Accounts of the growth of organizational theory usually start with Taylor and Weber, but, as Scott (1987) remarks, organizations were present in the old civilizations. Even if we did not find the type of public or private formal organization that dominates our societies today, organization was nevertheless a theme for discussion several thousand years ago as well—as I have tried to illustrate by the quotation from Plato’s The Republic in the caption for Chapter 5. In the section the quotation is taken from, Socrates has just started on his elaborate explanation of how a state comes into being, and how it should be organized in order to create just citizens. He is speaking to Glaucon and Adeimantus, two young men puzzled by the doctrine then current that justice is a mere matter of social convention, imposed from without. The two men are eager to discuss this with Socrates, who is of another opinion. Let us indulge ourselves by following Socrates for a few more paragraphs (the person replying here is Adeimantus):

We need not go any further to sense the conviction underlying the text in The Republic: that there is a best way to organize, and that men are more or less preordained for their different occupations. In The Laws, Plato expressed this view even more clearly, giving a very strong statement on the one best way to organize a city-state. The notion that there is a "best way" to organize, and that humans were destined for their eventual positions, held sway (more or less) for almost 2500 years—although it was never a theory in the modern sense of the word. It is more accurate to say that nations as well as public and private enterprises were organized according to tradition and tacit knowledge, and that the structures and methods employed were viewed as "natural" or even ordained by God.

Classical Theory

The explicit theory of the one best way to organize is normally ascribed to the "classical" theorists, notably Frederick Taylor (The Principles of Scientific Management, 1913) and Max Weber (his theory of bureaucracy in Wirtschaft und Gesellschaft, 1922), but it is, as we have seen, much older, even if it then only concerned social organization.

Taylor’s model sprang from factory production and Weber’s from the offices of public administration, but they had a lot in common—notably a reliance on standardization of work, control of quality, fine-grained division of labor, and a strict hierarchy. They both strongly believed that the organizational models they proposed would prevail and eventually supplant all others because they were the most efficient.

Weber’s interest was not in organization per se, but in the role it played in politics and economics in general. His discussion of bureaucracy therefore centered on its legal and political ramifications, as well as its part in the general rationalization of society—a result of the growing hegemony of rational means-ends relations. Weber viewed bureaucracy as the epitome of this development, working with supreme efficiency, and believed it would supersede all other organization forms. In Weber’s eyes, this development was not necessarily in humanity’s interest—on the contrary, he saw in the efficiency of bureaucracy a frightening potential to lock us into an "Iron Cage" of machine-like existence.

With Weber’s own definition of sociology in mind, it is difficult to understand how he could be so sure of the inevitable and total domination of a single organizational structure. In that definition, he bases sociology squarely on an understanding of individual action and interaction, based on individuals’ subjective understanding of their situation and the purpose of their own actions (Fivelsdal 1971). Supra-individual concepts such as structure, function, and system are rejected as causes. One should think that human variation would make room for more than one structural form, and at least that its grim advances could be blocked by a pervasive tendency among disgruntled individuals to choose (for subjective reasons) other solutions.

However, even if we take into account Weber’s inclination to discuss institutional features by representing them in their ideal-type form in order to make the analysis clearer (and thus ending up looking more dogmatic than he really was), there is little doubt that he believed in the technical superiority of bureaucracy as an organizational form, and in its eventual triumph.

Henri Fayol and later Luther Gulick and Lyndall Urwick (Papers on the Science of Administration, 1937) emphasized formal authority and the role of direct supervision (Mintzberg 1979), but the spirit of their work was the same as Taylor’s. You might even say, as Koolhaas (1982) does, that they were not really presenting theories of organization at all, but recipes—indicating the best solution for every type of activity, just as Plato did in the meticulous details of The Laws.

The early theorists’ belief in the existence of final, superior solutions and their inescapable triumph can be viewed as an expression of their times—of the rapid progress of science and technology; the immense success of the mass-producing factory, the general increase in rational attitudes; and a rather naive belief in the simpleness of human affairs and their resemblance to physical systems.

Buckley (1967) has suggested that such theories represented a continuation of the "Social Physics" that (according to Sorokin 1928) arose in the seventeenth century. Its central notion was that man was a physical object, a kind of advanced machine; that behavior and social relations were subject to natural laws of the same kind as the laws of physics; and that man and society could be analyzed and managed accordingly. In politics and history the Marxian visions of inevitable social transformations embodied much of the same spirit, even if the underlying analysis was more sophisticated.

The belief in the rationality and inevitability of things was thus a reflection of the contemporary beliefs in progress and technology, and the notion of the one best solution also appeals to our natural thirst for simplification—a faith in a "one best way" is much more reassuring than the acknowledgment of a bewildering array of optional solutions. As such, this view lingers on today—both in the minds of managers and in the offerings of consultants.

Implied in this view is a notion of technological determinism—if there is a one best way of organizing, there must also be a one best way to utilize any new tool. Such a one-to-one relationship between a tool and its optimal use means that the tool itself will, by necessity, have strong bearings on organizational design.

It is quite obvious that Taylor included tools and machinery in his designs for factory organization, and that the properties of those tools and machines were important determinants for the design of jobs and the relationships between them. The connection may not seem just as plain when we look at Weber and his theories of bureaucracy—there do not seem to be so many tools in use. However, the most important organizational tool in history (at least before the computer) has probably been the art of writing, and Weber’s bureaucracy is explicitly based on written procedures and written information. In other words, if bureaucracy is the one best way to organize administrative work in a literate society, and it presupposes the use of writing, the properties of writing (as a tool) must be regarded as one of the most the most important determinants of bureaucratic organization—maybe even the most important.

Simon’s Bounded Rationality

In the development of organization theory, the belief in the "one best way" and the closed, rational model of organizations (Scott 1987) gradually came under attack after World War II. One of the early attackers was Herbert A. Simon, who developed a new theory of decision making, opposing the reigning concept of unbounded rationality in organizational and economic matters.

In his Administrative Behavior, first published in 1945, he attacked both the economists’ image of "economic man" and the "rational manager" of the earlier management theorists. Although he seemed to accept the notion that there was an objective, theoretical "best way" in a given set of circumstances, he denied the possibility of finding this solution in practice (except perhaps, in some rare instances, by chance).

Simon’s great common-sense realization was that humans operate with limited information and wits in an exceedingly complex world, and that they have no choice but to simplify, to operate with a bounded rationality, to satisfice—not maximize. In the beginning of Chapter V, the second of the two core chapters in his book, he says (1976, p. 79):

In his introduction to the third edition of Administrative Behavior, he puts it even more strongly (1976, pp. xxvi–xxvii):

And a little later (1976, p. xxviii, italics in original):

It follows from this that the realization of an objective "best way" is not a practical possibility, even if it may exist in theory. The objective, practical goal of organizational members is therefore never to find the optimal solution (even if they may think so themselves), but to find one that is good enough for their ends—which usually also means good enough for the organization to survive. It also follows that there must be many such solutions, and that different people and different organizations will more often than not choose different solutions.

In the original edition of Administrative Behavior, Simon does not discuss tools or technology. In light of his attitude toward decisions and organizing, however, it seems reasonable to infer a similar attitude to tools: In theory, there is always a one best way to use a tool in a given set of circumstances, but we can never hope to achieve in real life more than an approximation of this solution. Two of the added chapters in the book’s third edition (1976) concern information technology. There, Simon discusses the effects IT will have on organization, but there are no deterministic prophecies, apart from the contention that computers will bring more automation and allow us to probe the alternatives more deeply when facing decisions.

Simon’s theme of bounded rationality in decision making was further developed in collaboration with James March (March and Simon 1958) to cover the organization in general. In this work the organization’s dependence on and interplay with its environment, which had been peripheral to the discussions in Administrative Behavior, were recognized as critical features. The central theme was that because of the limits in human decision making, organizations will never be fully rational—although managers will strive toward that goal. Organizations will also never be able to adapt completely to their environments, since the bounded rationality of their members will make their understanding of their environments incomplete and their adaptive behavior imperfect. The more rapid the changes in the environment are, the greater will their problems be. To help in this process, a repertory of short-term, adaptive responses are developed in organizations to cope with the more common variations in the environment without the great cost of developing new responses for each change. In highly volatile environments, adaptive change must be institutionalized as far as possible—although this is almost a contradiction in terms, and it will always be difficult to prevent adaptive structures from becoming rigid over time.

Scott (1987) also classifies the theory presented in Administrative Behavior as belonging to the closed, rational system model. This seems a bit unjust, since several passages in the book discuss interactions with the environment (for instance, the discussions in Chapter VI, "The Equilibrium of the Organization") and fully document that Simon does not believe that an organization is an island to itself. However, the theory of decision making that is developed in the book largely treats organizational decisions as something internal to the organization, and this may perhaps merit Scott’s classification. Because the environmental connection is more pronounced in the book coauthored with March (March and Simon 1958), the theory presented there is classified by Scott as belonging to the open, rational system models. These models represent organizations as predominantly rational systems, but they recognize that organizations are continuously dependent on exchanges with their environment and must adapt to it to survive.

Transaction Costs Analysis

Another approach in the open, rational systems category is the transaction cost analysis developed by Williamson (1975, 1985). However, Williamson’s interest in organizational structure centers on questions of organization size and the degree of vertical integration. He argues that the cost of exchanging goods or services between people, departments, or organizations will decide whether or not a function will be incorporated into the organization.

The primeval, "natural" state of business activities can be seen as a situation with individual producers exchanging goods and services through the market. If markets or tasks (or both) grow so complex that the cognitive limits of the producers become overloaded or if the transaction costs increase for other reasons, there will be a pressure to increase the level of organization in order to overcome these difficulties. Applied on the current situation, this implies that existing organizations will try to internalize transactions if they believe they can execute them more efficiently than the market or if they become so complex that market-based solutions become intractable. For instance, an auto manufacturer will develop or buy its own dealer network if it believes it can sell more cars or fetch a bigger profit that way; an aluminum producer will buy into bauxite mines if it believes that this will shield it from dangerous price fluctuations.

If, on the other hand, an organization (generally assumed to be a hierarchy) becomes too inefficient in its internal coordination and market transactions become cheaper, it will tend to either crumble, shed functions, or split up. The transactions will then flow back into the market, as when a PC manufacturer decides to stop making its own motherboards because state-of-the art boards can be bought cheaper and more conveniently on the OEM market.

Technology has a part in transaction theory insofar as it changes transaction costs in the market, inside the organizations, or both. Since information technology has a great potential for changing the conditions for coordination—both within and between organizations, it should be of great interest to the transaction cost perspective.

The Human Relations Movement

In his attack on the classical school of theorists, Simon was joined by the initiators of the human relations school of organizational thinking. The foundations for their arguments were laid down even before World War II, in the report from the Hawthorne studies by Roethlisberger and Dickson (1939), but, according to Scott (1987), it was Elton Mayo who gave the most influential interpretation.

The human relations school brought the individual and the social relations between individuals into focus. People in organizations were no longer seen only—not even mainly—as rational beings working to achieve the goals of the organization. It was discovered that they were just as much driven by feelings, sentiments, and their own particular interests—which could be quite different from what classical theory presupposed. Moreover, the new studies also showed that there was an informal structure in every organization, growing from the unofficial contacts people in the organization had with each other. This informal structure could be just as important as the formal one for predicting the outcome of decision-making processes—sometimes even more important.

According to Scott (1987), there were a number of main themes investigated by the different approaches within the human relations school, and most of them are still actively pursued by researchers. The most basic is the insistence on the importance of individual characteristics and behaviors in understanding organizational behavior. This easily leads to an interest in the effects of different leadership styles, as well as in the effects of race, class, and cultural background. Formalization in work is strongly repudiated on the grounds that it is detrimental to both worker commitment and psychological well-being, and participative management, job enlargement, or, at least, job rotation is prescribed.

In fact, human relations theorists have always been eager to promote changes in organizations to produce what they see as more humane work places, and they claim that the less formal, more participative organization will also be the most productive. It is not unreasonable, therefore, to criticize at least the most ardent proponents of these views for prescribing "one best way" solutions just as much as the classical theorists (Mohr 1971). Mohr specifically mentions Likert and groups him with Fayol, Gulick, and Urwick in this respect. Mintzberg (1979) is especially harsh in his criticism, also referring to Likert. Scott (1987) notes that several decades of research has failed to substantiate most of the claims of the human relations theorists, and that they have also been criticized on ideological grounds for advocating a manipulative attitude toward workers on the part of management.

With their emphasis on humans and their psychological and social properties, the human relations theorists were not especially interested in tools and technology except as a source of repressive formalization. However, even if we might say that they inherited a belief in optimal solutions from the classical theorists, their theories implied that it was human needs and qualities, and not technology, that dictated the optimal organizational forms. In other words, it was in their view possible to design and operate organizations principally on the basis of human characteristics, and thus thwart what others viewed as technological imperatives.

In Scott’s (1987) classification the human relations school belongs to the closed, natural system model. In contrast to the rational model, the natural model refuses the notion that organizations are rational instruments to achieve goals. On the contrary, they are first and foremost collectives of human beings, quite like social organizations such as families, neighborhoods, and societies. Their rational goals are often undermined by more personal or group goals, and the chief real goal of any organization tends to be survival at any cost. The informal structures are seen as the most the important ones, with the formal structures as little more than a stage set. Since the focus of the human relations theorists was clearly on the internal situation in organizations, it is not unreasonable to consider them closed system theorists, although there was also some concern for the effects of worker’s organizational membership on their situation outside the organization.

Woodward

Whatever the specific merits of the human relations movement, there is no doubt that it constituted a major intellectual shift in the thinking about organizations (Scott 1987, Hollway 1991). It largely created organizational studies as sociological and psychological disciplines, and the research on organizations increased rapidly. Among the new research projects were Woodward’s pathbreaking studies of a number of manufacturing companies in the southeast of England in the 1950s (Mintzberg 1979, Clegg 1990), in which she showed how three basic production technologies strongly correlated with a corresponding number of organization structures: Bureaucratization increased as one went from unit or small batch production via large batch or mass production to continuous-process production.

Sociotechnics

In England a group of researchers at the Tavistock Institute of Human Relations developed a distinct framework of their own. In addition to their "action" approach (seeking to induce change as part of their research), they also proposed that "the distinguishing feature of organizations is that they are both social and technical systems" (Scott 1987, p. 108). In their view, the core of the organization represented, so to speak, an interface between a technical system and a human (social) system. This implied that, in order to achieve maximum performance in an organization, it did not suffice to optimize only the technical or the social system, nor to search for the best match between existing technological and organizational elements. The goal should be a joint optimization of the two—creating a synergy that yielded more than could be achieved simply by adding the two together. Their preferred organizational solutions emphasized co-determination, internalized regulation, and workgroup autonomy.

They also discovered that changes at the workgroup level did not survive for long without compatible changes in the overlying structures—a discovery that was also made in a series of experiments with autonomous workgroups in Norwegian industry in the 1960s, inspired by the Tavistock group and directed by the newly founded Work Research Institute in Oslo (Thorsrud and Emery 1970). During their projects they also learned that the environment impinged on intraorganizational activities to a much larger degree than they had anticipated. Scott quotes Trist thus (Scott 1987, p. 108):

Scott (1987) classifies sociotechnical theory as belonging to the open, natural system models. An interesting aspect of the theory is the belief that there may be many optimal solutions to a specific problem—the "joint optimization" of a particular technical and human system can be implemented in different ways that can be equally efficient. From this it follows that sociotechnical theory completely dispels the notion of technological determinism, a conclusion supported by Eason (1988). It also follows that there is an intimate interdependence between technology, the social system, and individual roles.

In my view, sociotechnics is here taking a position that is particularly relevant for information technology, even if sociotechnics was established as a theoretical framework before computers started to make themselves felt to any significant degree. When working with information technology in organizations, it is of utmost importance to be aware of the intimate interdependence between the computer-based systems, the individuals using them, the manual routines, and the organizational structure. Any serious attempt to optimize the use of information technology must acknowledge this reciprocity.

It is therefore quite remarkable that sociotechnical theory has remained so much out of fashion for the last decade (or even two), just the period when the use of computers has really exploded. One reason may be the general lack of interest in information technology that has plagued the social sciences overall (which probably explains the lack of a sociotechnical revival in Norway and the rest of Scandinavia); another is that those who were interested within the sociotechnical tradition tended to be drawn toward research on the cognitive aspects of computer use, especially the (literal) user interfaces of computer systems, neglecting the overlying (and much more important) question of the broader interaction of humans and computer systems in structural terms. A notable exception is the book Information Technology and Organizational Change by Ken Eason of the HUSAT center in England, which discusses the design and implementation of computer-based systems in organizations, broadly based on the principles of sociotechnical analysis and design, and with due concern for organizational matters.

Marxist Theory

Another theoretical approach with strong normative foundations is Marxist theory, which has traditionally approached Western organizations (both business and public) as instruments for control, domination and appropriation of surplus value. This is of course a consequence of the broader political analysis that forms the basis of Marxism, which lies behind a Marxist approach in any field. The normative aspect of Marxist theory usually manifests itself in a critical approach to existing practices, and there is much less emphasis on prescriptions for alternative arrangements, except on a societal level. Most suggestions have been centered around models for collective ownership.

Because Marxist theory is explicitly political in its outlook, it will generally hold that investigators cannot remain neutral, but will have to choose whether to legitimate and support existing practices or to transform them in ways that serve the interests of the people employed in them. Most other theories are seen as—consciously or not—legitimizing exploitation and serving to maintain existing organizational forms. Therefore the dialogue between Marxists and people working within other approaches has, at times, been quite strained.

Focus in Marxist analyses has, quite naturally, been on organizations as power systems, and the objective has generally been to reveal the nature of the suppression they represent, especially as instruments for capitalists and capitalist society. Braverman (1974) is an example of such an approach.

Marxist approaches insist that the political and economic context will strongly affect organizational structure and functioning, and advocate the necessity of historical analyses. They therefore view organizations as open systems. In addition, they maintain that the internal structure and functioning of organizations is largely a consequence of the specific interests of persons or groups of persons, both within the organization and outside it—that is, they are definitely constructed—and not simply a consequence of more or less neutral contingencies. Scott (1987) therefore classifies these approaches as belonging to the open, natural systems category.

Marxist theory’s attitude to technology has been varied. Often, it seems like it espouses technological determinism. Braverman discusses this tendency, but argues vigorously that it is not grounded in Marx’s writings (1974, p. 20):

Braverman’s approach to technology thus resembles the action and constructivist approaches described later. However, Braverman’s conception is more narrow, as he seems to think that the "capitalist property relations" dictate both the form of social relations and the labor process within the enterprise. Instead of technological determinism, we therefore get a kind of social determinism, where one subset of relations—property relations—determines the rest and thereby indirectly determines the way technology is employed.

Contingency Theory

Woodward’s technologically based modifications of the "one best way" approach were soon supplemented by other studies, which refuted even the modified determinism she proposed (Clegg 1990, Mintzberg 1979). The Aston studies, for instance, pointed to size as the determining factor; other studies pointed to age, product diversification, the degree of stability in the environment, cultural factors, and so on. This new diversity was also corroborated on a more general basis by open systems theory, which showed that both equifinality (same result from different starting conditions) and multifinality (different result from same starting conditions) are facts of the systems world (Bertalanffy 1973). The emerging central theme for the theory building on these findings was that design decisions depend on environmental conditions, and that those organizations whose design and internal arrangements best match the challenges they meet in the environment will be the most successful (Scott 1987). The term contingency theory, which has since been used to designate this body of theory, was coined by Lawrence and Lorsch in their (1967) book Organization and Environment: Managing Differentiation and Integration.

Since then, studies of contingencies have made up the main body of organization research (see Mintzberg 1979 and Clegg 1990 for thorough discussions), and variations of systems approaches have dominated. Scott (1987) classifies the "classical" contingency theory of Lawrence and Lorsch as belonging to open, rational system theory—open because of its emphasis on the decisive dependencies between organizational structure and the environment; rational because it still views organizations as predominantly rational instruments for achieving specific goals.

Scott also identifies a direction he calls the strategic contingency approach. It shares the main tenets of contingency theory—that organizations are open systems that differentiate structurally in order to respond to challenges and opportunities in their environment. In addition, it acknowledges that individual organization members and organization departments vary in their interests, motives, and power, which means that organizations must be viewed as coalitions, and not monolithic actors. A primary source of power is the extent to which an individual or group is vital in dealing with the uncertainty posed by the environment. According to Scott, a seminal work in this tradition was Crozier‘s The Bureaucratic Phenomenon (1964).

For contingency theory, technology is just one of many contingencies (although one of the most important ones) that shape organizational structure and behavior. Although different production technologies are viewed as conducive to different organizational structures, the relationship is not seen as deterministic—contingency theory recognizes that there are generally too many variables that have bearing on organizational structure, both external and internal to the organization, for a single one of them to dominate completely. The diversity in real-life combinations will also make it unlikely that the same technology always produce the same organizational impetus.

Technology is, however, as far as I have been able to ascertain, generally perceived as production technology. It is therefore looked upon as something external to the organizational efforts themselves, a given factor that organizations must adapt to—not a tool that may change the possibilities or modes of organizational adaptation in itself, as sociotechnical theory implies and information technology promises.

Organization as Patterns of Action

The fact that organizational members and departments vary in their interest, motives, and power has been developed further to what Scott (1987) calls theories of negotiated order, covering both the symbolic interactionism of Goffman and the action approach of Silverman. Their central theme is that people do not behave; they act—only inanimate matter "behaves," and only its behavior can be understood through an observation of the behavior itself. People always interpret the situations they are in and their own actions, and they attribute meanings to them. To understand the logic of human actions, we can therefore not rely on observation alone; we must also understand their subjective meanings for the people involved. As Silverman says (1970, pp. 128–29):

Theories based on behaviorist views are therefore rejected—organizations cannot be seen as behavioral systems reacting to external and internal stimuli in order to adapt and secure survival. Processes cannot be treated as objective facts, something external to the individual actors (Silverman 1970, p. 130):

Meanings are given to us through the social environment we grow up in; they grow from the history and present structure of society and especially the part of it that we belong to. They are sustained through our everyday actions, through our compliance with role-expectations—when we believe we are acting "naturally," in the only possible way, we are in fact only reinforcing prevailing meanings. From this follows that meanings are also socially changeable, and changes can occur both through disruptive actions and through gradual developments, since we never comply completely with expectations, and new expectations may appear.

Organizations are therefore (just as society itself) social constructs, which exist because their members and their outside contacts continue to act according to sets of role-expectations—sets common to society and to peer and other groups, as well as sets peculiar to the organization. Therefore, Silverman argues (1970, p. 153):

Scott (1987) classifies this approach as belonging to the open, natural system models. Silverman (1970) himself viewed action and systems explanations as conflicting, since he perceived their views of the nature and consequences of social order as very different. Indeed, he did not consider the action approach as a theory at all, but more as a method, a "frame of reference" for the analysis of organizations.

Weick (1979) describes organizations as accomplished by processes, and those processes in turn contain "individual behaviors that are interlocked among two or more people" (p. 89). When the behavior of one person is thus contingent on the behavior of others, these contingencies are labeled interacts. When the interaction is reciprocal, so that both parties’ actions are contingent on the other, it is called a double interact. Regular patterns of interlocked behavior produce the organizational structure.

By and large, the writers in this tradition have, from the beginning, considered technology a part of the environment, something that organizational members relate to with reference to the meanings they attribute to it and to their own situation. At least for Silverman, this is quite evident from the preceding passage.

Some of the writers that inspired Silverman (notably Peter Berger and Thomas Luckmann) have also been foundational for a similar approach, called institutional theory (Scott 1987). They maintain, like Silverman, that social reality is a human construction, continuously recreated through social interaction. In this interaction there will be recurring patterns, and certain actions will acquire a commonly understood meaning. This process is defined as institutionalization (Berger and Luckmann 1967, p. 54, italics in original):

This approach has also been adapted to organizations, and Scott points to Meyer and Rowan (1977) as a prominent example. Although they recognize that many organizations are structured mainly according to the demands made by their technology and work activities, Meyer and Rowan argue that a large number of organizations "reflect the myths of their institutional environments instead" (p. 341). Norms of formal rationality have not only become sufficiently pervasive in modern societies to be institutionalized; they have become so entrenched that they have acquired the status of myths—beliefs so widely held that they are beyond objective testing, beliefs that are true precisely because they are believed. These myths of rationality will then not only present a compelling pattern for organization, but also provide the organization with legitimacy in the wider societal context.

More recently, theorists belonging to the tradition of social constructivism (such as Trevor Pinch, Wiebe Bijker, and John Law) have also turned to the subject of technology. They argue that technology is not a pure phenomenon extraneous to society, developing according to its own internal laws of scientific logic or technological necessity. On the contrary, even technology is socially constructed, under the influence of a wide range of heterogeneous factors—concrete technologies always represent compromises between human actors with specific interests. And, accordingly, these compromises (and hence the successful technologies) might always have been different. As Bijker and Law say (Bijker and Law 1992, p. 3, italics in original),

There are a number of other "traditional" approaches to organization theory as well—such as the population ecologists, who study selection processes among populations of organizations with concepts and methods adapted from biology (Hannan and Freeman 1977). However, I will wind up this section instead with some organizational aspects of what has been termed postmodern theory.

Whereas the 60s and 70s were dominated by the development of the various versions of contingency theory, the last 15 years have witnessed a rapid proliferation of a bewildering array of theoretical perspectives, moving far beyond the modest pluralism of the original contingency theories. As Reed says (Reed 1992, p.1),

Not only have factors such as size, production technology, environment, and age come into focus; likewise have power arrangements, politics, culture, and history. A number of organization researchers have left what they see as drab, constrictive old paradigms for the exciting turbulence of postmodernism, chiefly inspired by developments in philosophy and literary criticism, with the main illumination coming from the works of Jean Baudrillard, Jacques Derrida and Jean-François Lyotard (Hassard 1993). Parker (1992), Hassard (1993), and Thompson (1993) see two main schools of thought in postmodern theory, which also apply to postmodern organization theory: One is epistemological, presenting postmodernism as a theory of knowledge, a way of seeing the world; the other is ontological, presenting it as a description of a historical epoch. The difference between the two is quite substantial; indeed, as Parker (1992) and Thompson (1993) note, it is difficult to understand how the two approaches can be compatible.

Postmodernism as Epistemology

The epistemological approach is by far the one that constitutes the most decisive break with "conventional" science—according to Hassard, its defining feature is its insistence that there cannot be unequivocal relations between forms of representation (symbols, such as words and images) and an objective, external world. This means that we cannot ever really discuss "real," external phenomena, such as nature or even social structure—all such discussion becomes just a ("serious") play with words, whose meanings are impossible to ascertain. Theory formation the way we know it is therefore also impossible. Consider, for instance, Gergen’s position on organization theory (Gergen 1992, p. 210):

The method for revealing this lack of correspondence is Derrida’s deconstruction—a tool for exposing the inherent contradictions in any text. The term "text" is here given a wide interpretation, including both written and verbal communication, as well as the social context in which the communication and the deconstruction itself takes place (Hassard 1993). Derrida argues against the notion that language is just a medium for the communication of thoughts ("logocentrism"). If words cannot truly represent an objective, external world, it is equally impossible to have them represent "objective" aspects of the soul, mind, or reason. Deconstruction, meaning simply the opposite of construction, is a method for revealing the way text is constructed as well as the contradictions and dynamics inherent in it. Central here is the concept of différance, a word coined by Derrida to fuse the two meanings of the French word "différer"—to differ and defer. Gergen exemplifies (1992, p. 219, italics in original):

Différance is one of the five concepts emphasized by Hassard (1993) as the key elements of the postmodern approach to knowledge. The other four are representation (the genuine order of things cannot be discovered), reflexivity (one must be critical of one’s own assumptions), writing (language does not represent concepts with independent existence in the object world), and decentering the subject (the deconstruction of subjective awareness as an artifact of language).

Postmodernism as Ontology

The epistemological approach to postmodernism is not easy to understand, and it may be even harder to accept—it has altogether left what more "traditional" researchers are ready to accept as serious science. The ontological approach, however, is more accessible. The central notion here is that society is moving into a new era, which differ from the previous "modern" age in significant ways (Hassard 1993, p. 3):

The goal for this strain of postmodern research will therefore be to identify the features in the external world that supports this hypothesis (Hassard 1993). A part of this tradition is also the use of other "post-"combinations, such as post-Fordism, postcapitalism or postindustrialism, which all signify a break with the "modern." Within the field of organization research, Hassard points to Clegg’s Modern Organizations (1990) as a "work which reflects the ‘epoch’ orientation." Clegg is also an example of a postmodernist with a definite link to "conventional" theory and methods, and a comparison between Clegg and Gergen supports Parker (1992) and Thompson’s (1993) contention that the epistemological and ontological approaches within postmodernism cannot be accommodated under one roof. Clegg would probably disagree, but he acknowledges the distinction between the epistemological and ontological approach.

Clegg’s purpose with Modern Organizations is twofold: to present empirical evidence for the emergence and growth of postmodern organizational forms and, through the concomitant discussions, to tap into "important debates about the nature of modernity and postmodernity" (1990, p. 1). To highlight the main focus of his book, and also what he believes to be the defining quality of postmodernity, he continues (1990 p.1):

The core, then, of modernity is differentiation—in organizations, especially the increasingly fine-grained and rigid division of labor. The core of postmodernity is de-differentiation—the gradual integration of jobs, the blurring of areas of responsibility, the increasing overlap of functions, the increasing flexibility, and the team attitudes.

To Clegg, Weber is the fountainhead of modernity, the great creator of rational order who casts his shadow over the whole field of organizational analysis. "Of course, it is in the nature of a colossus to cast a large shadow," says Clegg, who goes on, "The shadow will be given a name: modernism" (1990, p. 3). However, this modernity has no singular appearance, "organizations have been represented in various modernist terms" (1990, p. 3), such as the ones that have been presented on the preceding pages of this dissertation. Clegg views these all as competing hypotheses, "capable of adjudication," but argues that they are insufficient to explain the postmodern phenomena (Clegg 1990, p. 5):

Instead of what he sees as the overarching frameworks of modern organizational theories, Clegg suggests the use of what he calls "modes of rationality," where the focus is "on what agents actually do in accomplishing the constitutive work involved in organization" (1990, p. 13).

An agent can be a person, an organization, or a subpart of an organization. Agents act under a subjective rationality: They attempt to accomplish projects "which make sense in terms of the calculation which agents have available to them" (1990, p. 7), and they are "knowledgeable actors with a healthy regard for self-conceptions of their own and other’s interests" (1990, p.13). However, subjective rationalities can differ widely, since any agent will be heavily influenced by the cultural and institutional values of their national frameworks, which of course vary considerably. Therefore, organizational forms and practices cannot be universal; they will vary greatly from setting to setting, both across nations and within them, as agents develop their strategies in accordance with their perceived interests and the constraints of their local environments. Clegg refers here to Granovetter’s (1985) concept of the social "embeddedness" of organizations.

These frameworks do not constrain agents absolutely, and they are not immutable, because they themselves are socially constructed phenomena. However, they change only slowly, and they represent the most important determinant both for how organizations are structured and for how the work. To substantiate this, and to counter what he perceives as a long-standing tendency to place excessive emphasis on empirical evidence from the Anglo-American sphere, Clegg draws his examples from other parts of the world, in Europe and (especially) Asia.

Clegg, then, views organizations as assemblages of agents on different levels, all acting under their respective local modes of rationality. They are significantly (although not totally) constrained by the cultural and institutional frameworks within which they are embedded, but they use their wits and whatever tools and methods they can muster to achieve their purposes (Clegg 1990, p. 153):

Around the world, Clegg argues, no discernible convergence toward a dominant organizational form can be found, as contingency theory, in his opinion would expect. Rather, we find a considerable range of functional alternatives.

After this tour of the craggy landscape of organizational theory, a newcomer to the field must feel inclined to ask if there is any help to get here at all—and a more veteran traveler may indeed question the possibility of establishing a sound platform for further inquiries. Organizations seem to be a little bit of everything. Gareth Morgan (1986) has struck a chord by presenting eight provocative metaphors—eight different outlooks on human organization, portraying organizations successively as machines, organisms, brains, cultures, political systems, psychic prisons, flux and transformation, and instruments of domination. By bringing forth this bewildering array of facts, theories, and conjectures, Morgan aptly illustrates the diversity in contemporary organization theory. His book and, especially, some of the essays in Reed and Hughes (1992) force the question whether there are any commonly acknowledged, defining features of organization left at all. Are they all lost in the diversity (or even anomie, as Reed [1992] suggests) of the theorizing of recent years?

To me, the answer is a clear no. That is, there may not be any commonly acknowledged defining features anymore, but this is more a consequence of the theoretical debates at this point in time than of the nature of organizations and the people who make them up. If you discard the idea that it is possible to accumulate at least a base of common theory about organizations or if you draw the conclusion that any perspective is just as valid as any other, then you do not pursue different perspectives; you pursue a rather sterile perspectivism. You may amuse yourself and a number of others, notably people without any practical responsibility for organizations, but you will end up without influence on how organizations are actually built and run.

I do not dispute the value of different perspectives—on the contrary, they are mandatory for illuminating the rich texture of actual organizations, which is indeed Morgan’s (1986) main purpose with the book. I therefore think some of the criticism, like Thompson’s (1993), is a bit unjust. Morgan’s stated purpose (in the introduction) is not primarily to build theory, but to explore and develop "the art of reading and understanding organizations." Organizations—or at least managers—use metaphors themselves, often quite consciously, to focus on certain interpretations. The slogan "marketing is war," which was quite popular some years ago, is an example of this.

However, the acceptance of different perspectives as useful tools for understanding organizations must not degenerate into the theoretical anomie that I feel is dangerously close to Gergen’s (1992) point of view. To me, the most basic proof for the possibility of building theory is the simple fact that experience makes you better equipped both to understand organizations and to operate inside them. We know that most people who work their way through a number of organizations develop a "feeling" for organization that makes it possible for them to understand the nature and peculiarities of a new exemplar both faster and better than people without such experience. If, in addition, they have special talent (sensibility) for understanding human affairs, they can become very adept, even in the absence of formal training or knowledge about organization theory. If organizations had nothing in common, if there were no basic rules that applied to how people behave in them, it would be simply impossible to learn from experience in this way—you would be back to square one every time you encountered a new specimen.

If practitioners, then, can develop such theories-in-use (Argyris 1980), theories they work by even if they do not make them explicit (or are even aware of them), there must be regularities and common traits in organizations that can be made the subject for explicit theories as well. And I think that existing theory, research findings, practical experience, and common human sense taken together makes it possible to build such theories.

However, accepting the viability of theory and the view that all perspectives are not equally important, it is just as vital not to fall in the other ditch—to select one narrow perspective and explain everything within its bounds, as classical theory, Marxist theory and transaction cost theory do. Theories built within one perspective are not necessarily wrong, but they are doomed to cover only a part of reality. With that in mind, however, we can accept them as valid for their field and use them when they are relevant for our particular problems. Perhaps this is as far as we will ever get—organizations are exceedingly complex, and our wits are indeed limited. Maybe, however, that exactly this proposition of Simon’s is an example that we may find valid general propositions after all! I cannot enter into that discussion here, however, because it would easily displace my real subject in this dissertation.