three cultures of management: the key - prescient...

Three Cultures of Management: The Key toOrganizational Learning

Edgar H. Schein

Reprint 3811

MassachusettsInstitute of Technology

Fall 1996

Volume 38Number 1

MITSMR022

9SLOAN MANAGEMENT REVIEW/FALL 1996 SCHEIN

Why do organizations fail to learn how tolearn and therefore remain competitivelymarginal? In this article, I try to explain

why organizational innovations either don’t occur orfail to survive and proliferate. Some typical explana-tions revolve around vague concepts of “resistance tochange,” or “human nature,” or failures of “leader-ship.” I propose a more fundamental reason for suchlearning failures, derived from the fact that, in everyorganization, there are three particular cultures amongits subcultures, two of which have their roots outsidethe organization and are therefore more fundamental-ly entrenched in their particular assumptions. Everyorganization develops an internal culture based on itsoperational success, what I call the “operator culture.”But every organization also has, in its various func-tions, the designers and technocrats who drive thecore technologies. I call this the “engineering culture”;their fundamental reference group is their worldwideoccupational community. Every organization also hasits executive management, the CEO and his or herimmediate subordinates — what I call the “executiveculture.” CEOs, because of the nature of their jobsand the structure of the capital markets, also consti-tute a worldwide occupational community in the sensethat they have common problems that are unique totheir roles.

These three cultures are often not aligned witheach other, and it is this lack of alignment that causesthe failures of organizational learning that I will dis-

cuss. The question is whether we have misconceivedthe initial problem by focusing on organizationallearning, when, in fact, it is the executive and engi-neering communities that must begin their own learn-ing process if we are to meet the challenges of thetwenty-first century.

Organizations Don’t Learn; InnovationsDon’t Last or Diffuse

The ability to create new organizational forms andprocesses, to innovate in both the technical and orga-nizational arenas, is crucial to remaining competitivein an increasingly turbulent world. But this kind oforganizational learning requires not only the inven-tion of new forms but also their adoption and diffu-sion to the other relevant parts of the organizationand to other organizations in a given industry. Orga-nizations still have not learned how to manage thatprocess. The examples of successful organizationallearning we have seen either tend to be short-runadaptive learning — doing better at what we are al-ready doing — or, if they are genuine innovations,tend to be isolated and eventually subverted andabandoned.

For example, a new product development teamin a large auto company worked with the MIT

Three Cultures of Management:The Key to Organizational Learning

Edgar H. Schein

Edgar H. Schein is the Sloan Fellows Professor of Management Emeritusat the MIT Sloan School of Management.

The three communities of executives, engineers, and operators do not really

understand each other very well. A lack of alignment among the three groups

can hinder learning in an organization.

Organizational Learning Center to develop a capaci-ty for learning. By using various techniques derivedfrom “action science,” systems dynamics, and orga-nization development, the team created high levelsof openness between hierarchical levels and increasedcommunication and trust among its members.1 Thisopenness and trust permitted team members to re-veal engineering design problems as they arose in-stead of waiting until they had solutions, as prior tra-dition in this company had dictated.2

Early identification of those problems was crucialin order to avoid later interactive effects that would re-quire costly, complex redesigns. For example, chang-ing the chassis design might increase weight, which

might require a different tire design, which, in turn,might cause more internal noise, and so on. By re-vealing such problems early, the team could view thewhole car more systemically and could thereforespeed up redesign.

However, the pileup of early problems caused upper-level managers to make a false attribution. They con-sidered the team to be “out of control” and ordered itto get itself back under control. The team realized thathigher management did not understand the value ofearly problem identification and continued to use itsnew learning, assuming that the ultimate results wouldspeak for themselves. The team was able to completethe design well ahead of schedule and with consider-ably lower costs, but, contrary to expectations, highermanagers never understood the reasons for these no-table results nor gave the team credit for having learneda new way of solving problems. Instead, higher man-agers gave themselves credit for having gotten theteam “under control.” They did not consider the teamto be particularly innovative and disbanded it. Theysubsequently encouraged several of its members and

leaders to take early retirement as part of the compa-ny’s general downsizing program.

In another example, an insurance company decid-ed to move toward the paperless office.3 Top manage-ment hired a manager to implement the new system,mandated a schedule, and provided whatever re-sources the manager needed to accomplish the task.In order to use the new system, employees had tolearn complex new computer routines to replace theirfamiliar work with paper. Because the company wasalso under financial pressure, it had instituted a num-ber of productivity programs that caused line man-agers to insist that all the daily work continue to beperformed even while the learning of the new systemwas supposed to take place. The new manager wasequally insistent that the system be implemented onschedule, causing employees to short-circuit certainroutines, to learn only the rudiments of the new sys-tem, and even to misrepresent the degree to whichthey were now working without paper.

The new manager, based on partial and incorrectinformation, declared that the system was implement-ed “on schedule” and was given public credit for thisachievement. However, the result was that the employ-ees did not learn the new system well enough to makeit more productive than the old paper system. In fact,productivity was lower with the new system because itwas so imperfectly implemented.

In a third example, a company decided to intro-duce automatic machine tools into its productionprocess.4 The idea originated with the engineers whosaw an opportunity to do some “real” engineering.The engineers and the vendors developed a proposalbased on technical elegance but found that middlemanagement would not push the proposal up to exec-utive management unless it was rewritten to showhow it would reduce costs by cutting labor. No accu-rate figures were available, so the team more or lessinvented the numbers to justify the purchase of theexpensive new machines.

As the proposal worked its way up the hierarchy,the labor union got wind of the project and insistedthat it would not go along unless management guar-anteed that no jobs would be lost and that all thepresent operators would be retrained. This not onlydelayed the project, but, when the machines were fi-nally installed, the production process proved to be

10 SCHEIN SLOAN MANAGEMENT REVIEW/FALL 1996

In most organizations, there are three different major

occupational cultures that donot really understand each other very well and that often work at

cross-purposes.

much less effective and much more costly than hadbeen promised in the proposal. The engineers werehighly disappointed that their elegant solution had,from their point of view, been subverted and that allthe operators that were to have been replaced hadmerely been retrained and kept on jobs that the en-gineers considered superfluous.

Beyond these three specific cases, the history oforganizational development, change, innovation,and learning shows over and over that certain lessonsseem not to take hold. Since the Hawthorne studiesof the 1920s, it has been recognized that employeeinvolvement increased both productivity and motiva-tion. Lewin, Argyris, McGregor, Likert, and manyothers showed how managers who treated people asadults, who involved them appropriately in the tasksthat they were accountable for, and who created con-ditions so employees could obtain good feedback andmonitor their own performance were more effectivethan those who did not.5

Programs such as the National Training Labs’ sensi-tivity training groups and Blake’s managerial gridwere, for several decades, touted as the solution to allour productivity problems, just as the human rela-tions and participatory management programs of theforties had promised.6 Yet these and other similarprograms have come and gone, and it is not at all

clear what organizations learned from them or whythese innovations have disappeared, only to be rein-vented under new labels such as empowerment, self-managed groups, and servant leadership.

The lesson of these and similar cases is complicat-ed. On the one hand, we can say that this is just nor-mal life in organizations. It is just politics or just

human nature. Or we can say that these projects andprograms were mismanaged, by either the projectteams or the executive managers above them. Or wecan say that all these human-relations-oriented pro-grams were misguided in the first place. However, Ihave begun to see deeper phenomena at work here.

The deeper issue is that in most organizations,there are three different major occupational culturesthat do not really understand each other very welland that often work at cross-purposes. These cul-tures cut across organizations and are based on whathave been described as “occupational communities.”7

The Concept of Culture and OccupationalCommunities

A culture is a set of basic tacit assumptions about howthe world is and ought to be that a group of peopleshare and that determines their perceptions, thoughts,feelings, and, to some degree, their overt behavior.8

Culture manifests itself at three levels: the level of deeptacit assumptions that are the essence of the culture,the level of espoused values that often reflect what agroup wishes ideally to be and the way it wants to pre-sent itself publicly, and the day-to-day behavior thatrepresents a complex compromise among the espousedvalues, the deeper assumptions, and the immediate re-quirements of the situation. Overt behavior alone can-not be used to decipher culture because situationalcontingencies often make us behave in a manner thatis inconsistent with our deeper values and assump-tions. For this reason, one often sees “inconsistencies”or “conflicts” in overt behavior or between behaviorand espoused values. To discover the basic elements ofa culture, one must either observe behavior for a verylong time or get directly at the underlying values andassumptions that drive the perceptions and thoughtsof the group members.

For example, many organizations espouse “team-work” and “cooperation,” but the behavior that theincentive and control systems of the organization re-ward and encourage is based more on a shared tacitassumption that only individuals can be accountableand that the best results come from a system of indi-vidual competition and rewards. If the external situa-tion demands teamwork, the group will develop somebehavior that looks, on the surface, like teamwork by


Aculture is a set of basic tacitassumptions about how theworld is and ought to be

that a group of people share andthat determines their perceptions,thoughts, feelings, and, to some

degree, their overt behavior.

conducting meetings and seeking consensus, butmembers will continue to share the belief that theycan get ahead by individual effort and will act accord-ingly when rewards are given out. I have heard manyexecutives tell their subordinates that they expectthem to act as a team but remind them in the samesentence that they are all competing for the boss’s job!

Cultures and Subcultures Cultures arise within organizations based on theirown histories and experiences. Starting with thefounders, those members of an organization who haveshared in its successful growth have developed as-sumptions about the world and how to succeed in it,and have taught those assumptions to new membersof the organization.9 Thus IBM, Hewlett-Packard,Ford, and any other company that has had severaldecades of success will have an organizational culturethat drives how its members think, feel, and act.

Shared assumptions also typically form aroundthe functional units of the organization. They areoften based on members’ similar educational back-grounds or similar organizational experiences, whatwe often end up calling “stove pipes” or “silos.” Weall know that getting cross-functional project teamsto work well together is difficult because the mem-bers bring their functional cultures into the projectand, as a consequence, have difficulty communicat-ing with each other, reaching consensus, and imple-menting decisions effectively. The difficulty of com-munication across these boundaries arises not onlyfrom the fact that the functional groups have differ-ent goals, but also from the more fundamental issuethat the very meaning of the words they use will dif-fer. The word “marketing” will mean product devel-opment to the engineer, studying customers throughmarket research to the product manager, merchan-dising to the salesperson, and constant change in de-sign to the manufacturing manager. When they tryto work together, they will often attribute disagree-ment to personalities and fail to notice the deeper,shared assumptions that color how each functionthinks.

Another kind of subculture, less often acknowl-edged, reflects the common experiences of given lev-els within a hierarchy. Culture arises through sharedexperiences of success. If first-line supervisors discover

ways of managing their subordinates that are consis-tently successful, they gradually build up shared as-sumptions about how to do their job that can bethought of as the “culture of first-line supervision.” Inthe same way, middle management and higher levelswill develop their own shared assumptions and, ateach level, will teach those assumptions to newcom-ers as they get promoted. These hierarchically basedcultures create the communication problems associat-ed with “selling senior management on a new way ofdoing things,” or “getting budget approval for a newpiece of equipment,” or “getting a personnel requisi-tion through.” As each cultural boundary is crossed,the proposal has to be put into the appropriate lan-guage for the next higher level and has to reflect thevalues and assumptions of that level. Or, from the

viewpoint of the higher levels, decisions have to beput into a form that lower levels can understand,often resulting in “translations” that actually distortand sometimes even subvert what the higher levelswanted.

So far, I have focused on the cultures that arise with-in organizations from the unique experiences of itsmembers. But “occupational communities” also gener-ate cultures that cut across organizations.10 For exam-ple, fishermen around the world develop similar world-views, as do miners, as do the members of a particularindustry based on a particular technology. In thesecases, the shared assumptions derive from a commoneducational background, the requirements of a givenoccupation such as the licenses that have to be ob-tained to practice, and the shared contact with othersin the occupation. The various functional cultures inorganizations are, in fact, partly the result of member-ship in broader cross-organizational occupational


Decisions have to be put into aform that lower levels can

understand, often resulting in“translations” that actually distort and

sometimes even subvert what thehigher levels wanted.

communities. Salespeople the world over, accountants,assembly line workers, and engineers share some tacitassumptions about the nature of their work regardlessof who their particular employer is at any given time.

Such similar outlooks across organizations alsoapply to executive managers, particularly CEOs. CEOsface similar problems in all organizations and in allindustries throughout the world. Because executivesare likely to have, somewhere in their history, somecommon education and indoctrination, they form acommon worldview — common assumptions aboutthe nature of business and what it takes to run a busi-ness successfully.

Three Cultures of Management

The learning problems that I have identified can be di-rectly related to the lack of alignment among three cul-tures, two of which are based on occupational commu-nities — (1) the culture of engineering, (2) the cultureof CEOs, and (3) the culture of operators — and theshared assumptions that arise in the “line units” of agiven organization as it attempts to operate efficientlyand safely. To understand how these three cultures in-teract, let us examine their shared assumptions.

The Operator Culture The culture of operators is the most difficult to de-scribe because it evolves locally in organizations andwithin operational units (see the sidebar). Thus wecan identify an operator culture in the nuclear plant,the chemical complex, the auto manufacturing plant,the airplane cockpit, and the office, but it is not clearwhat elements make this culture broader than thelocal unit. To focus on this issue, we must considerthat the operations in different industries reflect thebroad technological trends in those industries. Atsome fundamental level, how one does things in agiven industry reflects the core technologies that cre-ated that industry. And, as those core technologiesthemselves evolve, the nature of operations changes.For example, as Zuboff has persuasively argued, in-formation technology has made manual labor obso-lete in many industries and replaced it with concep-tual tasks.11 In a chemical plant, the worker no longerwalks around observing, smelling, touching, and ma-nipulating. Instead he or she sits in a control room

and infers the conditions in the plant from the vari-ous indexes that come up on the computer screen.

The operator culture is based on human interac-tion, and most line units learn that high levels ofcommunication, trust, and teamwork are essential togetting the work done efficiently. Operators also learnthat no matter how clearly the rules are specified as towhat is supposed to be done under different opera-tional conditions, the world is to some degree unpre-dictable and one must be prepared to use one’s owninnovative skills. If the operations are complex, as in anuclear plant, operators learn that they are highly in-terdependent and must work together as a team, espe-cially when dealing with unanticipated events. Rulesand hierarchy often get in the way in unpredictedconditions. Operators become highly sensitive to thedegree to which the production process is a system ofinterdependent functions, all of which must work to-gether to be efficient and effective. These pointsapply to all kinds of “production processes,” whethera sales function, a clerical group, a cockpit, or a ser-vice unit.

The tragedy of most organizations is that the opera-tors know that, to get the job done effectively, theymust adhere to the assumptions stated above, but thatneither the incentive system nor the day-to-day man-agement system may support those assumptions.Operators thus learn to subvert what they know to betrue and “work to rule,” or use their learning ability tothwart management’s efforts to improve productivity.To understand why this happens, we must examinehow two other major cultures operate in organizations.


Assumptions of the Operator Culture

• Because the action of any organization is ultimately the ac-tion of people, the success of the enterprise depends on peo-ple’s knowledge, skill, learning ability, and commitment.

• The required knowledge and skill are “local” and based onthe organization’s core technology.

• No matter how carefully engineered the production processis or how carefully rules and routines are specified, operatorsmust have the capacity to learn and to deal with surprises.

• Most operations involve interdependencies between sepa-rate elements of the process; hence, operators must be ableto work as a collaborative team in which communication,openness, mutual trust, and commitment are highly valued.

The Engineering Culture In all organizations, one group represents the basic de-sign elements of the technology underlying the workof the organization and has the knowledge of how thattechnology is to be utilized. This occupational com-munity cuts across nations and industries and can bestbe labeled the “engineering culture.”12 A colleague whoworks for a company driven by the engineering culturetold me that in the parking lot of his company, signssay, “Maximum Speed Limit: 5.8 Miles Per Hour.”Although this culture is most visible in traditional engi-neering functions, it is also evident among the design-ers and implementers of all kinds of technologies —information technology, market research, financial sys-tems, and so on. The shared assumptions of this com-munity are based on common education, work experi-ence, and job requirements (see the sidebar).

Engineers and technocrats of all persuasions are at-tracted to engineering because it is abstract and im-personal. Their education reinforces the view thatproblems have abstract solutions and that those solu-tions can, in principle, be implemented in the realworld with products and systems free of humanfoibles and errors. Engineers, and I use this term inthe broadest sense, are designers of products and sys-

tems that have utility, elegance, permanence, efficien-cy, safety, and maybe, as in the case of architecture,even aesthetic appeal, but they are basically designedto require standard responses from their human oper-ators, or, ideally, to have no human operators at all.

In the design of complex systems such as jet air-craft or nuclear plants, the engineer prefers a techni-cal routine to ensure safety rather than relying on ahuman team to manage the possible contingencies.Engineers recognize the human factor and design forit, but their preference is to make things as automaticas possible. Safety is built into the designs themselves.

When I asked an Egyptian Airlines pilot whether hepreferred Russian or U.S. planes, he answered imme-diately that he liked the U.S. planes because theRussian planes have only one or two back-up systems,while the U.S. planes have three back-up systems. In asimilar vein, during a landing at the Seattle airport, Ioverheard two engineers saying to each other that thecockpit crew was totally unnecessary. A computercould easily fly and land the plane.

In other words, a key theme in the culture of engi-neering is the preoccupation with designing humansout of the systems rather than into them. For example,the San Francisco Bay Area Rapid Transit (BART) usestotally automated trains. But the customers, not theoperators, objected to this degree of automation, forc-ing management to put human operators on eachtrain even though they had nothing to do except to re-assure people by their presence.

In the earlier example of the company introducingautomated machines tools into production processes,the engineers were very disappointed that the opera-tions of the elegant machine they were purchasingwould be constrained by the presence of more opera-tors than necessary, by a costly retraining program,and by management-imposed policies that had noth-ing to do with “real engineering.” In my own researchon information technology, I found that engineersfundamentally wanted the operators to adjust to thelanguage and characteristics of the particular comput-er system being implemented and were quite impa-tient with the operators’ “resistance to change.” Fromthe viewpoint of the users — the operators — notonly was the language arcane, but they did not con-


Assumptions of the Engineering Culture

• Engineers are proactively optimistic that they can and shouldmaster nature.

• Engineers are stimulated by puzzles and problems and arepragmatic perfectionists who prefer “people free” solutions.

• The ideal world is one of elegant machines and processesworking in perfect precision and harmony without human in-tervention.

• Engineers are safety oriented and overdesign for safety.

• Engineers prefer linear, simple cause-and-effect, quantitativethinking.

Akey theme in the culture of engineering is the preoccupation with

designing humans out of the systemsrather than into them.

sider the systems useful for solving the operationalproblems.13

Both operators and engineers often find them-selves out of alignment with a third critical culture,the culture of executives.

The Executive Culture The “executive culture” is the set of tacit assumptionsthat CEOs and their immediate subordinates shareworldwide. This executive worldview is built aroundthe necessity to maintain an organization’s financialhealth and is preoccupied with boards, investors, andthe capital markets. Executives may have other preoc-cupations, but they cannot get away from having toworry about and manage the financial survival andgrowth of their organization.14 (For the assumptionsof the executive culture, see the sidebar.)

What I have identified as the executive culture ap-plies particularly to CEOs who have risen throughthe ranks and been promoted to their jobs. Foundersof organizations or family members appointed tothese levels have different assumptions and often havea broader focus.15 The promoted CEO, especially,adopts the exclusively financial viewpoint because ofthe nature of the executive career. As managers rise inthe hierarchy, as their level of responsibility and ac-countability grows, they not only have to becomemore preoccupied with financial matters, but alsofind that it becomes harder to observe and influencethe basic work of the organization. They discoverthat they have to manage from afar, and that discov-ery inevitably forces them to think in terms of con-trol systems and routines that become increasinglyimpersonal. Because accountability is always central-ized and flows to the top of organizations, executivesfeel an increasing need to know what is going on,while recognizing that it is harder to get reliable in-formation. That need for information and controldrives them to develop elaborate information sys-tems alongside the control systems and to feel in-creasingly alone in their position atop the hierarchy.

Paradoxically, throughout their careers, managershave to deal with people and recognize intellectuallythat people ultimately make the organization run.First-line supervisors, especially, know very well howdependent they are on people. However, as managersrise in the hierarchy, two factors cause them to be-

come more “impersonal.” First, they become increas-ingly aware that they are no longer managing opera-tors, but other managers who think like they do, thusmaking it not only possible but also likely that theirthought patterns and worldview will increasingly di-verge from the worldview of the operators. Second, asthey rise, the units they manage grow larger and larg-er until it becomes impossible to personally know ev-eryone who works for them. At some point, they rec-


Assumptions of the Executive Culture

Financial Focus

•Executives focus on financial survival and growth to ensurereturns to shareholders and to society.

•Financial survival is equivalent to perpetual war with one’scompetitors.

Self-Image: The Embattled Lone Hero

•The economic environment is perpetually competitive and po-tentially hostile, so the CEO is isolated and alone, yet appearsomniscient, in total control, and feels indispensable.

•Executives cannot get reliable data from subordinates so theymust trust their own judgment.

Hierarchical and Individual Focus

•Organization and management are intrinsically hierarchical;the hierarchy is the measure of status and success and theprimary means of maintaining control.

•The organization must be a team, but accountability has to beindividual.

•The willingness to experiment and take risks extends only tothose things that permit the executive to stay in control.

Task and Control Focus

•Because the organization is very large, it becomes deperson-alized and abstract and, therefore, has to be run by rules, rou-tines (systems), and rituals (“machine bureaucracy”).

•The inherent value of relationships and community is lost asan executive rises in the hierarchy.

•The attraction of the job is the challenge, the high level of re-sponsibility, and the sense of accomplishment (not the rela-tionships).

•The ideal world is one in which the organization performs likea well-oiled machine, needing only occasional maintenanceand repair.

•People are a necessary evil, not an intrinsic value.

•The well-oiled organization does not need people, only activi-ties that are contracted for.

ognize that they cannot manage all the people direct-ly and, therefore, have to develop systems, routines,and rules to manage “the organization.” They in-creasingly see people as “human resources” to betreated as a cost rather than a capital investment.

The executive culture, thus, has in common withthe engineering culture a predilection to see people asimpersonal resources that generate problems ratherthan solutions. In other words, both the executiveculture and the engineering culture view people andrelationships as means to the end of efficiency andproductivity, not as ends in themselves. If we musthave human operators, so be it, but let’s minimizetheir possible impact on the operations and their costto the enterprise.

Dysfunctional Interactions among theThree Cultures

In many industries, there is enough initial alignmentamong the needs of the task as defined by the opera-tors, the needs of the engineers for reliable and effi-cient operations, and the needs of the executives forminimizing costs and maximizing profits so that thereare no problems. But when organizations attempt tolearn in a generative way, when they attempt to rein-vent themselves because the technologies and envi-ronmental conditions have changed drastically, thesethree cultures collide, and we see frustration, lowproductivity, and the failure of innovations to sur-vive and diffuse.

For example, in their research on nuclear plants,Carroll and Perin found that plant operators under-stood very well the interdependencies and interactionsof all the systems.16 They lived in an environment thathad its own ecology in which interdependence wasvisible and in which the management of interdepen-dencies through teamwork was crucial to safety andproductivity. But one or two levels above the plant,management saw only specific technical and financialissues, driven very much by the outside forces of theNuclear Regulatory Agency and their own worldviewas executives, a view that could best be described as a“machine bureaucracy,” while the operators’ world-view could better be described as a “sociotechnicalsystem.”

The plants were different in how they operated,

but each developed its own concept of how to im-prove its operations. Such improvement plans oftenrequired additional allocations of money for trainingand plant redesign, and also often required bendingsome formal rules and procedures mandated by theindustry and the government. When such require-ments were articulated, the engineering communityfocused primarily on finding standard solutions toproblems, preferably solutions free of human inter-

vention, and executive management focused primar-ily on money and cost control. The lack of align-ment among the three cultures often led to inactionand the continuation of practices that were viewed asless efficient or effective.

In some situations, like that in an airplane cockpit,the executive and operator cultures can collide in adrastically dysfunctional way. Blake’s research hasshown that some airline crashes are due to communi-cation failures in the cockpit resulting from obsessionwith rank and hierarchy.17 For example, in one crash afew miles short of the runway, the flight recorder re-vealed that the flight engineer had shouted for severalminutes that they were running out of gas, while thepilot, functioning as the CEO, continued to circleand tried to fix a problem with the landing gear. Whenthis situation was run in a simulator, the same phe-nomenon occurred; the pilot was so busy with his op-erational task and so comfortable in his hierarchicalexecutive position that he literally did not hear criticalinformation that the flight engineer shouted at him.Only when the person doing the shouting was a fel-low pilot of equal or higher rank did the pilot pay at-tention to the information. In other words, the hierar-chy got in the way of solving the problem. The engi-neering solution of providing more warning lights orsounds would not have solved the problem either, be-


Whereas the IT specialistsaw networking as away of eliminating

hierarchy, executives saw hierarchyas intrinsic to organizational control and coordination.

cause the pilot could easily rationalize them as com-puter or signal malfunctions.

At the boundary between the engineering and ex-ecutive cultures, other conflicts and problems of com-munication arise. In my research on executive viewsof information technology (IT) contrasted with theviews of IT specialists with an engineering mentality,the IT specialists saw information as discrete, pack-ageable, and electronically transmittable, while exec-utives saw information as holistic, complex, impre-cise, and dynamic.18 Whereas the IT specialist sawnetworking as a way of eliminating hierarchy, execu-tives saw hierarchy as intrinsic to organizational con-trol and coordination. Whereas IT specialists saw thecomputer and expert systems as the way to improvemanagement decision making, executives saw thecomputer as limiting and distorting thinking by fo-cusing only on the kinds of information that can bepackaged and electronically transmitted. And if exec-utives did buy into IT implementations for reasonsof cost reduction and productivity, they often man-dated it in a way that made it difficult for the opera-tors to learn to use the systems effectively because in-sufficient time and resources were devoted to therelearning process itself, as the earlier insurance com-pany example showed.

Of course, the way in which technology is used isinfluenced by the values and goals imposed by the ex-ecutive culture, as some of my examples have shown.And those values are sometimes more stable than thetechnological possibilities, causing technologies likeinformation technology to be underutilized from theviewpoint of the engineering culture.19 In the earlierexample, the engineers were thwarted by the execu-tive culture, and the solution that resulted from unionpressure reflected the executives’ short-run financialfears.

The lack of alignment among the executive, engi-neering, and operator cultures can be seen in other in-dustries such as health care in which the needs of theprimary care physicians (the operators) to do healthmaintenance and illness prevention conflicts with theengineering desire to save life at all costs and the exec-utive desire to minimize costs no matter how thismight constrain either the engineers or the operators.

In education, the same conflicts occur betweenteachers who value the human interaction with stu-

dents and the proponents of sophisticated computer-ized educational systems on the one hand and thecost constraints imposed by school administrators onthe other hand. If the engineers win, money is spenton computers and technologically sophisticated class-rooms. If the administrators win, classes becomelarger and undermine the classroom climate. In eithercase, the operators — the teachers — lose out, andhuman innovations in learning are lost.

Implications of the Three Cultures

There are several important points to note about thethree cultures. First, the executive and engineering cul-tures are worldwide occupational communities thathave developed a common worldview based on theireducation, their shared common technology, and theirwork experience. This means that even if an executiveor engineer in a given organization learns to think likean operator and becomes more aligned with the oper-ator culture, his or her eventual replacement will mostprobably return the organization to where it was. Thefield of organization development is replete with ex-amples of innovative new programs that did not sur-vive executive succession. In other words, the execu-tive’s or the engineer’s reference group is often outsidethe organization in his or her peer group, whose defi-nition of “best practice” may differ sharply from whatis accepted inside the organization. Executives and en-gineers learn more from each other than from theirsubordinates.

Second, each of the three cultures is “valid” from itsviewpoint, in the sense of doing what it is supposedto. Executives are supposed to worry about the finan-cial health of their organization, and engineers aresupposed to innovate toward the most creative people-free solutions. To create alignment among the threecultures, then, is not a case of deciding which one hasthe right viewpoint, but of creating enough mutualunderstanding among them to evolve solutions thatwill be understood and implemented. Too often intoday’s organizational world, either the operators as-sume that the executives and engineers don’t under-stand, so they resist and covertly do things their ownway, or executives and/or engineers assume that theyneed to control the operators more tightly and forcethem to follow policies and procedure manuals. In ei-


ther case, effectiveness and efficiency will suffer be-cause there is no common plan that everyone can un-derstand and commit to.

Third, both the executive and engineering cul-tures are primarily task focused and operate on theimplicit assumption that people are the problem, ei-ther as costs or as sources of error. In the case of theengineers, the assumption is already implicit in theireducation and training. The ultimately elegant solu-tion is one that always works and works automatical-ly, in other words, without human intervention. Inthe case of the executives, the situation is more com-plex. Either executives have come from the engineer-ing culture where people were not important in thefirst place, or they learned as they were promotedand began to feel responsible for hundreds of peoplethat they had to think in terms of systems, routines,rules, and abstract processes for organizing, motivat-ing, and controlling. And as they became chief exec-utives accountable to the financial markets and theirstockholders, they learned to focus more and moreon the financial aspects of the organization. Thegradual depersonalization of the organization andthe perception that employees are mostly a cost in-stead of a capital investment is thus a learned occu-pational response.

It is not an accident that chief executives tend toband together and form their own culture becausethey come to believe that no one except another chiefexecutive really understands the lonely warrior role.With that sense of aloneness come related assump-tions about the difficulty of obtaining valid informa-tion and the difficulty of ensuring that subordinatesdown the line will understand and implement whatthey are asked to do, leading ultimately to fantasies ofspying on their own organizations like the Caliph ofBaghdad who donned beggar’s clothes to mingleamong the people and find out what they were reallythinking. Even though the CEO’s immediate subor-dinates are humans, increasingly the chief executivesees them as part of a larger system that must be man-aged impersonally by systems and rules. CEOs oftenfeel strongly about not fraternizing with subordinatesbecause, if the organization gets into trouble, thosesubordinates are often the first to be sacrificed as evi-dence of “fixing” things.

Fourth, the engineering and executive cultures may

agree on the assumption that people are a problem,but they disagree completely on how to make organi-zations work more effectively. Executives recognizethat their world is one of imperfect information, ofconstant change, and of short-run coping while at-tempting to maintain a strategic focus. Engineers seekelegant permanent solutions that are guaranteed towork and be safe under all circumstances and, there-fore, typically produce solutions that cost much morethan the executives believe they can afford. So the ex-ecutives and the engineers constantly battle about howgood is good enough and how to keep costs downenough to remain competitive.

What is most problematic is that we have come toaccept the conflict between engineering and manage-ment as “normal,” leading members of each culture todevalue the concerns of the other rather than lookingfor integrative solutions that will benefit both. A fewcreative companies have sent engineers to talk to cus-tomers directly to acquaint them with business reali-ties and customer needs. Some executives aware of thisconflict involve themselves from time to time in oper-ations and product development so they do not losetouch with the realities and strengths of the other cul-tures. But this kind of remedy deals only with the or-ganizational level. The dilemma of twenty-first centu-ry learning is broader.

The Dilemma of Twenty-First CenturyLearning

Organizations will not learn effectively until they rec-ognize and confront the implications of the three oc-cupational cultures. Until executives, engineers, andoperators discover that they use different languagesand make different assumptions about what is impor-tant, and until they learn to treat the other cultures asvalid and normal, organizational learning efforts willcontinue to fail. Powerful innovations at the operatorlevel will be ignored, subverted, or actually punished;technologies will be grossly underutilized; angry em-ployees will rail against the impersonal programs ofreengineering and downsizing; frustrated executiveswho know what they want to accomplish will feelimpotent in pushing their ideas through complexhuman systems; and frustrated academics will won-der why certain ideas like employee involvement,


sociotechnical systems analyses, high-commitmentorganizations, and concepts of social responsibilitycontinue to be ignored, only to be reinvented undersome other label a few decades later.

First, we must take the concept of culture moreseriously than we have. Instead of superficially ma-

nipulating a few priorities and calling that “culturechange,” we must recognize and accept how deeplyembedded the shared, tacit assumptions of execu-tives, engineers, and employees are. We have lived inthis industrial system for more than a century andhave developed these assumptions as an effective wayto deal with our problems. Each culture can justifyitself historically, and each has contributed to thesuccess of the industrial system that has evolved.

Second, we must acknowledge that a consequenceof technological complexity, globalism, and universaltransparency is that some of the old assumptions nolonger work. Neither the executives nor the engineersalone can solve the problems that a complex socio-technical system like a nuclear plant generates. Wemust find ways to communicate across the culturalboundaries, first, by establishing some communica-tion that stimulates mutual understanding rather thanmutual blame.

Third, we must create such communication bylearning how to conduct cross-cultural “dialogues.”Recently, the concept of “dialogue” has substantiallyimproved our understanding of human thought andcommunication and promises to make some under-standing across cultural boundaries possible.20 If peo-ple from the different cultures will sit in a room to-gether, which is hard enough, they must reflectivelylisten to themselves and to each other, which is evenharder. Fortunately, the understanding of what it takesto create effective dialogues is itself coming to be bet-ter understood.

The engineering and executive cultures I have de-scribed are not new. What is new is that the operatorculture in all industries has become much more com-plex and interdependent, which has thrown it moreout of alignment with the other two cultures. Theimplication is that each community will have to learnhow to learn and evolve some new assumptions. Wehave directed our efforts primarily at the operationallevels of organizations and viewed the executive andengineering cultures as problems or obstructions,partly because they do not sufficiently consider thehuman factor. Yet these cultures have evolved andsurvived and have strengths as well as weaknesses.

The key to organizational learning may be in help-ing executives and engineers learn how to learn, howto analyze their own cultures, and how to evolve thosecultures around their strengths. These communitiesmay learn in different ways, and we will have to de-velop appropriate learning tools for each community.Learning may have to be structured along industrylines through consortia of learners rather than alongindividual organizational lines.21 And business andengineering education itself will have to examinewhether the assumptions of academics are evolving ata sufficient rate to deal with current realities.

We are a long way from having solved the prob-lems of organizational learning, but thinking aboutoccupational communities and the cultures of man-agement will begin to structure these problems so thatsolutions for the twenty-first century will be found. �

References1. C. Argyris, R. Putnam, and D. Smith, Action Science (San Francisco:Jossey-Bass, 1985);P. Senge, The Fifth Discipline (New York: Doubleday, 1990); andR. Beckhard and R.T. Harris, Organizational Transitions: ManagingComplex Change, 2nd ed. (Reading, Massachusetts: Addison-Wesley, 1987).2. G.L. Roth and A. Kleiner, “The Learning Initiative at the AutoCompany Epsilon Program” (Cambridge, Massachusetts: MIT Organi-zational Learning Center, working paper 18.005, 1996).3. G.L. Roth, “In Search of the Paperless Office” (Cambridge,Massachusetts: MIT, Ph.D. dissertation, 1993).4. R.J. Thomas, What Machines Can’t Do (Berkeley, California:University of California Press, 1994).5. D.M. McGregor, The Human Side of Enterprise (New York:McGraw-Hill, 1960).6. E.H. Schein and W.G. Bennis, Personal and Organizational Changethrough Group Methods: The Laboratory Approach (New York: JohnWiley, 1965); andR.R. Blake, J.S. Mouton, and A.A. McCanse, Change by Design(Reading, Massachusetts: Addison-Wesley, 1989).


Organizations will not learn effectively until they

recognize and confront the implications of the three

occupational cultures.

7. J. Van Maanen and S.R. Barley, “Occupational Communities: Cul-ture and Control in Organizations,” in B.M. Staw and L.L. Cummings,eds., Research in Organizational Behavior, vol. 6 (Greenwich, Connecticut:JAI Press, 1984).8. E.H. Schein, Organizational Culture and Leadership, second edition(San Francisco: Jossey-Bass, 1992a).9. E.H. Schein, “The Role of the Founder in the Creation of Organi-zational Culture” Organizational Dynamics, Summer 1983, pp. 13-28. 10. Van Maanen and Barley (1984).11. S. Zuboff, In the Age of the Smart Machine: The Future of Work(New York: Basic Books, 1988).12. G. Kunda, Engineering Culture: Control and Commitment in aHigh-Tech Corporation (Philadelphia: Temple University Press, 1992).13. E.H. Schein, “The Role of the CEO in the Management of Change:The Case of Information Technology,” in T.A. Kochan and M. Useem,eds., Transforming Organizations (New York: Oxford University Press,1992b). 14. G. Donaldson and J.W. Lorsch, Decision Making at the Top (NewYork: Basic Books, 1983).

15. Schein (1983).16. J. Carroll and C. Perin, “Organizing and Managing for Safe Produc-tion: New Frameworks, New Questions, New Actions” (Cambridge,Massachusetts: MIT Center for Energy and Environmental PolicyResearch, Report NSP 95-005, 1995).17. Blake et al. (1989).18. Schein (1992a, 1992b).19. L. Thurow, The Future of Capitalism (New York: William Morrow,1996).20. W.N. Isaacs, “Taking Flight: Dialogue, Collective Thinking, andOrganizational Learning,” Organizational Dynamics, Winter 1993,pp. 24-39; andE.H. Schein, “On Dialogue, Culture, and Organizational Learning,”Organizational Dynamics, Winter 1993, pp. 40-51.21. E.H. Schein, “Building the Learning Consortium” (Cambridge,Massachusetts: MIT Organizational Learning Center, working paper10.005, 1995).

Reprint 3811


Copyright © 1996 by the

SLOAN MANAGEMENT REVIEW ASSOCIATION.

All rights reserved.

Reprints

To order reprints, or requesta free reprint index, pleasecontact: Sloan Management ReviewReprints, E60-10077 Massachusetts AvenueCambridge, MA 02139-4307

Telephone: 617-253-7170Fax: 617-258-9739E-mail: [email protected]

Copyright Permission

To request permission tocopy one or more articles,please contact: Sloan Management ReviewPermissions, E60-100 77 Massachusetts AvenueCambridge, MA 02139-4307

Telephone: 617-258-7485Fax: 617-258-9739E-mail: [email protected]

MIT

three cultures of management: the key - prescient...

Documents