Why do two organizations with identical strategies, comparable talent, and access to the same markets so often behave as though they inhabit different worlds? One escalates a safety concern within hours; the other buries it for years. One treats dissent as data; the other treats it as disloyalty. The usual explanations of incentives, structure, and leadership quality capture part of the answer, but they leave a stubborn residue. That residue is what we have learned to call organizational culture.

The term is used loosely, often as a synonym for perks, mood, or a list of values printed on a lobby wall. Edgar Schein's more demanding formulation locates culture deeper. For Schein, culture consists of the shared, taken-for-granted assumptions a group has learned while solving its problems of external adaptation and internal integration. These assumptions work so reliably that members stop experiencing them as choices. They become, for those inside, simply the way reality is.

This raises a fundamental question about the nature of social reality. If people produce culture, how does it come to stand over them, shaping what they notice, what they consider reasonable, and who they believe themselves to be? Answering it requires tracing three movements: how cultures form, how they nest within larger national and industrial systems, and why deliberate attempts to change them so often fail. Each movement reveals the same underlying dynamic, in which minds build structures that then rebuild minds.

Culture Formation Mechanisms: How Contingency Becomes Common Sense

Organizations are not born culturally neutral. Arthur Stinchcombe's concept of imprinting captures how the conditions present at founding stamp themselves onto structures and practices that persist long after those conditions have vanished. These conditions include available technologies, local labor markets, and prevailing institutional logics. Research by James Baron and Michael Hannan on Silicon Valley startups found that founders' early employment blueprints shaped later outcomes, including turnover and the likelihood of going public. Some founders built around commitment, some around star talent, and some around engineering challenge.

Founders matter disproportionately because they act before anything is settled. Schein argued that leaders embed assumptions less through declarations than through behavior. What they systematically attend to, measure, and control carries the message. So do their reactions to crises, their allocation of rewards and status, and their decisions about whom to hire, promote, and expel. A founder who interrogates every expense teaches frugality more effectively than any memo. Members learn these lessons vicariously, by watching which behaviors are rewarded and which are quietly punished.

Yet leaders only propose; shared experience disposes. When a founder's assumption coincides with success, such as surviving a near-bankruptcy or defeating a larger rival, the group credits the assumption rather than luck or circumstance. Repeated validation moves the belief from espoused value to tacit premise. Karl Weick's work on sensemaking illuminates this process. Members collectively construct retrospective narratives that render ambiguous events coherent, and those narratives harden into the organization's working theory of how the world operates.

Transmission then becomes self-reinforcing. Benjamin Schneider's attraction-selection-attrition framework explains how this happens. People drawn to an organization tend to resemble its existing members, selectors favor candidates who seem to fit, and misfits gradually leave. Over time, the population homogenizes. Socialization completes the work, because newcomers learn not only formal procedures but also the identity the group offers. Social identity theory predicts that members internalize the group's norms to the extent that membership becomes part of how they define themselves.

Henri Tajfel's minimal group experiments showed that even arbitrary category assignments produce in-group favoritism. Organizations supply far richer categories than that, complete with histories, rivals, and heroes. The result is a system that recruits, trains, and retains the people who will reproduce it. This is culture's great strength, since it allows coordination without constant negotiation, and also its latent hazard. The mechanisms that create coherence also screen out the variation an organization may later need, determining which future problems are even perceptible.

Takeaway

Culture is less what leaders announce than what a group has watched succeed. Whatever an organization rewards during its formative crises will eventually feel like common sense to people who never witnessed them.

Multilevel Culture Analysis: Nations, Industries, and Firms

No organization's culture is self-contained. Each sits within a nested hierarchy of cultural systems, including national, regional, industrial, and occupational cultures, and each level shapes the assumptions members bring through the door. The analytical challenge is to determine which level explains which behaviors, and how the levels interact rather than simply stack. Treating organizational culture as a closed system produces both weak theory and costly managerial errors, particularly in multinational and cross-border contexts.

The foundational evidence came, paradoxically, from a single company. Geert Hofstede analyzed survey data from IBM employees across dozens of countries and found systematic national differences in values such as power distance, individualism, and uncertainty avoidance. These differences appeared despite a strong corporate culture designed to make employees alike. Because employer and occupation were held roughly constant, the national variation was difficult to dismiss. The later GLOBE project extended this work across sixty-two societies and distinguished cultural practices as they are from values as respondents wished them to be.

Hofstede drew an important conclusion that is frequently overlooked. National cultures differ primarily in values, which people acquire early in life. Organizational cultures differ primarily in practices, which people learn later through socialization at work. An organization can teach new routines, symbols, and rituals relatively easily. It struggles to rewrite the deeper orientations toward hierarchy, ambiguity, and the self that employees absorbed long before their first day. In this view, corporate culture is a thinner layer than its champions often assume.

Between nation and firm sits industry, a level that receives less attention than it warrants. Firms in the same industry share technologies, regulatory regimes, competitive tempos, and labor pools, and these produce recognizable cultural signatures. Examples include the risk tolerance of venture-backed software, the procedural caution of nuclear power, and the hierarchy of surgical teams. Research by Jennifer Chatman and Karen Jehn found substantial cultural variation across industries alongside notable similarity within them. Occupational subcultures cut across firms entirely, binding engineers or nurses to professional norms that no single employer controls.

Michele Gelfand's work on cultural tightness and looseness offers a useful integrating lens. Her research suggests that societies facing greater ecological and historical threat develop stronger norms and lower tolerance for deviance. Similar logic appears to apply to industries and firms operating under high-consequence conditions, where culture at each level is calibrated to perceived threat. The practical lesson is that organizational culture is best understood not as an independent creation but as a negotiated settlement among competing cultural claims.

Takeaway

An organization's culture is a negotiated settlement between the values members inherited and the practices the firm teaches. Strategies that ignore the deeper layers mistake a thin coat of paint for the structure beneath it.

Culture Change Challenges: Why Transformation Resists Intention

Few managerial ambitions are announced more confidently, or abandoned more quietly, than culture change. Consultancies routinely claim that most transformation efforts fail. The precise figures are contested and often poorly sourced, but practitioners and researchers recognize a consistent pattern. Ambitious programs generate workshops, value statements, and new vocabularies, while the underlying behavior of the organization stays remarkably stable. Understanding why requires taking culture's psychological function seriously rather than treating it as a communications problem.

The first problem is that change efforts typically target the wrong level. Schein's model distinguishes visible artifacts, espoused values, and basic underlying assumptions. Most interventions operate on the first two, through redesigned offices, refreshed mission statements, and new competency frameworks. Assumptions are the level that actually guides behavior, and they are largely invisible to those who hold them. When espoused values diverge from operative assumptions, members learn to perform the new language while continuing the old practice, and the gap breeds cynicism.

The second problem is identity. If culture provides members with a social identity, then members experience an attack on the culture as an attack on the self. Social identity theory predicts that threatened groups intensify in-group loyalty and derogate the source of threat. Mergers illustrate this vividly. Research on post-merger integration repeatedly finds that employees cling to pre-merger identities, cast the acquiring firm as an out-group, and resist integration in ways that financial models of synergy rarely anticipate.

Schein framed the core dynamic as a contest between two anxieties. Survival anxiety is the recognition that the current path is untenable. Learning anxiety is the fear of incompetence, lost status, and lost group membership that accompanies new ways of working. Change becomes possible when survival anxiety exceeds learning anxiety, and Schein argued that reducing learning anxiety works better than amplifying threat. Many programs do the reverse. They raise the threat level without providing psychological safety, which Amy Edmondson's research identifies as a precondition for experimentation and honest error reporting.

Successful transformations tend to share recognizable conditions. A credible crisis or disconfirmation unsettles old assumptions. Leaders model the new behavior consistently, especially in visible moments of trade-off. Structures, metrics, and promotion criteria change alongside rhetoric, so that the organization actually rewards the new assumptions. Change also builds on elements of the existing culture rather than repudiating it wholesale, which gives members a continuous identity. Culture changes when new practices succeed repeatedly, the same mechanism that formed it initially.

Takeaway

Culture cannot be decreed into existence; it can only be learned through repeated experience of what works. Change efforts succeed when they make learning feel safe while making the old way clearly untenable.

Organizational culture is perhaps the clearest laboratory for observing how minds and social structures constitute one another. Individuals, acting under particular founding conditions and leadership, generate shared solutions to common problems. Those solutions are validated, narrated, and institutionalized. The resulting system then selects, socializes, and shapes the very individuals who sustain it.

The multilevel view complicates any heroic account of this process. Organizations operate within national and industrial cultures that limit what they can teach, and their members carry values formed long before employment. Culture is therefore less an asset to be engineered than an ecology to be understood, with dynamics that resist unilateral control.

The broader implication extends beyond firms. Every institution, whether a university, a ministry, or a professional body, converts historical contingencies into common sense. Recognizing that what feels obvious was once a choice may be the first condition for choosing differently. For leaders and researchers alike, the task is to make the invisible architecture visible enough to examine.