Research Paper No.

1847

Organizational Culture

Joanne Martin March 2004

RESEARCH PAPER SERIES

Organizational Culture Joanne Martin Stanford University

March, 2004

To be published in N. Nicholson, P. Audia, and M. Pillutla (Eds.), The Blackwell Encyclopedic Dictionary of Organizational Behavior, Second edition, Oxford, England: Basil Blackwell Ltd., in press.

Organizational culture Organizational culture is embedded in the everyday working lives of all cultural members. Manifestations of cultures in organizations include formal practices (such as pay levels, structure of the HIERARCHY, JOB DESCRIPTIONS, and other written policies); informal practices (such as behavioral norms); the organizational stories employees tell to explain how things are done around here; RITUALS (such as Christmas parties and retirement dinners); humor (jokes about work and fellow employees); jargon (the special language of organizational initiates); and physical arrangements (including interior decor, dress norms, and architecture). Cultural manifestations also include values, sometimes referred to more abstractly as content themes. It is essential to distinguish values/content themes that are espoused by employees from values/content themes that are seen to be enacted in behavior. All of these cultural manifestations are interpreted, evaluated, and enacted in varying ways because cultural members have differing interests, experiences, responsibilities and values. Culture consists of the patterns of meanings that link these manifestations together, sometimes in harmony, sometimes in bitter conflicts between groups, and sometimes in webs ofambiguity, paradox, and contradiction. For these reasons, it is much too simple to define culture in unifying, harmonious terms, for example, in terms of values that are espoused by management and supposedly shared by most employees.

The l980s brought a renaissance of interest in organizational culture. The resulting proliferation of research was accompanied by fundamental and fruitful disagreements about what culture is, whether it should be studied using quantitative or qualitative methods, if its content can be controlled by management, and whether a particular kind ofculture can result in stronger organizational performance. This dissension among cultural researchers, regarding such 2

fundamental issues, makes it difficult to define culture and summarize the results ofthis growing literature in terms of linear progress toward greater, widely accepted knowledge.

Theoretical Hairsplitting or D~fferencesof Consequence?

Given this dissension, it is reasonable to ask why applied researchers and practitioners should care about cultural research. Some managers have sought to replicate the supposedly strong cultures ofprofitable companies, while others have tried engineering values to generate COMMITMENT to a philosophy of management, in the hopes of increasing loyalty, PRODUCTIVITY, or profitability. Some top executives have sought to create a culture cast in their own image, to perpetuate an organizational culture reflecting their own personal values, thereby attempting to achieve an organizational form of immortality. Usually practitioners respond to promises of easy solutions and quick fixes with well-deserved skepticism, but organizational culture, at first, seemed immune from such skepticism. Later disillusionment set in and many dismissed culture as yesterdays fad. The seesaw between credulity and disillusionment has caused considerable waste oftime and money. Practitioners need to know enough to judge
--

with appropriate skepticism

--

what

researchers and strategic advisors are focusing on and what they are ignoring. Without some understanding ofwhy researchers come to different conclusions about how cultures change and whethertransitions can be managed, it is impossible to judge whetherthe results of a particular study are in some sense valid and whether they have practical applications in a given cultural context. For all these reasons, theoretical differences of opinion are not just hairsplitting debates

of interest only to ivory-tower scholars.

Overview

Three theoretical traditions can be used to describe most organizational culture research to date: the Integration, Differentiation, and Fragmentation perspectives (Martin, 1992). This essay defines the premises of each perspective, summarizes results of representative studies, identifies problems inherent in each viewpoint, and reviews multiple-perspective studies that transcend some ofthe difficulties associated with single-perspective studies (Martin, 2002).

Integration Perspective

Of the three perspectives that have come to dominate organizational culture research, the integration perspective is the most popular and, ironically, the least well supported empirically. Integration studies of culture implicitly or explicitly assume that a culture is characterized by consistency, organization-wide consensus, and clarity. According to Integration studies, consistency occurs because people at the higher levels of an organization articulate a set of espoused values, sometimes in the form of a mission statement; these values are then reinforced by a variety of cultural manifestations that allegedly generate organization-wide value consensus. In Integration studies, there is clarity concerning what the organizational values are and should be, what behaviors are preferable, and what a particular story or ritual means. Organizational members apparently know what they are to do, and they agree why it is worthwhile to do it. In

the few instances when ambiguity is acknowledged, or subcultural differences emerge, they are described as not part of the culture or as evidence of a failure to achieve a strong culture. For example, Schein (1985) focused attention on individual corporate leaders who attempt to generate company-wide consensus regarding their personal values and corporate goals through a wide range of consistent corporate policies and practices. Using a similarly functionalist approach, Collinson and Porras (2002) assume an Integration view of culture and argue that such strong (integrated) cultures are a key to firm profitability (see also Kotter, 1992). In contrast to such functionalist research, other Integration studies take a more symbolic approach (Schultz and Hatch, 1996). For example, Barley (in Frost et al., 1991) described how Funeral Directors use a series ofpractices and rituals (e.g., putting make-up an a corpse, changing sheets on a death bed, etc.) to reinforce the idea that death can be life-like. In most but not all Integration studies, culture supposedly originates in the values articulated by top management; these values are then reinforced by selectively hiring people with similar priorities and by attempting to socialize new employees thoroughly (see SOCIALIZATION). The Integration perspective conceptualizes cultural change as an organization-wide cultural transformation, whereby an old unity is replaced
--

hopefully

--

by a new one. In the interim,

conflict and ambiguity may occur, but these are interpreted as evidence of the deterioration of culture before a new unity is established. Much of the research that initially generated the renaissance ofinterest in culture, particularly in the United States, falls within the Integration perspective. This view of culture still has some acceptance, in part because such a harmonious and clear environment is attractive, particularly to executives who would like to think that they could create a vision and enact a culture that would inspire such consensus.

For example, most Integration research takes a specialist approach, studying just one (or at most a few) manifestations, usually measures of agreement with a set of espoused values or selfreports ofbehavioral norms. This limitation creates problems. Self-reports of values and nonns are especially liable to reflect HALO EFFECTS related to overall JOB SATISFACTION, social desirability of particular responses, and IMPRESSION MANAGEMENT considerations. Meanings associated with a small sample ofmanifestations may not be consistent with meanings associated with the full range of manifestations of a culture. In addition, most Integration studies rely primarily on the views of managerial and professional employees, although it cannot be assumed that the views of this minority of powerful individuals are shared by all employees, particularly given the likelihood of differences of opinion across levels of a hierarchy. Finally, employees behavioral COMPLIANCE to top managements preferences or policies cannot be taken as evidence of their personal approval of values, interpretations, or norms. Thus, Integration studies often take evidence of a limited subset of a cultures manifestations or a small and unrepresentative sample of its members, assume consistency and consensus, and generalize from these limited findings to the whole culture, as perceived by all or most of its members. This part-whole error characterizes much Integration research (Martin, 2002). Because of part-whole errors, Integration studies run the risk oftautology: culture is defined in terms of consistency, consensus, and clarity, and data regarding any manifestations, interpretations, or cultural members that do not conform to this view are excluded as not part ofthe culture or dismissed as evidence of a weak culture. Because ofpart-whole errors, and because so many other variables (economic, marketing, and strategic) affect firm performance, oft-repeated claims that strong cultures are a key to improved organizational profitability, should be regarded as, at best, unproved (Siehi & Martin,

1990). Many critics of Integration research make a stronger claim: that it is highly unlikely that any organizational culture, studied in depth, would exhibit the consistency, organization-wide consensus, and clarity that Integration studies have claimed to find (e.g., Alvesson, 2002; Martin, 1992, 2002; Turner, 1986). Thus, Integration studies offer managers and researchers a seductive promise of harmony and value homogeneity that is empirically unmerited and unlikely to be fulfilled.

Differentiation Perspective

Differentiation studies describe organizations as composed of overlapping, nested subcultures that coexist in relationships of intergroup harmony, CONFLICT, or indifference. For example, in a Differentiation study, Bartunek and Moch (in Frost et al, 1991), show how five subcultures in a food production firm reacted differently to managements imposition of a QUALITY OF WORKING LIFE intervention Top management was primarily concerned with control. In-house consulting staff members were cooperative. The management of the local plants where the program was implemented was paternalistic, using imagery of employees as children to managerial parents. Line employees exhibited a dependent reaction, following managements preferences. Machinists, historically an active, independent, and comparatively well-paid group, actively resisted the intervention. Thus, in Differentiation studies, to the extent that consensus exists, it exists within subcultural boundaries. A hierarchical alignment of subcultures is also evident in Van Maanens (in Frost et al., 1991) study ofride operators at Disneyland. Food vendors (pancake ladies and coke blokes) were

allocated to the bottom ofSTATUS ranking and male operators of yellow submarines and jungle boats shared high status. Tension between ride operators, customers, and supervisors was evident, as ride operators arranged for obnoxious customers to be soaked with water when submarine hatches opened and supervisors were foiled in their constant attempt to catch operators breaking rules. In Youngs (Frost et al, 1991) study of bag ladies in a British manufacturing plant, tensions between management and labor were evident, and the younger and older workers fissioned into different subcultures. As these examples indicate, subcultures often appear along lines of functional, occupational, and hierarchical DIFFERENTIATION. Also evident in these studies is a subtext: many of these subcultural differences also reflect demographic differences (e.g., class, RACE, ethnicity, AGE, and GENDER), creating working environments that are racially segregated (Cox, 1993) and/or deeply gendered (e.g., Gherardi, 1995; Aaltio and Mills, 2002). Inconsistency across cultural manifestations is also evident in Differentiation studies. For example in the food production firm studied by Bartunek and Moch, top management said one thing to employees, and did something different. At Disneyland, ride operators appeared to conform to managements rules, while doing what they pleased. In a particularly detailed examination ofthe effects of such inconsistencies on individuals, Kunda (1991) studied engineers reactions as they conformed to a company ritual designed to exhibit commitment to supposedly shared company values. During moments of ease while off stage, the engineers usedhumor and sarcastic side remarks to express their disapproval, skepticism, or ambivalence. As these examples indicate, espoused values, behavior mandated by formal policies, and informal norms are often observed to be inconsistent. Whereas some Differentiation studies describe subcultures in functionalist terms, as reflections of occupational socialization, other

Differentiation studies take a more critical approach (Alvesson, 2002; Willmott, 1993), conceptualizing culture as a partially successful attempt by management to exercise hegemonic control over lower ranking employees, eliciting a mix of compliance and resistance. To summarize, a Differentiation study includes evidence of inconsistency between one cultural manifestation and another. Consensus is evident, but only within the boundaries of a subculture. Within a subculture all is clear, but ambiguities do appear at the interstices where one subculture meets another. When viewed from the Differentiation perspective, the organization is no longer seen as a cultural monolith; instead, it is a collection of subcultures. Some ofthese subcultures enthusiastically reinforce the views of the top management coalition or operate cooperatively with each other. Others become pockets of ignorance or resistance to top management initiatives. From the Differentiation perspective, change is localized within one or more subcultures, alterations tend to be incremental, and change is triggered (if not determined)by pressures from an organizations environment (see ORGANIZATION AND ENVIRONMENT). That environment is likely to be segmented, so different subcultures within the same organization experience different kinds and rates of change. Of the three perspectives, the Differentiation viewpoint is most congruent with research that emphasizes environmental determinants of organizational behavior. As is the case with Integration research, the methodological choices made in Differentiation studies partially determine what results are found. For example, as can be seen in the Disneyland and bag lady studies, there is a tendency for Differentiation research to focus on relatively lowranking employees or first-line supervisors
--

people who are less likely to share the views of top

management. And within any subculture, there is a tendency to focus on the ways subcultural

members share the same views, rather than on the ways subcultural members views differ or what they find ambiguous. As a result, Differentiation studies do not distance themselves far enough from the oversimplifications and distortions ofthe Integration view; within a subculture, consistency, consensus, and clarity still predominate, and ambiguities are relegated to the interstices among subcultures.

Fragmentation Perspective

In Fragmentation studies of culture, claims of clarity, consistency, and consensus are shown to be idealized oversimplifications that fail to capture the confusing complexity of contemporary organizational functioning. The Fragmentation perspective offers a quite different alternative. Rather than banning ambiguity from the cultural stage (the Integration view) orrelegating ambiguity to the interstices between subcultures (the Differentiation view), Fragmentation studies see ambiguity as the defining feature of cultures in organizations. In these studies, ambiguity is defined to include multiple meanings, paradox, irony, and inescapable contradictions. Such ambiguity pervades all but the most routine and trivial aspects of organizational functioning. Therefore, the meanings that different cultural members attach to particular cultural manifestations are neither clearly consistent nor clearly in conflict. There are many plausible interpretations of any one issue or event, making the idea of a single clear, shared cultural reality highly unlikely. To the extent that consensus exists, it is issue-specific and transient: problems or issues get activated, generate positive and negative reactions, and then fade from attention as other issues take center stage, creating temporary, issue-specific networks

10

of connection that disappear and reconfigure themselves in a constant flux. From the Fragmentation perspective, culture looks less like a monolith, and less like a collection of subcultural islands, and more like a room full of spider webs, constantly being destroyed and rewoven. For example, Feldman (in Frost et a!., 1991) studied federal policy analysts who analyze policy options, write reports that never get read, and if they are read, probably never will impact policy decisions. Robertson and Swan (2003) studied highly educated consultants working within a knowledge-intensive firm where project work was inherently fluid, complex, and uncertain, embracing ambiguity. Meyerson (in Frost et al., 1991) studied the ambiguities of social work. Goals were unclear; there was no consensus regarding the appropriate means to achieve those goals; success was hard to define and even harder to assess. For social workers, ambiguity was the salient feature of their working lives and any cultural description that excluded ambiguity would be dramatically incomplete. Whereas the Fragmentation studies described above focused on occupations coping successfully with ambiguous work, Weick (in Frost et al., 1991) has used the Fragmentation perspective in a context where the effects of ambiguities were less benign
--

a foggy airport in Tenerife where

one airplane was attempting to land while another waited to take off. Weick focused on talk among pilots, cockpit crews, and air traffic controllers, as they coped with the complexities of making themselves understood across barriers created by differences in native language, occupational and national prestige, and incompletely shared knowledge. Hundreds ofpassengers died in the ensuing crash, making this study a powerful illustration ofthe conclusion that most Fragmentation studies draw: that an understanding of ambiguities should be a central component of any cultural study that claims to encompass the full range of cultural members working lives.

11

In Fragmentation studies of culture, POWER is diffused broadly at all levels of the hierarchy and throughout the organizations environment. Culture has no specific point of origin; fleeting affinities are issue-specific. Change is a constant flux, rather than an intermittent interruption in an otherwise stable state. Change is largely triggered by the environment or other forces beyond an individuals control, so that Fragmentation studies of change offer few guidelines forthose who would normatively control the change process. The methodological choices made in Fragmentation research enable these kinds of conclusions to be drawn. For example, Fragmentation studies tend to focus on highly ambiguous occupations (i.e., social worker, policy analyst) and contexts (e.g., cross-national communication, literally in the fog). As noted regarding research conducted from the other two perspectives, Fragmentation studies exhibit a form ofmethodological tautology: these researchers define culture in a particular way and then find what they are looking for.

Advantages of Using Multiple Perspectives in a Single Study

These problems of methodological tautology, and the theoretical blind spots associated with any single perspective, can be minimized if a single cultural context is studied from each of the three perspectives, permitting a more complete understanding to emerge. While most studies utilize only one of the three perspectives, more recent research indicates that any organizational culture contains elements congruent with all three viewpoints (Martin, 2002). If any organization is studied in enough depth, some issues, values, and objectives will be seen to generate organization-wide consensus, consistency, and clarity (an Integration view). At the same time, 12

other aspects of an organizations culture will coalesce into subcultures that hold differing opinions about what is important, what should happen, and why (a Differentiation view). Finally, some problems and issues will be ambiguous, in a state of constant flux, generating multiple, plausible interpretations (a Fragmentation view). A wide range of organizational contexts have been examined using the three-perspective framework, including studies of a temporary educational organization for unemployed women in England, a newly privatized bank in Turkey, the problem of truancy in an urban high school in the United States, changing organizational cultures in the Peace Corps/Africa, a search for a university provost, and professional subcultures in an Australian home care service (e.g., Bloor and Dawson, 1994; Eisenberg, Murphy, and Andrews, 1998).

A Subjective Approach

Sometimes the three perspective framework is mistakenly taken to mean that some organizations will be correctly described by an Integration viewpoint, while other contexts may better fit the Differentiation or Fragmentation perspectives. This is a misunderstanding. Although these perspectives are empirically derived from the perspectives of cultural members, they are not objective representations of the views of cultural members. Who a researcher is affects what he or she sees and what questions she or he seeks to address. And the identity of a cultural member affects what information they are exposed to, what information they absorb, and what reactions they exhibit. The measurement, collection, and interpretation of qualitative or quantitative cultural data are inevitably affected by subjective factors, whether quantitative or qualitative

13

methods are used (see RESEARCH METHODS). This subjective orientation is counter-intuitive for many. Often one theoretical perspective, labeled the home viewpoint, is easy for cultural members and researchers to see, while the other two perspectives can be more difficult to access. The harder it is to see applicability of a particular perspective, the more likely it is that, in changed circumstances, insights from that perspective may be crucial for organizational survival. For example, if most cultural members would like to see their organization as strongly Integrated, perhaps around the personal values of a well-respected leader, they may repress, or avoid seeing, evidence of any kind of subgroup conflict. If then that leader were to leave the organization, subcultural conflicts might surface in a totally unanticipated way. As this example illustrates, awareness of the perspectives that are less easily seen may provide a key to anticipating, or at least understanding, organizational change. This brief review has explained why and how many cultural researchers have disagreed about such fundamental ideas as to what culture is, how it should be studied, and whether it generates (and reflects) harmony, conflict, and/or ambiguity. These disagreements have been fruitful. In part because ofthe efforts of culture researchers, qualitative methods are now more broadly accepted in organizational studies. Novel applications of quantitative methods have been used to study aspects ofculture (e.g., Kilduff and Corley in Ashkanasy et al., 2000, on network measures of the three theoretical perspectives). New approaches to writing about cultural theory have been developed (e.g., Czarniawska, 1999), and innovative theoretical approaches are being explored (e.g., see Strati, 1999, on aesthetics of organizing). Work in related areas of inquiry, such as organizational identity, international cultures, industry cultures, and organizational climate, have moved from single-perspective, Integrationist assumptions, to more complex, multi-perspective

14

views (e.g., Ashkanasy et a!., 2000; Hatch and Schultz, 2000; Sackman, 1997). Post-structural cultural theorists (e.g., Jeffcutt, 1995; Wilmott, 1993) (see POSTMODERNISM), have shown how cultural studies can reveal the hidden biases and the silenced voices in organizational accounts, broadening the scope of our inquiries to include a wider range of contradictory theories and a greater number of cultural members viewpoints, without coming to any single conclusion about the superiority or predominance ofany ofthese theoretical views. As hoped by its early proponents in the 1980s, the study of culture has brought fresh ideas into organizational studies, showing how theoretical and methodological dissension can breed new insights.

See also: Organizational Identity, National Culture, Industry culture, Organizational Climate,

Symbolism; Group culture

15

References

Aaltio, I. & Mills, A. (Eds.), (2002). Gender, identity and the culture oforganizations. New York: Routledge.

Alvesson, M. (2002). Understanding organizational culture. London: Sage.

Ashkanasy, N., Wilderom, C., & Peterson, M. (2000). Handbook oforganizational culture and climate. Thousand Oaks, CA: Sage.

Bloor, G. & Dawson, P. (1994). Understanding professional culture in organization context. Organization Studies, 15, 275-295.

Collins, J. & Porras, J. (2002). Built to last: Successful habits ofvisionary companies. New York: Harper Business.

Cox, T, (1993). Cultural diversity in organizations: Theory, research, and practice. San Francisco: Berrit-Koehler.

Czamiawska, B. (1999). Writing management. Organization theory as a literary genre. London: Oxford University Press. Eisenberg, E., Murphy, A., & Andrews, L. (1998). Openness and decision making in the search for a university provost. Communication Monographs, 65, 1, PA Research II Periodicals.

Frost, P., Moore, L., Louis, M., Lundberg, C. & Martin. J. (Eds.), (1991). Reframing organizational culture. Newbury Park, CA: Sage.

Gherardi, S. (1995). Gender, symbolism, and organizational cultures. London: Sage.

Hatch, M.J. & Schultz, M. (2000). Scaling the Tower of Babel: Relational differences between identity, image, and culture in organizations. In M. Schultz, M. J. Hatch & H. H. Larsen (Eds.), The expressive organization, (pp. 11-35). London: Oxford University Press.

Jeffcutt, P. (1995). Culture, and symbolism in organizational analysis. Newbury Park, CA: Sage.

Kilduff, M. & Corley, K. (2000). Organizational culture from a network perspective. In Ashkanasy, N., Wilderom, C., & Peterson, M. (Eds.), Handbook of organizational culture & climate, (pp. 211-221). Thousand Oaks, CA: Sage.

Kotter, J. (1992). Corporate culture andperformance. New York: Free Press.

Kunda, G. (1991). Engineering culture: Control and commitment in a high-tech corporation. Philadelphia: Temple University Press.

Martin, J. (1992). Cultures in organizations: Three Perspectives. New York: Oxford University Press.

Martin, J. (2002). Organizational culture: Mapping the terrain. Newbury Park, CA: Sage.

Robertson, M. & Swan, J. (2003). Control

what control? Culture and ambiguity within a

knowledge intensive firm. Journal ofManagement Studies, 40, 83 1-858.

Sackman, 5. (1997). Cultural complexity in organizations: Inherent contrast and contradictions. Thousand Oaks, CA: Sage.

Schein, E. (1985). Organizatinal culture and leadership. San Francisco, CA: Jossey-Bass.

Schultz, M. & Hatch, M. J. (1996). Living with multiple paradigms: The case ofparadigm interplay in organizational culture studies. Academy ofManagement Review, 21, 529-557.

Siehi, C. & Martin, J. (1990), Organizational culture: The key to financial performance? In B. Schneider (Ed.), Organizational climate and culture, (pp.241-281). San Francisco: Jossey-Bass.

Strati, A. (1999). Organization and aesthetics. London: Sage.

Turner, B. (1986). Sociological aspects of organizational symbolism. Organizational Studies, 7, 101-115.

Wilmott, H. (1993). Strength is ignorance; slavery is freedom: Managing culture in modern organizations. Journal ofManagement Studies, 30, 5 15-552.