fastest phone answers in the port. New software licences from both previous rollouts, paid for and unused. Saturday errands run personally for crews, unbilled.
- Espoused beliefs and values
- The website and the MD's own induction speech say: any ship, any hour; the client deals with a person, never a system; three generations of trust; we are the master's agent first and everyone else's second. Internally, managers espouse professionalism and say they welcome modernisation, and minuted meetings from both failed rollouts record unanimous support.
- Basic underlying assumptions
- Group sessions built on the contradictions surface four. First, service is personal: anything that puts a screen between the clerk and the master degrades the product, so using the platform felt like shirking. Second, the family name is the guarantee, so mistakes are corrected quietly in-house and never systematised into logs a client might see. Third, real knowledge lives in people: the senior clerk's memory of a charterer's quirks is the asset, and writing it down feels like both a demotion and a redundancy plan. Fourth, outsiders do not understand the trade, which is why both software vendors, and by extension their products, were politely frozen out.
The read. The rollouts did not fail for want of training budget; they collided with assumptions that equate documentation with depersonalisation and systems with distrust. The diagnosis suggests a design that works with the grain: position the platform as the way the firm keeps its any-hour promise as call volumes grow, have the senior boarding clerks author the workflow templates so codified knowledge carries their names, and keep the personal greeting ritual deliberately intact. The deeper assumptions will shift only if the first digitised container-feeder calls visibly succeed and are celebrated as the family standard upheld, which argues for one carefully chosen pilot client rather than a fleet-wide launch.
Pitfalls
- Reading culture off the artifacts. Open-plan offices and values posters support confident wrong conclusions; the same artifact can encode opposite assumptions.
- Treating the espoused level as the culture. Value statements predict speeches, and only sometimes behaviour; the diagnosis lives in the gap.
- Diagnosing without a presenting problem, which yields an interesting portrait and no decision.
- Assuming one culture. Departments, sites and generations may hold different assumptions, and averaging them hides exactly the fault lines that break change programmes.
- Launching frontal assaults on assumptions. Naming an assumption does not dissolve it; only repeated success with new behaviour does, and that takes design and time.
- Confusing climate with culture. Mood surveys move quarterly; assumptions move over years.
What the critics say
Martin's three-perspective critique argues Schein's model embodies an integration view, culture as shared and consensual, and that real organisations simultaneously show differentiation (subcultures in conflict) and fragmentation (ambiguity with no stable consensus), so a single stack of shared assumptions can misdescribe the object it claims to diagnose.
Martin, J. (1992) Cultures in Organizations: Three Perspectives. New York: Oxford University Press.
The deepest level resists measurement almost by definition: assumptions are unconscious and elicited through clinical inquiry, which makes findings hard to replicate or compare. Large-scale empirical work concluded that shared perceptions of daily practices, rather than deep values, are the measurable core of organisational culture, a direct methodological challenge to building diagnosis on the assumption level.
Hofstede, G., Neuijen, B., Ohayv, D. D. and Sanders, G. (1990) 'Measuring Organizational Cultures: A Qualitative and Quantitative Study across Twenty Cases', Administrative Science Quarterly, 35(2).
The model is descriptive rather than predictive: it explains failures convincingly after the fact but offers weak guidance on how much cultural resistance a given change will meet in advance, and the clinical method depends heavily on the skill of the individual diagnostician.
Run it as a workshop
Set expectations honestly before the room opens: a single workshop surfaces artifacts and espoused values well, but the deepest layer, the basic underlying assumptions, needs real observation and inquiry over weeks, not one session. Use this workshop to generate the presenting question, not to close it.
Running order
| Name the actual presenting problem this diagnosis serves | 15 min |
| Catalogue visible artifacts honestly | 25 min |
| Surface espoused beliefs and values, and where they diverge from behaviour | 30 min |
| Discuss what basic assumptions might explain the gap | 30 min |
| Agree what further observation or inquiry is actually needed | 20 min |
| Total | 2h |
You will need
- A three-layer template (artifacts, espoused values, assumptions) drawn large
- Photographs or examples of real artifacts if possible (physical space, rituals, language)
- The organisation's actual published values statement, for honest comparison
- A parking-lot sheet for assumptions the room can only guess at, flagged for real inquiry
Traps to avoid
- Reading culture straight off the artifacts. Open-plan offices and values posters invite confident, wrong conclusions; the same artifact can encode opposite assumptions in different organisations.
- Treating the espoused layer as the culture itself. Value statements predict what people say in speeches, only sometimes what they actually do; the real diagnosis lives in the gap between the two.
- Running this with no presenting problem in mind. A diagnosis with nothing it is meant to explain produces an interesting portrait and no decision.
- Presenting workshop-level findings as though they were the deep, verified level-three diagnosis. Say plainly in the room that this is a first pass, not the final word.
Sources and further reading
- Schein, E. H. (1985) Organizational Culture and Leadership. San Francisco: Jossey-Bass.
- Schein, E. H. and Schein, P. A. (2017) Organizational Culture and Leadership, 5th edn. Hoboken: Wiley. ↗
- Schein, E. H. (1999) The Corporate Culture Survival Guide. San Francisco: Jossey-Bass.
- Martin, J. (1992) Cultures in Organizations: Three Perspectives. New York: Oxford University Press.