This file is also available in Adobe Acrobat PDF format
CULTURE IN ORGANIZATIONS
Consider the culture of the organization that currently employs you. Jot down a few words or phrases to describe its essential features. In doing this exercise, ask yourself about the specific beliefs embodied in the culture, and about the particular practices and behaviors regarded as appropriate and inappropriate in the organization.
Now think about how the culture was yesterday, how it was last week, how it was last month, how it was last year, and how it was five years ago, if you know. Odds are, if your organization is typical, that while some parts of the culture may have changed over the longer periods, the cultural essence of the organization has shifted little, if at all. Yet many other things about the organization likely did change: employees came and went, jobs changed hands with regularity, top managers moved around, suppliers were shifted, customers left and new ones were cultivated, competitors emerged and fell by the wayside, and the economy exerted itself in different ways.
Cultural stability in organizations, even over very long periods of time, is a fact of social life. It may not be a universal fact, as we can all point to particular organizations where significant cultural change has occurred. But it is a highly prevalent fact, and an amazing one at that. Given the vast amount of demographic and other change that occurs in the typical organization, it seems reasonable to expect some kind of cultural drift or erosion as the normal state of affairs--after all, culture resides in and gets reproduced by the specific persons present, and when they come and go regularly, it would seem that cultural elements might too. Instead, most social scientists would contend that organizational cultural usually persists over time with great force, even in the midst of strong pressures apparently to the contrary. The famous phrase of nineteenth-century French novelist Alphonse Karr (1849), "The more things change, the more they remain the same," seems an apt way to describe the cultures of many modern organizations.
This book investigates the amazing social fact of cultural persistence. We aim to understand when cultures persist and when they change. We pay special attention to dynamics, especially the demographic dynamics of personnel. This is an unusual approach to the study of organizational culture, and its potential value may not be immediately obvious. Accordingly, we start in this first chapter by painting the backdrop to our research program and its theoretical orientation, by explaining how it connects with the extant literature on culture and what we think it has to offer.
The chapter begins with a short observation about the state of cultural analysis in social science and then moves quickly to a brief overview of the more contemporary literature on culture in organizations. Following that, in the next two sections, we draw a distinction between two general approaches to cultural analysis, the popular content-based approach that focuses on the actual substance of cultural symbols and practices, and the lesser-known distributive approach that concentrates on the variability of cultural interpretation across the members of a population. We then explain how we use aspects of both approaches to develop a limited synthetic approach to study and model organizational culture. In our approach, we conceptualize organizational culture as comprising the distribution of managerially preferred cultural content across individuals within the organization and across time. Viewing organizational culture in this way gives us a general model for analyzing the dynamics and morphology of cultural systems. Next, we hint at the power of our approach by offering summary glimpses of some of the findings presented in the book. Finally, we conclude with an overview of the book and some suggestions for how to read it.
1.2 CULTURE IN SOCIAL SCIENCE
Although probably controversial, a case could be made that the scientific study of the social world began with early investigations of culture. It would start by making clear, if it is not so already, that ethnography qualifies as a scientific method. It would proceed by claiming that the dispassionate narrative accounts of premodern societies found in the journals of explorers and natural historians constitute a kind of early ethnography. The case would then be finalized by showing that these kinds of activities were earlier and more widespread in the literature than the statistical analysis of human mortality conducted by John Graunt (1662), commonly cited as marking the origins of social science.
Whatever the lineage of social science, professional accounts of cultures can be found in the journals of geography and natural history from at least the early nineteenth century. For instance, early issues of the Journal of the Royal Geographic Society of London contain numerous descriptive accounts of adventurers' and naturalists' explorations, including descriptions of native customs, religious practices, laws, polity, organization, and behavior. Heavily descriptive and often narrative, these reports typically have the tone of objective and systematic exposition, even when making moralistic judgments. At times, they are even somewhat analytical, as in the ethnographic article of J. C. Prichard (1839, 200) on "high Asia" which focuses on language differences and their origins:
On comparing the Mongolian, the Mandschu and Turkish languages, in relation to their grammatical structure, a series of remarkable analogies is discovered. The resemblance of the Mongolian and the Mandschu is much closer than between either of them and the western dialects of the Turkish language. These dialects, especially that of the Osmanli, have been subjected to a foreign influence and culture, the result of intercourse with Persians and Arabians, and the introduction of Mohammedan literature among the Turks.
Discussion of culture can be found in the early journals and books of each of the modern social science disciplines, including economics, sociology, and political science. But, of course, nowhere is culture's early prominence greater than in anthropology, where the discipline essentially constituted itself around the concept. From its foundation, anthropology's concern with premodern societies showed a special fascination with cultural phenomena (Tylor, 1871).
Given such early interest, it might seem natural to expect that modern social science would have either reached a consensus about how to regard culture scientifically or eschewed use of the concept altogether. But neither characterization accurately describes the current situation. Remarkably, ample dissensus still exists in many quarters over what culture is, what it means to social life, and how to analyze it (see Brumann, 1999, and associated commentary).
Despite disagreement about such fundamental questions, culture maintains a grand presence on the stage of contemporary social science. Indeed, the last decade or so has witnessed a resurgent, and perhaps deeper, interest in culture (often dubbed "the cultural turn"), initiated by a widespread belief about its importance in explaining social life (see Lie, 1997; Mohr, 1998; Perrow, 2000). Current fascination with culture as an explanation of behavior is so widespread across areas and disciplines that some social scientists wonder if anthropology has any distinctive approach left for analyzing the social world.
1.3 CULTURE IN ORGANIZATION THEORY
Cultural studies in the field of organization theory show several notable parallels with its more general development. As with the basic disciplines, one can find cultural issues raised and discussed seriously in some foundational texts.
For instance, the pre-paradigmatic classic book of Barnard (1938) does not use the word culture, but it is replete with references to informal organization and cooperation.1 Later, as the nascent field of organization theory began to develop, the eminent sociologist Talcott Parsons (1956: 67) wrote in the first issue of Administrative Science Quarterly:
Like any social system, an organization is conceived as having a describable social structure. This can be described and analyzed from two points of view, both of which are essential to completeness. The first is the cultural-institutional point of view which uses the values of the system and their institutionalization in different functional contexts as its point of departure; the second is the group or role point of view which takes suborganizations and the roles of individuals participating in the functioning of the organization as its point of departure . . . primary attention will be given to the former.
Although early organizational theorists (most notably the human relations and institutional schools) developed ideas about informal structures, as suggested by Barnard, Parsons, and others, a specific focus on organizational culture did not come until around 1980.2 During this period, large Japanese firms rose rapidly in prominence in many world markets, making analysts acutely aware of their apparently homogeneous cultures. A flurry of studies attributed the rise of Japanese corporations to their peculiar cultural designs, touting the potential performance benefits of homogeneous cultures and offering recipes for cultural management in organizations (Ouchi, 1981; Deal and Kennedy, 1982; Peters and Waterman, 1982).
The surge of interest in organizational culture rapidly developed to include more tempered scholarly analyses. Among other things, studies of culture pointed to the potential liabilities of cultural homogeneity (e.g., inertia, discrimination, and the suppression of creativity), the difficulties of analyzing culture (e.g., its multidimensionality, ambiguity in interpretation, and incomparability), and the subtle complexities of attempting to manage a culture (e.g., the reactive nature of human behavior and the development of subversive subcultures). Studies also demonstrated, however, that cultures vary across organizations and that these differences affect behavioral outcomes (Hofstede, 1980). Moreover, few, if any, analysts argued that culture was irrelevant or could be ignored in organizational analyses.
Given the short but intense history of theory and research on organizational culture, an outside observer might anticipate that analysts would have either settled basic questions about organizational culture or dismissed it as a nonviable path for investigation. Again, as with culture in anthropology and the basic disciplines, this view would be wrong. Theories, methods, and even conceptual approaches to the issue reflect stark and often inconsistent differences. Dissensus is so strong that Martin (2002) refers to intellectual exchanges in the area as "culture wars."
Yet almost no one denies the potential importance of organizational culture. Social scientists may argue about many aspects of organizational culture. But there is little disagreement about its potential ability to control behavior or its role in assisting members to interpret and derive meaning from their experiences while inside particular organizations. Analyses of culture can be found today in most issues of the leading academic journals on organizations and in every major textbook used in organizational courses throughout the world. Culture receives credit for the spectacular long-term success of companies such as Nordstrom, Southwest Airlines, and Hewlett-Packard (Tushman and O'Reilly, 1997) and blame for tragedies such as the Columbia space shuttle accident at the National Aeronautics and Space Administration (NASA; Columbia Accident Investigation Board, 2003). Most experienced managers will also tell you that culture is one of the great intangibles in managing a business or complex organization, especially the modern multinational corporation that spans different businesses and countries. For example, CEO Lou Gerstner (2002: 182) of IBM writes that "culture isn't just one aspect of the game--it is the game."
1.4 THE CONTENT APPROACH TO CULTURE
Despite the many different approaches to culture, social scientists typically maintain a strong focus on what we call the content of culture. By content, we refer to the substance of the culture--the actual sets of ideas, beliefs, values, behaviors, symbols, rituals, and the like that characterize the culture. In various studies, content might be described in terms of specific cultural features, such as teknonymy, the practice whereby individuals are referred to not by their personal names but as derived from their offspring's names (e.g., Lishan's mother).3 Or, content might be characterized in more general terms, such as collectivist or individualistic (Hofstede, 1980).
Research organized around cultural content typically tries to discover or identify the important elements of the culture. It then attempts to understand how members of the culture interpret these elements and to explain how content shapes social interaction and behavior. The usual study also assumes that for a cultural element to be important, it must be more or less shared, a coherent collective phenomenon. Cultural analysis thus tends to concentrate on the shared content, an approach consistent with social science's general orientation of analyzing relationships between the means of variables.
Much theoretical debate over culture involves disputes about what exactly constitutes the important general elements of cultural content, the specific ways that they control or guide behavior, and their interrelationships with each other.4 Perhaps the most fundamental difference lies in the distinction between the materialist view of cultures as adaptive systems and the ideational view of cultures as cognitive, structural, or symbolic systems (see Keesing, 1974). In the materialist conception, cultures are viewed broadly as "systems (of socially transmitted behavior patterns) that serve to relate human communities to their ecological settings. These ways-of-life communities include technologies and modes of economic organization, settlement patterns, modes of social grouping and political organization, religious beliefs and practices and so on" (Keesing, 1974: 75). By this view, the material conditions of life constrain and shape the form of culture that develops and adapts as conditions change. In the ideational view, ideas define the content: "A society's culture consists of whatever it is one has to know or believe in order to operate in a manner acceptable to its members . . . culture is not a material phenomenon, it does not consist of things, people, behavior or emotions . . . it is the forms of things that people have in their mind, their models for perceiving, relating and otherwise interpreting them" [Goodenough (1957: 167)]. An attempt to elaborate the ideational elements of content yielded the following phenomena: forms, propositions, beliefs, rules, values, recipes, routines, customs, meaning, and function (Goodenough, 1981).
Contemporary debate over cultural content has shifted somewhat, with the materialist view receiving far less attention. According to Hannerz (1992), current studies of culture primarily address one of two general dimensions of content: (1) ideas and modes of thought, or (2) forms of externalization. He considers the first to be primary in anthropological study:
Understanding structures of knowledge, belief, experience, and feeling in all their subtlety, and in their entire range of variations at home and abroad, is reasonably enough the core of cultural analysis. Secondarily, perhaps, anthropologists have occupied themselves with . . . the way meaning finds expression in a somewhat limited range of manifest forms--speech, music, graphic arts, or certain other communication modes. (Hannerz, 1992: 9-10)
Criticisms of the culture concept commonly consist of criticisms of the content approach, especially its emphasis on shared content. In summarizing the critiques, Brumann (1999: S1) notes that "the major concern of the sceptical discourse on culture is that the concept suggests boundedness, homogeneity, coherence, stability, and structure whereas social reality is characterized by variablity, inconsistencies, conflict, change and individual agency." Rodseth (1998: 55) sees the critics as claiming that culture "is tainted by essentialism, by holism, by ahistoricity; the concept inevitably suggests that human variation comes packaged in neatly bounded systems of changing forms: primordial, homogeneous and overly coherent."
Organizational Cultural Content
Content issues also dominate theory and research on organizational culture. Consider, for instance, definitions of culture from three sets of organizational researchers with different disciplinary backgrounds:
- "Culture: a pattern of basic assumptions--invented, discovered, or developed by a given group as it learns to cope with its problems of external adaptation and internal integration--that has worked well enough to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think and feel in relation to those problems" (social psychologist Edgar Schein, 1985: 9).
- "We define culture as a system of shared values (that define what is important) and norms that define appropriate attitudes and behaviors for organizational members (how to feel and behave)" (psychologically oriented management theorists Charles O'Reilly and Jennifer Chatman, 1996: 160).
- "Culture refers to the commonly held values and beliefs of individuals within the organization and, accordingly, the evaluative criteria used to make both large and small decisions" (economists Garth Saloner and Andrea Shepard and sociologist Joel Podolny, 2001: 76).
While differing in the particular cultural elements highlighted (basic assumptions, shared values and norms, or commonly held beliefs), each of these definitions directs us to look at the patterns of more or less shared content found in an organization. Typical aspects of cultural content would include beliefs, values, norms, special languages, and mental maps that make sense of the organization.
As with culture generally, theory and research on organizational culture vary greatly in terms of the specific content emphasized. For example, Schein (1992) identifies three levels of cultural content: (1) basic assumptions, (2) values, and (3) artifacts and creations, including technology, art, and visible and audible behavior patterns. Likewise, Trice and Beyer (1993) demarcate two general types of content: (1) ideologies, abstractions composed of belief, norms, and values; and (2) forms, concrete cultural manifestations composed of symbols, languages, narratives, and practices. Moreover, dissensus prevails about what to highlight. In her review of the field, Martin (1992: 53) notes that prominent studies5 of organizational culture "disagree about what exactly is being shared," including beliefs, values, patterns of meaning and expectations, understandings, basic assumptions, and communication rules.
Designing Cultural Content
A major difference in views of organizational culture revolves around the question of design--the extent to which culture can be engineered to look like one wants. Of course, most social scientists believe that culture is a naturally occurring phenomenon that every formal organization possesses. For instance, political theorist James Q. Wilson (1989: 91) states, "Every organization has a culture . . . Culture is to an organization what personality is to an individual."
Compared with societies, formal organizations differ in several ways that plausibly affect cultural design. First, formal organizations possess much greater specificity in goals or purposes (Parsons, 1956). Second, formal organizations typically exhibit relatively smaller scales. Third, formal organizations find it relatively easy to control their boundaries with the outside. Fourth, formal organizations usually do not encompass all their members' lives but only some part of them. All except the last factor would seem to make cultural design highly salient, and even partiality of membership might make design easier to implement, since individuals will be less invested in the culture.
Nonetheless, opinions differ about the extent to which the content of the organization's culture can be designed and controlled effectively, especially by top management. At one end of the spectrum, management theorists such as Michael Tushman and Charles O'Reilly (1997: 35) claim that "great managers effectively manage the short-term demands for . . . bolstering today's culture and the periodic need to transform their organization and recreate their unit's culture." At the other end are those who believe that culture is too amorphous and too reactive to be designed and managed in ways that produce precise intended outcomes. Failure in implementing a cultural design might emanate from any number of managerial or other factors,6 but morphologically it takes either of two forms: (1) a discrepancy in content between the design and enacted culture, such that values, beliefs, and norms of the two differ radically even to the point of opposition; or (2) a wide variation in cultural content within the organization, such that the design's values, beliefs, and norms are only one small set among a much broader range represented and widely held in the organization. In either case, the culture winds up differing from the intended design and quite possibly interfering with it.
The issue rarely comes up in analysis of societal culture, which is usually assumed (at least implicitly) to be naturally evolving according to its own logic, and not easily influenced by individual actors, let alone designed, except perhaps in small and dense societies with a strong central government such as Singapore or under extraordinary political circumstances such as the Cultural Revolution under Mao Tsetung. But it seems natural that it should be an issue for formal organizations, given their goal specificity and their smaller scales (Parsons, 1956; Scott, 2003).
In the business world, the executives of many companies and organizations unquestionably believe that they can design and implement a culture to good effect. Many of them try to put in place a culture that embraces the values and norms they believe will make their organization better, as a competitor or as a place to work. Few, however, think that they can do so quickly or without considerable attention, effort, or money.
A GLIMPSE OF TWO DESIGNED CULTURES
A stereotype of designed organizational culture holds that it consists of content primarily of a caring or worker-friendly nature. While this is true of many designed cultures, warmheartedness is hardly ubiquitous. The contents of the cultures that executives and employees design and implement in their companies range wildly, from the secretive and spartan coldness of candy maker Mars, Inc. (Brenner, 1999), to the exclusive gentility of investment bank Goldman Sachs (Endlich, 1999), to the hippie egalitarianism of ice cream manufacturer Ben & Jerry's (Lager, 1994).
A vivid illustration of such differences and their implications for individual behavior can be seen in two companies we are familiar with, Dreyer's Grand Ice Cream and Cypress Semiconductor. Both companies grew at fast paces in the 1980s and 1990s. Dreyer's went from a small shop to a dominant national business with more than 4,500 employees and sales of over 1 billion dollars. Cypress went from a start-up to a multinational corporation with over 4,100 employees and revenues that have passed 1 billion dollars. Both companies were still actively led in the year 2000 (the date of information we use here) by the executives responsible for designing their cultures.7
Dreyer's Grand Ice Cream. Gary Rogers and Rick Cronk initiated the modern incarnation of Dreyer's in 1977 when they bought a local ice cream shop in Berkeley, California.8 As the company grew and expanded geographically to Los Angeles, they realized that different values, norms, and behaviors were developing in the various locations. They sought to unify the culture by articulating their values and beliefs. In 1988, they put together a pamphlet describing their philosophy, which they had been practicing since 1981. They named it the "I Can Make a Difference" philosophy:
At Dreyer's Grand Ice Cream, we believe in the individual . . . In return for committing their energy and enthusiasm to their jobs and identifying personally with their company's goals and challenges, employees today expect that company to value and respect their hearts and souls as well as their minds and bodies . . . People today also want to be trusted . . . People want their company to just assume that they will come to work on time, work hard . . . without needing time clocks, stop watches, forms, and rigid procedures to ensure this behavior . . . Finally, the days when employees were blindly willing to do what they were told and follow their bosses' instructions or their company's policies without question are gone forever. Today, most people have an opinion on almost everything and want a major voice in how their work environment functions. (Chang, Chatman, and Carroll, 2001: 11)
Rogers and Cronk expressed the philosophy more concretely in ten tenets, or "Grooves," which were the company's values, its culture, and the "way we do things around here."9 The essentials of the Grooves are shown in table 1.1.10 Rogers stated that the overall goal of the Grooves was to maintain a "day one attitude in its employees . . . an employee has a perfect attitude on day one, wanting to progress in his or her career, enthusiastic, committed . . . our job is not to change this."
Cypress Semiconductor. T. J. Rodgers and six partners founded Cypress Semiconductor of San Jose, California, in 1982. For many years a niche producer, making high-end innovative chips, the company eventually ramped up scale to become a large producer as well.
Cypress prides itself on its aggressiveness in markets, including the labor market. The company proactively recruits top-notch engineers and gives them great responsibility, provided they can demonstrate worthiness to their colleagues. Employees are attracted to the company in part because they see it as highly meritocratic and devoid of organizational politics. Table 1.2 gives the stated core values that underlie the culture at Cypress.11
Obviously, many of the values, norms, and behaviors endorsed by these two cultural designs differ radically. In a nutshell, Dreyer's embraces many of the "feel-good" principles of the human potential movement, including self-determination and mutual respect as manifested in civility of exchange. People treat each other nicely at Dreyer's, even when they are being frank about less-than-pleasant realities. The culture makes members feel as though they are in a warm and caring family. Employees at Cypress feel they are part of a very different kind of family--one with a tough, uncompromising paternal leader. Cypress people pride themselves on their hard edges, their stark views of reality, and their acceptance of adversity as a great motivator. Feeling good and being nice are not valued nearly so dearly as winning.12 Despite these dramatic differences in content, both companies attribute at least part of their success to their cultures.
As noted earlier, a designed culture can be superfluous and not reflected in the actual culture of an organization. Although we do not possess hard evidence for these two cases, it is our observation about Dreyer's and Cypress in 2000, based on casual and occasionally systematic fieldwork by ourselves and colleagues, that these designs reflect a great deal about the actual cultures in place. This is not to say that every employee had likely internalized, accepted, or even knew all aspects of each cultural design. But we venture with some confidence that a sound empirical study, ethnographic or otherwise, undertaken at either place would have turned up cultural accounts and behaviors mainly consistent with these articulated values, norms, and behaviors. If so, then these examples suggest that questions about designed culture should be not so much about its possibility as about its degree. For any given cultural design, it makes sense to assess the extent to which the enacted culture of the organization agrees with it or not.
As a constraint on action, the cultural content of an organization matters for how it behaves compared with other organizations, and how life is experienced by its individual members. Moreover, in the eyes of many scholars and executives, cultural content also ultimately affects how the organization performs. Why?
Analysts generally agree that the key content issue concerns alignment or congruence between the culture, on one hand, and the organization's competitive strategy, on the other. For instance, Tushman and O'Reilly (1997: 101, 128) proclaim: "To manage organizational culture effectively, managers must be clear in their own minds about the type of culture and the specific norms and values that will help the organization reach its strategic goals . . . The overarching search is either for needed norms that are not aligned or for norms that are currently valued which need to be eliminated or downplayed in the future." Alignment of cultural content with other dimensions of an organization's design may also be an issue. Particularly important in many analysts' frameworks is the organization's formal structure, including the way it groups activities and people, as well as its compensation scheme.
Textbooks commonly prescribe a high level of overall alignment among an organization's strategy, cultural content, and formal structure because of the purported effects on organizational performance. With high or strong alignment, analysts regard an organization as well designed. Lack of alignment leads to deficient or at least substandard performance and could be corrected by realignment of one or several dimensions (Saloner, Shepard, and Podolny, 2001).
For example, a tight-button-downed, control-oriented cultural content, such as that found at Cypress Semiconductor, is typically not thought to be well aligned with a strategy of technical innovation, where slack resources and independence are usually associated with success. As O'Reilly and Pfeffer (2000: 229) surmise:
Cypress says it wants the best, but the company has created an environment that is, in many ways, unlikely to attract some of the talented, creative entrepreneurial people it claims to be seeking. Cypress has built a culture that does attract people who work hard, like to compete, and want to win--all potentially important ingredients for its primary strategy of short-term incremental innovation. But the values that promote this capability may also run counter to those needed for its secondary strategy of breakthrough technical innovation.
Because innovation is typically fostered in loose, carefree cultures, this observation claims an inconsistency or lack of alignment or congruence between Cypress's cultural content and its business strategy of innovation.
With Dreyer's, a milder criticism can be made, namely, that the cultural content fostered by the Grooves bears no direct relationship to the strategy but stands apart from it. That is, the Grooves do not apparently have much to do--positively or negatively--with how the company competes in the ice cream market. As a result, to some the Grooves may represent an indulgence rather than an inconsistency; the cost is in the missed opportunity for aligning the culture rather than any inherent contradiction or tension.
Despite the intuitive appeal of such analysis, we should note that in our view no scientific theory of cultural content addresses questions of alignment in a precise enough way, giving definitive judgments about dimensions of content relative to each other. This means that there exist no strong scientific criteria by which one can assess a priori the objective compatibility or incompatibility of a given culture and a particular strategy. In other words, ad hoc arguments supporting any judgment about content congruence or alignment can be advanced in a compelling manner.
1.5 THE DISTRIBUTIVE APPROACH TO CULTURE
Although historically less popular, the distributive approach to culture represents a conceptual alternative to the content approach. It starts from the view that there is more cultural variation within a society (or other unit of analysis) than advocates of the content approach commonly recognize.13 By the distributive approach, this variation should not be ignored or deemphasized but rather recognized as an inherent part of the phenomenon. As Rodseth (1998: 55) explains, this recognition leads to a very different depiction of culture, one that does not necessarily reflect a coherent or integrated system:
[The distributive approach] construes culture not as integrated system or text and not as a mere aggregation of traits or behaviors but as a semantic population--a population of meanings. These meanings have definite material embodiments. They may be stored in human brains, expressed in speech and other forms of action, or transmitted in writing and other artifacts, but they are always things in the world, rather than mere abstractions. From this perspective, meanings exist not in formal or idealized systems but in spatiotemporal distributions, which may or may not be orderly, coherent or stable.
A distributive analysis of culture first identifies patterns of cultural variation across time and space. It then seeks to account for them. Because the variation occurs within a society or organization, the approach finds differences within the society or group valuable for explanation. For instance, Sperber (1996) emphasizes the distributive approach's ability to deal with cultural change, especially as a function of lower-level (micro) processes. Using the word "representation" (rather than Rodseth's meaning), he claims that in the distributive model of culture
the causal explanation of cultural facts amounts, therefore, to a kind of epidemiology of representations. An epidemiology of representations will attempt to explain cultural macro-phenomena as the cumulative effect of two types of micro-mechanisms: individual mechanisms that bring about the formation and transformation of mental representations that, through alterations of the environment, bring about the transmission of representations . . . in the epidemiological perspective, the explanation of a cultural fact--that is, of the distribution of representation--is to be sought not in some global macro-mechanism, but in the combined effect of countless micro-mechanisms.
Hannerz (1992: 8) takes the approach a step further, arguing that cultural distributions serve to structure and shape future meanings and representations. He claims that "complexity in [cultural] distribution and in externalization has a large part in breeding complexity in the order of meaning."
Perhaps the most influential practitioner of the distributive approach is anthropologist Fredrik Barth. In his study of Inner New Guinea, Barth (1987: 1) analyzes the cosmological knowledge and rituals of the population. When he does, he looks at "the (variety of) ideas that it contains, and how they are expressed; the pattern of their distribution, within and between communities; the processes of (re)production in this tradition of knowledge, and how they may explain its content and pattern of distribution; and thus, the processes of creativity, transmission and change." He goes on to explain that such an approach is not simply about cultural diffusion because "it seeks to identify the developments, departures, and dogmatisms of each of the small local centres within a tradition of knowledge, to discover the patterns of variation and thereby the underlying processes of thought innovation and stimulus at work within it" (pp. 19-20).
Barth advocates a naturalistic research strategy, whereby "analytical operations . . . should model or mirror significant, identifiable processes that can be shown to take place among the phenomena they seek to depict" (Barth, 1987: 8). Some other distributive theorists show more of an inclination to abstraction. Schwartz (1978), for example, notes that between the extremes of total homogeneity and total heterogeneity, there is a structure of commonality. It "consists of all of the intersects among idioverses [implicit cultural constructs of individuals]--among all pairs, all triads, etc.--all subsets of the total set of idioverses of members of a society" (Schwartz, 1978b: 428). And, for any given common social function, Schwartz sees a social structure of commonality.
A difficulty with the distributive approach to culture involves its tractability in actual analysis. Consider, for instance, that Schwartz (1978b) defines culture as the union of all individuals' sets of cultural constructs (what he calls "idioverses"). Goodenough (1981) regards the shared content (the intersection of the idioverses) as the culture but still requires us to consider the "cultural pool" (the union of the idioverses). The two models are "extensively similar" in Schwartz's view because they both involve looking at the diversity and the commonality in content simultaneously. But if there are no exclusionary criteria on content (or the criteria are loose), then just about any and all content held by an individual might be legitimately considered cultural. Moreover, in the case of extreme heterogeneity, where each individual's set shows little or no overlap with others, there seems to be a danger of losing the intuitive sense that something commonly regarded as cultural might be operating.
Even with a precise criterion for relevant content, the distributive approach often has trouble demarcating the population's boundaries. As Schwartz notes, the difficulty arises where individuals' networks of interaction are overlapping, non-discrete and inclusive, implying that similar content will be widely distributed. At the extreme, "we may find that the maximal extension of a given society ramifies widely and indefinitely through the networks of its individuals, until such interlacing and overlapping networks include all of the geographically accessible populations of an area in chains of relatedness" (Schwartz, 1978a: 222). In these cases, the analytical solution involves using some arbitrary threshold to decide what degree of interaction or shared cultural content constitutes the boundaries (Brumann, 1999).
Organizational Cultural Distribution
Perhaps unwittingly, a common type of cultural analysis in organizational theory is appropriately classified as using the distributive approach. Organizational theorists and practitioners express a strong interest in what they call "strong-culture" or "high-commitment" organizations. They typically define a strong-culture organization in purely morphological terms as one where wide consensus exists about the content and where participants believe in the importance of the content (O'Reilly, 1989). Content consensus implies homogeneity of beliefs, norms, and values--a high degree of shared understanding among the members of the organization. Importance of content means that participants take it seriously as the appropriate way to behave, internalize the content as something personal rather than something imposed, and act to enforce the content by sanctioning others when they do not conform (Baron, Burton, and Hannan, 1996).
Cultural strength in organizations is especially important because social control may be exercised informally through the culture rather than formally through a bureaucratic system. Strong culture also provides a common focus, generates intrinsic motivation, and engages peers in monitoring and evaluation. As a result, a widespread tacit understanding of the organization's goals develops that helps to generate consistent actions in novel situations and serves to lower conflict. In general, a strong culture economizes on administrative structure and even management, making it a potentially powerful and very attractive way to run an organization.14 That is, irrespective of content details, homogeneity of content may affect organizational performance as well as its variability over time (Sørenson, 2002a).
Surprisingly to some, cultural content also does not seem to matter much in generating a strong culture. Indeed, O'Reilly and Chatman (1986: 179) provocatively compare designs and processes of cults, corporations, and other organizations. They argue that "religious organizations, self-help organizations, cults and strong culture firms use participation as means for generating commitment, symbolic action to convey a sense of purpose, consistent information to shape interpretations and extensive reward and recognition systems to shape behavior the underlying psychology of social control is fundamentally the same across these types of behavior." Organizational practice shows clearly that, by putting in place certain structural (architectural) design features, it is possible to create and maintain a strong-culture organization. O'Reilly (1989) points to the following design features as key: (1) a rigorous selection of members based on cultural criteria; (2) an intensive period of socialization; (3) a social system with high levels of broad participation entailing numerous incremental commitments; (4) a comprehensive reward and recognition system, including especially intrinsic rewards; and (5) the use of symbolic actions and language to reinforce the content of the culture. This explains why, despite radically different contents, our example companies of Dreyer's and Cypress can both develop and maintain strong cultures.
Demography and Cultural Distribution
Demography concerns the composition of a population and its change over time. The distributive approach to culture sensitizes analysts to demographic fluctuations, as these may radically affect the distribution of content in sparse pockets of a population. Moreover, in comparison to societies and large ethnic groups, the distribution of cultural content in formal organizations heightens the salience of three major demographic issues.
First, the overall distribution of cultural content is driven at least in part, and we suspect heavily, by organizational structure and over time by organizational demography. In modern societies, people belong to many organizations, and most of these memberships last for only a short period compared with their lives or careers. Examine almost any organization over time and you will see a sizable churning of new members replacing old ones, as well as transfer and relocation among the remaining members. Unless cultural content automatically and instantaneously derives from location within the organization, these shifts of personnel must affect the distribution of cultural content.
Second, inertia and change of an organization's culture over time--its dynamics--likely vary by the form and pace of demographic activity. It appears obvious that culture will persist in an organization like the old Jesuit order of clergy, where overall growth is slow, socialization is intense, and turnover is ordinarily low because members intendedly join for life. But what about an organization with fast growth, weak socialization, and high rates of lumpy (uneven spurts of) turnover? It should be possible to map differences in demographic regimes onto cultural stability.
Third, much of the demography of an organization is directly under management control. This means that management potentially affects strongly the distribution of cultural content: it can define the type of content considered ideal, and it controls many of the levers driving the internal demographic system. In particular, management typically decides (or decides to delegate) how many persons will be hired at any time, how selective (from a cultural viewpoint) the hiring process will be, and who is hired. Management also often designs and operates its own socialization process through mechanisms such as orientation and training, as well as placing constraints on the ease with which peer socialization processes operate through mechanisms such as grouping design, job rotation, and incentives. Finally, management sets limits on how culturally tolerant the system will be, and those who deviate culturally beyond the permissible limits will be likely to leave or to experience termination. So, in explaining the distribution of cultural content in formal organizations, it is important to analyze managerial preferences and the policies designed to implement them.
1.6 A LIMITED SYNTHETIC APPROACH TO ORGANIZATIONAL CULTURE
As our review suggests, the content and distributive approaches to culture have both proved insightful but also limited in fundamental ways. It is, however, encouraging that advocates of both approaches recognize the value of the other. For instance, content theorist Schein (1990: 271) notes, "Revolution, restructuring and massive replacement of people may occur, and possibly a new cultural paradigm may occur." Likewise, distributive theorist Goodenough (1981: v) states that "whatever its other characteristics, [culture] has content and organization." And psycholinguist Sperber (1996: 35) proclaims that the "study of cultural representations cannot ignore their contents."
What seems to be called for is a synthesis. Of course, the difficulty of any full synthesis entails combining the attractive features of approaches while dispensing with, or at least minimizing, their deficiencies. We make no pretense of being able to accomplish this formidable task in a general way. But we do believe that the organizational context provides a natural setting whereby a specific limited type of synthesis becomes possible and proves extremely valuable.
Indeed, a particular synthesis motivates and sustains the efforts in this book to understand and model organizational culture. It initially entails representing cultural content in terms of an individual's propensity to embrace the values and norms of a particular organizational culture by a single measure indicating the degree to which an individual organizational member fits management's cultural ideal. We refer to this measure of cultural fitness as the individual's enculturation level; it represents a very specific conception of organizational cultural content, a certain fitness measure of an individual to a given culture. With this enculturation score in hand for each individual, we are then able to examine the distribution of cultural content across the entire organization and the structure of this distribution across different parts of the organization. By looking at the mean level of enculturation, we can also examine the overall degree to which the culture jibes with management's preferences.
Of course, this representation does not, in and of itself, yield a model of organizational culture or its transmission. To create that, the representation needs to be embedded in a set of social processes (e.g., socialization) that reflect some particular way(s) that culture plausibly operates among individuals within an organization. We construct such a model in chapter 4. For present purposes, a key aspect of the representation is that it can be used in conjunction with modeling the organization's demography--allowing individuals to enter the organization, to interact with each other as they experience management enculturation efforts while in the organization, and to leave the organization. Within the organization, individual enculturation scores--as well as their distributions--can change or remain stable, according to the effects of specified mechanisms such as socialization, or social influence, and decay. Demographic considerations thus allow us to tie together distributive-based notions such as cultural heterogeneity with a managerially based content formulation.
Why does this limited or partial synthesis make sense? Generally speaking, the reason has to do with the ways particular characteristics of formal organizations allow us to overcome, or at least bypass, certain troubling limitations of both approaches to culture. Recall that one major problem with the content approach concerns figuring out which elements of content should be singled out for study; different elements of content might give different views, and different persons might share different elements. There is also no precise scientific theory of cultural content. But collapsing content on to an overall abstract preference ordering is justified in this context because some individuals in this context (specifically, those in top management) are endowed with the authority to pursue the limited specific goals of the organization, or to change them. That is, this construct lets the cultural content of the organization potentially be any or all of the elements noted earlier, depending on managerial preferences. The justification for placing the ordering in top management's hands resides in the theoretical definition of a formal organization as a limited-purpose entity with relatively specific goals. Note that this would not be the case with a family, tribe, ethnic group, or society, where goals are more diffuse and rights to change them less clear-cut. Also, it is important to recognize that the preference-ordering construct leads to little loss in content generality, as we show in chapter 3. Recall, too, that the chief limitation of the distributive approach involves its inherent unboundedness of both content and persons. The preference ordering construct again takes care of the content issue. As for the boundedness of persons, the natural membership boundaries of a formal organization provide a clear point of delineation. Although there have been occasional debates, for the most part organizational boundaries are clearly circumscribed and identifiable. Not only do these boundaries coincide with the general limits of managerial control and discretion, but most demographic activity occurring across and within them also falls under the purview of top management. Both factors reinforce the content construct used in the limited synthesis.
With this conceptual synthesis as our background, we are able to construct a model of cultural transmission. We use it here to explore the dynamics of culture in formal organizational settings. We believe that the research program we report in this book represents the most systematic and sustained study ever conducted on the demographic foundations of cultural transmission. In our view, it yields much insight and leads to many new observations about organizational culture. For example:
- Cultural systems in organizations apparently equilibrate under very different demographic regimes. However, the time required to reach equilibrium and the managerial preferredness and homogeneity of culture obtained at equilibrium do vary by demographic activity. These differences underly some fundamentally different organizational types such as bureaucracy and entrepreneurial organizations (see chapter 5 for details).
- Very rapid organizational growth sometimes facilitates rather than impedes cultural stability, when stability is viewed as the quickness with which the system reaches equilibrium or rebounds to it after perturbation. Moreover, some previously observed cultural patterns ascribed to behavioral processes might be accounted for entirely by demographic processes. Cultural intensity, for example, is greater in declining organizations because of the dynamics of attrition, not necessarily because of some psychological reaction to the decline (see chapter 6).
- Common substantive interpretations about heterogeneity in tenure found in the organizational demography literature are potentially spurious. Specifically, an aggregated set of independent individuals with negative duration dependence in the rate of departure generates a statistical--but not behavioral--relationship between heterogeneity in tenure (length-of-service) distributions among members of an organization and the likelihood of a departure. In addition, the strength of the linkage between demography and culture varies by organizational conditions, potentially explaining the inconsistent findings of a large and popular research program in organizational demography (see chapter 7).
- Social influence networks may affect cultural variability in surprising ways. For instance, greater overall inequality of influence in the network lowers cultural variability within the organization. However, when a single individual has very high influence, cultural variability increases after controlling for the overall structure of influence (see chapter 8).
- While some governmental actions can weaken terrorist networks and their cultures, they still retain significant capabilities. Terrorist networks show a high degree of resilience and are not easily destroyed. However, the culture of a typical terrorist network appears to be more vulnerable to the incremental removal of terrorists rather than the elimination of full cells. Also, when facing governmental policy directed toward inhibiting replacement of vacancies in the network, the more viable strategy appears to be delaying replacement of cell vacancies rather than forcing the terrorist organization to skimp on the selection and training of replacements (see chapter 9).
- Post-merger cultural integration increases more quickly in some organizational situations than in others. Organizations with relatively larger acquirers, higher hiring selectivity, stronger management socialization processes, and more severe sanctions for cultural misfits (what we call an alienation effect) facilitate faster integration. Surprisingly, only small effects can be attributed to the layoff policies of the acquirer, the degree of cultural inertia in the acquiree, and the level of turnover and growth. Moreover, in a substantial number of cases, an unexpected pattern occurs: cultural heterogeneity rises in the short term, even when the mean level of enculturation is increasing in the merged organization. A strong alienation effect can overcome this pattern and produce consistent integration patterns (see chapter 10).
- Cultural heterogeneity can play a significant role in organizational mortality processes often linked to organizational aging and known as the liability of newness. Specifically, age dependence in organizational mortality can be accounted for by the effects of small size and cultural heterogeneity. So, observed age dependence in mortality might be a spurious relationship masking an effect of cultural heterogeneity. Organizations that grow without developing homogeneous cultures continue to experience higher risks of failure than those that do. Overall, it appears that cultural heterogeneity mediates size and growth effects on organizational mortality. This general finding suggests that classic arguments about the liability of newness imply a complex relationship between organizational growth, size, and culture--a relationship that is not obvious from purely age-based considerations (see chapter 11).
1.7 READER'S GUIDE
This book might be read in a number of ways, depending on one's background and interest. We have tried to deemphasize the technical aspects of the theory and research, but a certain amount of this material is necessary to ensure precision and to allow possible replication by others.
Chapter 2 discusses the use of formal models, and especially simulation models, to study social phenomena generally but especially organizational culture. It is somewhat didactic and could be skipped by those with technical training; however, it does at points explain the pragmatic philosophy of science we use in the book. Accordingly, this chapter serves as a guide to some distinctive views implicit in our research strategy.
Chapter 3 describes in some detail our way of representing and modeling organizational culture. Obviously, we adopt a synthetic approach that takes into account both cultural content and the distribution of culture. Specifically, we envision a managerial ideal of culture and then model its distribution over time and over different organizational structures as a function of demography. Chapter 3 analyzes the relationship between this representation and the more conventional method of measuring cultural content on multiple dimensions. This chapter could be skimmed or skipped in an initial reading of the book; however, we believe that any serious attempt to deal with the issues we tackle will need, at some point, to address the fundamental issues and findings there.
Chapter 4 presents our basic model of cultural transmission, the foundation for most of the book. The model contains three core components for the experiences of an individual in the organization: (1) a hiring function, (2) a socialization function, and (3) a departure function. We regard this chapter as essential reading for understanding the research program, the book, and many subsequent chapters. It lays open for inspection the bare skeleton of the model that we dress with different costumes in subsequent chapters.
Chapters 5 and 6 describe much of the basic behavior of the model; they should be read together as a module. Chapter 5 organizes findings by type of organization, relying on seven distinctive configurations of demographic features. Chapter 6 continues the investigation of the basic model in the context of these configurations, looking in detail at growth and decline processes in organizations.
The remaining chapters contain applications of the models and extensions or modifications. With an understanding of the basic model in hand, each of these chapters pretty much stands on its own and can be read in any order, except that chapter 9 builds on the network model developed in chapter 8 and might require some backpedaling as a result. Chapter 7 looks at heterogeneity in organizational demography, especially the length-of-service distribution. It scrutinizes a key assumption made in the literature on internal organizational demography and makes some suggestions for future research in this area. Chapter 8 elaborates the socialization component of the model by specifying variations in internal organizational networks of influence. Chapter 9 explores the structure of network-based terrorist organizations and their cultural stability over time. Chapter 10 analyzes cultural integration following mergers from a demographic viewpoint. Chapter 11 models the relationship between cultural heterogeneity and organizational growth and mortality. Finally, chapter 12 concludes the book with some brief remarks.
Return to Book Description
File created: 8/7/2007
Questions and comments to: email@example.com
Princeton University Press