ABSTRACT
The emerging global economy is characterized increasingly by knowledge intensive firms which require that diverse, specialized knowledge workers develop unique knowledge competences, and also collaborate in ways to create new knowledge that enhances the performance of the organization. Information technologies are increasingly playing an integrative role in knowledge intensive firms as a way of achieving mutual learning. However, the information systems field has predominantly been driven by the notion of integration as a rational design process and an end state to be achieved through a static incorporation of knowledge domains. It has failed to consider the interpretive dynamics associated with the integration of differentiated knowledge and expertise. Argues a new role for information technology, one that supports the exploration of differentiated theories of meaning and knowledge and facilitates the conduct of dialogue among highly differentiated experts as a basis for integration.
The relentless change in market expectations and the demands for new products has seen the gradual replacement of capital and labour intensive firms by knowledge intensive firms, and routine work by knowledge work (Starbuck, 1992). Knowledge work involves the creation of new understandings of nature, organizations or markets and their application by a firm in valued technologies, products or processes (Boland and Tenkasi, 1995). It is easy to observe the importance of knowledge work in new product development environments. However, commentators have noted that practically all organizations across the service, industrial and government sectors are becoming more knowledge intensive (Drucker, 1988).
A critical feature of knowledge work is that it requires multidisciplinary expertise and mutual learning in order to achieve a complex synthesis of highly specialized state-of-the-art technologies and knowledge domains (Dougherty, 1992). For example, pharmaceutical product development relies on integration of multiple and differentiated pools of expertise such as molecular biology, physiology, biochemistry, synthetic chemistry, pharmacology and even arcane fields such as molecular kinetics (Henderson, 1994). Also generally, a central element of organizational learning is the ability of organizations to create systemic knowledge by integrating their various interdependent parts (Senge, 1990).
In order to maintain their expertise, each of these domain experts must develop their unique understandings and nowledge base separately. More importantly, in order for the firm to develop new products, the separate domains must interact and learn from each other in ways that will yield new insights into a problem and new ideas for improved products successfully. The knowledge intensive firm requires differentiation and further integration of its multiple and differentiated forms of expertise.
Information technology planning and design today recognizes the need for supporting integration among differentiated knowledge subunits in contem-porary organizations (Rockart and Short, 1991). However, we argue that the current tradition of information systems lacks a strong basis of what it is to integrate differentiated knowledge and expertise and facilitate mutual learning. Integration is treated as a rational process that unproblematically can be achieved by focusing on multiple channels which make available common images, data and words, and better channel management which includes precise procedures for encoding and decoding information. In assuming that a universal consensus of human understanding is possible through rational devices, current information technologies neglect to consider knowledge workers as constituting distinct communities socialized into specialized language games with their unique theories of meaning and their referent interpretations of words, concepts and the world. Words may look similar but the referent meanings associated with apparently similar words can be vastly different. There is a need for elaborating differentiated meaning systems as a basis for dynamic integration. Exploring diversity in knowledge intensive firms is, we propose, a new frontier for the development of information technologies.
A legal advisory firm in England invested in a very expensive rule-based system that enabled the automatic application of logical decision rules to legal problems. However, the system fell into disuse quickly. On investigation the lawyers said they had little use for arguments based on logical inference compared with their need for support in resolving conflicts about meanings. Frequently, lawyers had differing conceptions of what it was to be logical, and the meaning of similar issues and terms varied depending on their legal speciality and experience background. For example, given the task of advising a mining company on closing uneconomic mines the lawyer specializing in corporate law saw it as a balance sheet problem. The labour lawyer had a different definition of an uneconomic mine - it was not so much a balance sheet problem as it was the future unemployment benefits that had to be paid out, and long-term relations with the union mattered. Unfortunately the conflict surfaced only in front of the client since the legal team had assumed that the task and the meaning of the term "uneconomic" was clear, which it was from their respective perspectives.
The rest of the article proceeds in the following manner. First we discuss knowledge integration and mutual learning as a perspective taking process. Second, we review the predominant tradition of information systems that look at integration and mutual learning as a rational, channel management process. Third, we argue the necessity of exploring diversity in theories of meaning as a necessary condition for perspective taking, followed by a presentation of perspective taking as a process of dynamic meaning integration between differentiated organizational communities. Finally, we identify some examples of technologies that are beginning to display a concern with supporting the exploration of knowledge diversity and some research issues that should accompany these emerging systems.
Knowledge integration as perspective taking
In knowledge intensive firms, competitive advantage and product success are a result of collaborative, ongoing learning. Success depends not only on how effectively the diverse individuals are able to organize and develop their unique knowledge competences, but also how they can integrate and utilize their distinctive knowledge both effectively and synergistically (Dougherty, 1992; Nonaka, 1994).
Such collaborative, ongoing, mutual learning by both the individual expert and the team is required because the problem solving and innovation that characterize knowledge intensive firms are unsolveable epistemologically by any one person and require continual insights from a variety of perspectives. Duncan and Weiss (1979, p. 86) summarize this process as one in which: "The overall organizational knowledge base emerges out of the process of exchange, evaluation and integration of knowledge. Like any other organizational process, it is comprised of the interactions of individuals and not their isolated behavior". It requires a process of mutual "perspective taking" where distinctive individual knowledge is exchanged, evaluated and integrated with that of others in the organization.
Much of social behaviour is predicated on assumptions an actor makes about the knowledge, beliefs and motives of others. This is the beginning of the process of perspective taking, and is fundamental to communication. In any communication, the knowing of what others know is a necessary component for co-ordinated action to take place (Bakhtin, 1981; Krauss and Fussell, 1991; Mead, 1934). As Rommetevit observed: "An essential component of communicative competence in a pluralistic social world … is our capacity to adopt the perspectives of different others" (Rommetveit, 1980, p. 126).
Integration as channel management
The process of perspective taking in knowledge intensive firms is essentially the process of integrating the firm's differentiated knowledge structures. The importance of differentiation and integration in determining the adaptability and success of organizations was formalized over two decades ago by Lawrence and Lorsch in their classic Organization and Environment (1967). They observed that the particularities of the environment, task and technologies faced by an organization's subunits were associated with differences in cognitive and emotional orientations among managers in those units. The more differentiated these subunits became, the greater the need for methods of integration to achieve a collaborative outcome.
The mechanisms proposed by Lawrence and Lorsch to enable the "knowing of what others know" were predominantly structural in nature (liaisons, project teams, matrices etc.). They were rational devices for enriching the interdepartmental communication of the organization through better management of the channels of communication, and that generally glossed over the problem of human meaning and interpretation. Galbraith (1977) popularized this view of organization design as information processing capacity, and this rational imagery has been adopted by the information systems community as a basic tenet of design as evident in the introductory textbooks of the field (Alter, 1992).
The rational information processing school and its derivative the conduit model of communication (Reddy, 1979) is based on the assumption that communication is a message sending and message receiving process through a transmission channel with a limited channel capacity. Words, information and data are viewed as conveying objective knowledge and as having fixed meanings. Organizational actors can achieve an universality of understanding and share each other's knowledge since the fixed meanings of words can be communicated objectively from one person to another. The problem to be solved is one of making information available and not one of facilitating subjective, mutual understanding. If there is an issue with understanding it can be remedied by better channel management.
Communication can be improved and integration achieved by reducing noise in the channel, with noise defined as the possibility of error contaminating the message on its route from sender to receiver. Noise can be reduced by increasing channel capacity, and an increase in the amount and complexity of information to be processed warrants a variety of channels or more complex channels. This is manifest in Galbraith's (1977) levels of integrating mechanisms from direct contact, and liaison roles to the matrix form. Similarly, Daft and Lengel's (1984) media richness theory advances different types of information channels (written formal, telephone and face-to-face) as increasingly complex structures to match the complexity of the information to be processed.
Making channels of communication more universally available (such as shared databases), and providing more reliable data storage and retrieval facilities are ways of ensuring better channel management. Another way of reducing noise is by refining the procedures for encoding and decoding messages such as heuristic or algorithmic based decision protocols and meta-level integrative techniques.
Information systems today are an important part of the larger intellectual environment reinforcing the felt need for an integrative image for the corporation. Interestingly, the information systems field in its quest for integration has relied primarily on Lawrence and Lorsch and Galbraith as a kind of organizational paradigm for guiding design of information technologies. Just as the structural elements (teams and matrices) were seen as enabling integrative information processing by acting as devices for better channel management, the information systems community, has seen its technologies as primarily rational integration devices.
This tradition is observed with the early images of the integrated firm-wide system and of the corporate model as a basis for defining shared decision requirements and information needs in system design (Ackoff, 1967) and continues to the recent work of Rockart and Short (1991) who make explicit use of Galbraith as a basis for identifying integrative mechanisms in the networked organization. In achieving integration in the new organizational forms of the twenty-first century, the emphasis is on amplifying channel capacities by increasing communication bandwidths that allow more information to travel at the same time on a common communication line, vastly expanded connectivity with more people and machines linked together, and, enabling communal capabilities in communication through technologies that store and retrieve information from shared databases (Fulk and DeSanctis, 1995).
Enterprise modelling and data architecture are concerned primarily with defining a single, unified data structure for supporting an integrated system (Scheer, 1992). End user computing stresses the establishment of standards and common structures for data and models (Brown and Bostrom, 1988). Model management systems concern themselves with encoding and decoding messages through a variety of meta-level integrative techniques to unify the diversity of knowledge in management decision models (Tam, 1990). In the group support area, the emphasis is on algorithmic-based decision protocols aimed toward gaining group consensus (Tyran et al., 1992) or reducing conflict (Sambamurthy and Poole, 1992).
Most of these designs rely on an assumption that as long as information is made available, shared meaning is achieved unproblematically - that all users employ a standard and shared set of interpretive structures to gain meaning from the data. Integration is seen as a state that is easily attained by emphasizing channels and rules for encoding and decoding messages without attention to how knowledge integration might actually take place. When examples of integration are given, it is an image of operational integration as in the value-chain model (Rockart and Short, 1991) or in the form of heuristics that reconcile independent knowledge bases by focusing on common words and sentences without any human involvement (Tam, 1990).
In sum, the rational information processing model and its manifestations in information system design have looked at knowledge integration with an unproblematic treatment of the notion of a message, a continuing tradition in knowledge intensive firms. It assumes that words, information and data have fixed meanings for all actors and the symbolic or interpretive character of a message is neglected. It symbolizes Varela's "control" approach towards information/knowledge in contrast to his "autonomous" approach. In the autonomous approach, information is always relative to processes of interactions of the domain in which they occur, and to the observer community that describes them. In the control sense information is seen as referential where it is possible to map one set of terms onto a corresponding set. It assumes fixed interactional and observational positions of all actors, and it is instructive, that is it tells one how to act in regard to a particular goal. It fails to include the active role the observer plays in the construction of the meaning of such information/knowledge systems (Cooper and Burrell, 1988).
Need for exploring differentiation of meaning systems as the basis of perspective taking
To appreciate the dynamics of perspective taking and the inherent difficulties of treating integration and mutual learning as purely a channel management process, it is best to cast specialized knowledge workers as members of distinct communities socialized into specialized "language games" with their unique theories of meaning. The notion of "language games" comes from the seminal work of the German philosopher Wittgenstein (1922/1961). Wittgenstein spent the early part of his life trying to define the essence of language, searching for a stable, ideal meaning of words and sentences, and the principles of logic that could be relied on to provide unambiguous and coherent knowledge. His effort was to locate the fixed internal structure of words where the same meaning can be associated with the same words. He later rejected the notion of an ideal language where meaning was uniquely identifiable and stable. Instead he came to see how words and sentences did not have a characteristic internal structure and the meaning of words is specified by the rules of intelligibility embedded in the institutional context in which language is employed. Further, he realized that these linguistic referents are the vehicles of our knowledge and comprehension of the world. We perceive nothing except through the meaning structures of our language in which perception and knowledge is embedded (Astley and Zammuto, 1992; Thachankary, 1992).
Gee (1992) building on Wittgenstein has suggested that meaning of even ordinary taken-for-granted words are dependent on the "semantic mediational theories" (or language games) unique to a community. These semantic mediational theories constitute the assumptive domain that characterizes the meaning and use of a word in a community. They come between the word and the world it indicates. Even commonly accepted words such as water, jerk and tiger can have completely different (and ambiguous) referent domains depending on the community that uses them. For example, the word bachelor is a simple word that to many definitively means unmarried male. Gee (1992) then raises the question: Is a thrice divorced man a bachelor? Is the Pope a bachelor? Is a gay person a bachelor? The answer based on Gee's experiment was no, despite the fact that they all involve cases of unmarried males. This is because the word "bachelor" in the Western culture is connected to a semantic mediational theory that indicates that men marry women (and not other men) and a bachelor is one who stays unmarried beyond an accepted age.
However, as Gee (1992) suggests, unless they are questioned critically and their individual theories of meaning are explored, people from different backgrounds subject to normal discourse agree unproblematically at the level of words and truly believe that they are indicating the same referent domain of meaning giving rise to an illusion of consensus. We can only grasp reality through the various descriptions and classificatory mechanisms afforded by our language games. They provide us with the conceptual apparatus (rules, meanings, conventions) that constitute our perceptual possibilities. As Wittgenstein indicates, what this means is that, even while there may seem to be agreement of understanding on the surface as we encounter other human practices, we can understand these other language games only through an appeal to and reliance on the conventions governing the language game where we originally acquired our language and meaning.
An organizationally relevant example of this uncognizant semantic ambiguity and false illusion of consensus is provided by Dougherty (1992) in her field study of new product development projects. In many of the unsuccessful product development projects, the various departments involved such as research and development, marketing, business planning and manufacturing were consensually together on the need for the product to be market oriented. However, on questioning, each department indicated a different meaning of the term "market oriented". For the research and development group market orientation meant product specifications and technical features - the emphasis was on what the product can do. To the manufacturing people a market oriented product was a durable and reliable - lowering the number of features and specifications would ensure that. For the marketing people market orientation was what each and every customer wanted in the product. The business planning group thought that to be market oriented was having the product in the right market niche.
Further the illusory consensus was not due to a lack of routines (channels) for communication. Dougherty (1992) reports that many of these organizations were replete with traditional integrative mechanisms such as project teams, liaison roles and matrix structures to facilitate better communication and that are often recommended for innovation. A major reason for failure was the inability of the team members to surface and explore their individual theories of meaning and reconcile dissimilarities in their knowledge frames of reference. Failure to achieve perspective taking through depicting and exchanging representations of their unique understandings dramatically reduced their possibilities for successful team learning and knowledge integration.
Procedures for reducing communication noise by encoding and decoding messages as in meta-level integrative techniques and algorithmic-based decision protocols found in group decision support systems can also foster a false illusion of consensus and certainty. An example drawn from field work with a new product development team will illustrate the point . The task for this team was to select a non-human analogue such as a rat or rabbit to conduct tests on a drug that they were developing. The various members representing disciplines such as life sciences, chemistry, toxicology and biopharmaceutics resorted to a popular groupware product and its voting system to arrive at a consensus. Based on the voting procedure, a rat analogue was chosen. Unfortunately, the rat analogue was not suitable for the task, but the team only found that out at the human clinical trials. The poor choice had by then cost the company considerable expense and three years of development time. The rule of consensus through voting embodied in the system hampered the members from exploring their individual theories of meaning that was necessary for making an informed choice. The groupware helped reduce noise in the communication and the members came away from that experience with an illusion of consensus, integration and certainty.
Actors can participate in an illusion of consensus of mutually contradictory theories of meaning unless they are engaged in an experience that requires them to bare open their unique theories of meaning. The ready availability of the actor's own perspective may lead the actor to overestimate the likelihood that the perspective will be shared by others. This false consensus effect, in which subjects assume that others are more similar to themselves than is actually the case (Ross et al., 1977) is a form of bias particularly relevant to the perspective taking process. Steedman and Johnson-Laird (1980) have proposed that "the speaker assumes that the hearer knows everything that the speaker knows about the world and about the conversation, unless there is some evidence to the contrary" (p. 129). This effect should lead to overestimates of the extent to which a speaker's knowledge and meaning is shared by others, and studies support the existence of such a bias (Krauss and Fussell, 1991).
Facilitating the perspective taking process as "dynamic" meaning integration
In order for perspective taking to proceed, there is a need for actors to elaborate on their individual theories of meaning through facilities for self-indication and self-representation. Differentiated theories of meaning and knowledge held by individuals in the organization must be represented in its uniqueness and made available for others to incorporate in a perspective taking process. Enabling each type of expert to make unique representations of their understandings, and assisting actors with different expertise to better recognize and accept the different ways of knowing of others, is the foundation for perspective taking. It can be provided by information systems that include an emphasis on exploring differentiation.
Our behaviour and perception, our logic and thought, comes within the control system of a language (Wittgenstein, 1922/1961). Truth and objectivity have meaning only with reference to some accepted system of verification. No human can be logical, unless there is an agreement among members of a universe of discourse as to the validity of some general conceptions of good reasoning. Thus, each language game or semantic mediational theory affords an interpretive tructure that realizes circumscribed interpretations within a wider set of possible explanatory accounts. Semantic mediational theories offer unique but limited models of the relationship between a given word or type of word and the world that it indicates. They set up central and marginal cases, hierarchies of experiences, and have a tendency of marginalizing certain things, certain experiences or certain explanations (Gee, 1992). The approach to a problem, the level on which a problem happens to be formulated; the stage of abstraction and stage of concreteness one hopes to attain are in a way bound up with the unique theory of meaning of a community.
Exploring differentiated meaning structures is then a requirement for perspective taking and achieving integration. However, integration is not meant as an act of smoothing over differences and arriving at one single, unified understanding. Rather it is a way of sharing unique understandings that can result in expansion of a meaning structure's frame of reference. It implies a process of dynamic, evolutionary integration versus a static incorporation that relies on a one-to-one mapping of meanings. In every day life a community operates on a background consensus of interpretations in a taken-for-granted way. However, a process of surfacing and examining interpretations allows a shaking of the background of consensus and opens the possibility of mutual interpretation that enables the achievement of a new definition of the situation in which all participants can share (Habermas, 1979). Exploring the basis and referents of unique meanings and terms can give rise to new linguistic referent structures that are co-created by intersecting communities of knowing and can represent new understandings of nature, markets or organizations. For example, in the case presented earlier (Dougherty, 1992) the signifier for the concept "market orientation of the product" could have acquired a more expansive meaning if the various parties in the product development process had elaborated on their unique theories of meaning.
In contrast to the objectivists conception of learning as the discovery of objective facts, prominent philosophers of science consider learning and understanding as a change in outlook of concepts, theories and meanings (Feyerabend, 1975). Such a change in outlook in turn affects our general beliefs and expectations and also our experiences and conceptions of reality. Learning and understanding is really a process where frequently known objects and phenomena acquire a new meaning (Phillips, 1977). It is the ability to see "this" as "that", where we see the same object from a new perspective. The Aristotelians saw swinging stones, but Galileo's impetus theory of motion gave a new meaning to swinging stones as pendulums (Kuhn, 1962). Further, such learning is a process of continual interpretation - the horizon of our meaning structures may change, but new questions arise. Meaning does not remain invariant if one is engaged in the process of explanation, self-reflection and exchange of representations. Such a process forces the individual to confront the reasonableness and validity of tacit cause-effect understandings (Weick, 1990).
In essence, developing a comprehensive knowledge base among a community of highly differentiated yet reciprocally dependent individual specialists requires an ongoing process of mutual perspective taking where individual knowledge and theories of meaning are surfaced, reflected on, exchanged, evaluated and integrated with others in the organization. In summary then, the problem of integration of knowledge in knowledge intensive firms is not a problem of combining, sharing or making data commonly available. It is a problem of perspective taking in which the unique understandings representing different theories of meaning are first made visible and accessible to oneself and others. Only after knowledge is differentiated and represented do the actors have something to integrate and engage in mutual learning through communication. An emphasis on integration is acceptable when the diversity of knowledge is sufficient and established, and when the problem of organization control is one of unifying across the diversity to achieve co-ordinated action.
Examples of an emergent tradition of information systems for exploring knowledge diversity
In the age of knowledge intensive firms which we seem to be approaching, a balanced emphasis on both the differentiation of greater varieties of expertise and their integration into collaborative knowledge intensive networks becomes the central challenge for information technologies. Both differentiated knowledge capabilities and collaborative mechanisms are in need of invention, development and enhancement in knowledge intensive firms. There are a few, isolated examples of information systems that seem to truly respect the exploration of diversity of knowledge and ways of knowing in the modern firm. These are a beginning for supporting differentiation in the firm. We will review some of them next, but the central point of this article focuses on the need to face up to this new horizon of system development highlighted by knowledge intensive firms, and inventing ways in which diversity in knowledge can be supported through information technology. None of these examples is "the answer" for the problem of differentiation and integration of knowledge. But they do suggest some beginnings.
One set of alternatives involves blackboard-like systems such as post-mechanistic groupware (Johnson-Lenz and Johnson-Lenz, 1991) or Amsterdam conversation environment (Dykstra and Carasik, 1991) that provide open-ended spaces for users with minimal amounts of structure. The focus is on stimulating interaction rather than producing a product out of the interaction. It is up to the users to structure their environment and interactions instead of relying on a structure provided by a task-oriented application.
Another approach to valuing diversity of knowledge in the modern firm is provided by OVAL (Malone et al., 1992) in which four primitives (objects, views, agents and links) are defined by users to build a knowledge representation and communication tool that can reflect their unique way of knowing. Different knowledge representations, developed by one domain of expertise can be exchanged through a mail system with members from other knowledge domains, thus beginning the perspective taking process.
Cognitive mapping tools are increasingly being used to represent knowledge of an organizational domain. COPE (Eden, 1988) is a general purpose cognitive mapping tool that has primarily been used to construct group understandings. SPIDER (Boland et al., 1994) uses cognitive maps as one element in a hyper link environment for depicting an understanding as a set of contextual layers of representations. In SPIDER, a cognitive map may be the central document in representing an understanding, but each factor in a map is in turn linked to other maps, spreadsheets, graphs or text for revealing underlying assumptions, layer by layer. In contrast to Eden's work, SPIDER is designed so that diverse understandings can each be represented separately and can be exchanged, analysed, and used as a basis for dialogue among knowledge workers. Current experiments using SPIDER with physicians in a health care setting have shown some interesting results. Based on a process of surfacing and exchanging representations the physicians are considerably expanding their understandings of the concept of quality in medical care (Boland and Tenkasi, 1995).
Thach and Woodman (1994) in a detailed survey of emerging forms of information technologies have provided several examples of systems that can facilitate perspective taking. One such future application is a new form of cyberspace that will be a computer maintained virtual world and will enable the enaction of abstract processes such as scientific visualization, where "cognitive entities would take on tangible form to facilitate access and manipulation" (p. 37). Participants' thoughts can be translated into computer forms for exchange, mutual dialogue and creative brainstorming.
Information technology developments, like those mentioned above, break with the theme of rational and static integration that has dominated the information systems field for the last three decades, and begin to explore the issue of differentiation and the exploration of diversity of knowledge within a firm. As other systems are developed to address this new frontier of knowledge diversity and perspective taking (Thach and Woodman, 1994), a new set of research questions should open up. Some research issues that intrigue us include the quality of organizational trust, the process of self-revelation, the quality of interpersonal dialogue and the effectiveness of organizational learning.
Technologies that support knowledge diversity and perspective taking should result in an increase in organizational trust (Zucker, 1986). Representing and exchanging ways of knowing and enabling an improved perspective taking among organization members should lead to increased confidence in the knowledge of the firm and to the experience of successful collaboration with others, both of which would be associated with increased levels of trust.
The surfacing of tacit assumptions during the process of building representations of distinct ways of knowing should also be associated with an increase in self-revelation, or the putting forward of beliefs and ideas that are normally kept to oneself. An increase in self-revelation would also be expected to lead to an increase in organizational trust. Finally, the impact of technologies that support knowledge diversity would have little point if the dialogue in the organization - the perspective taking behaviour - did not improve as a result. Better dialogue would in turn be expected to result in improvements in organizational learning.
These research themes can only be hinted at here. Our community's focus on integration, emphasizing shared and common knowledge, data and models has not stressed their importance, and to be honest, we do not have a good way to think about trust, self-revelation, or dialogue right now. As our attention shifts to questions of differentiation, however, and to questions of better supporting the development and representation of unique, diverse knowledge within the firm, we expect these issues to gain importance and our ability to conceptualize and measure them to improve.
References
Ackoff, R.L. (1967), "Management misinformation systems", Management Science, Vol. 14 No. 4, pp. 147-56.
Alter, S. (1992), Information Systems: A Management Perspective, Addison-Wesley, Reading, MA. Astley, W.G. and Zammuto, R.F. (1992), "Organization science, managers and language games", Organization Science, Vol. 3 No. 4, pp. 443-60.
Bakhtin, M.M., (1981) "Discourse in the novel", in Holquis, M. (Ed.), The Dialogic Imagination, University of Texas Press, Austin, TX.
Boland, R.J. and Tenkasi, R.V. (1995), "Perspective making and perspective taking in communities of knowing", Organization Science, Vol. 4 No. 6, pp. 350-72.
Boland, R.J., Tenkasi, R.V. and Te'eni, D. (1994), "Designing information technology to support distributed cognition", Organization Science, Vol. 5 No. 3, pp. 456-75.
Brown, C.V. and Bostrom, R.P. (1988), "Effective management of end-user computing: a total organization perspective", Journal of MIS, Vol. 6 No. 2, pp. 77-92.
Cooper, G. and Burrell, G. (1988), "Modernism, postmodernism, and organizational analysis: an introduction, Part I", Organization Studies, Vol. 9 No. 1, pp. 91-112.
Daft, R. and Lengel, R. (1984), "Information richness: a new approach to managerial behavior and organization design", in Cummings, L.L. and Staw, B.M. (Eds), Research in Organizational Behavior, Vol. 6, JAI Press, Greenwich, CT, pp. 191-223.
Dougherty, D. (1992), "Interpretive barriers to successful product innovation in large firms", Organization Science, Vol. 3, pp. 179-202.
Drucker, P. (1988), "The coming of the new organization", Harvard Business Review, January-February, pp. 45-53.
Duncan, R. and Weiss, A. (1979), "Organizational learning: implications for organizational design", in Cummings, L.L. and Staw, B.M. (Eds), Research in Organizational Behavior, Vol. 1, JAI Press, Greenwich, CT.
Dykstra, E.A. and Carasik, R.P. (1991), "Structure and support in cooperative environments: the Amsterdam conversation environment", International Journal of Man-machine Studies, Vol. 34, pp. 419-34.
Eden, C. (1988), "Cognitive mapping", European Journal of Operations Research, Vol. 13, pp. 1-13. Feyarabend, P. (1975), Against Method, New Left Books, London.
Fulk, J. and DeSanctis, G. (1995), "Electronic communication and changing organizational forms", Organization Science, Vol. 6 No. 4, pp. 337-49.
Galbraith, J. (1977). Organization Design, Addison-Wesley, Reading, MA. Gee, J.P. (1992), The Social Mind, Bergin and Garvey, New York, NY.
Habermas, J. (1979), Communication and the Evolution of Society, Beacon Press, Boston, MA.
Henderson, R. (1994), "Managing innovation in the information age", Harvard Business Review, January-February, pp. 100-05.
Johnson-Lenz, P. and Johnson-Lenz, T. (1991), "Post-mechanistic groupware primitives: rhythms, boundaries and containers", International Journal of Man-machine Studies, Vol. 34, pp. 395-417.
Kuhn, T.S. (1962), The Structure of Scientific Revolutions, University of Chicago Press, Chicago, IL.
Krauss, R.M. and Fussell, S.R. (1991), "Perspective-taking in communication representation of others' knowledge in reference", Social Cognition, Vol. 9 No. 1, pp. 2-24.
Lawrence, P.R. and Lorsch, J.W. (1967), Organization and Environment, Harvard University Press, Cambridge, MA.
Malone, T.W., Lai, K.Y. and Fry, C. (1992), "Experiments with oval: a radically tailorable tool for cooperative work", Proceedings of ACM Conference on Computer Supported Cooperative Work, Toronto, pp. 289-97.
Mead, G.H. (1934), Mind, Self and Society, University of Chicago Press, Chicago, IL.
Nonaka, I. (1994), "A dynamic theory of organizational knowledge creation", Organization Science, Vol. 5 No. 1, pp. 15-37.
Phillips, D.L. (1977), Wittgenstein and Scientific Knowledge: A Sociological Perspective, Rowman and Littlefield, Totowa, NJ.
Reddy, M.J. (1979), "The conduit metaphor", in Ortony, A. (Ed.), Metaphor and Thought, Cambridge University Press, Cambridge, pp. 284-324.
Rockart, J.F. and Short, J.E. (1991), "The networked organization and the management of interdependence", in Scott Morton, M.E. (Ed.), The Corporation of the 1990s, Oxford University Press, New York, NY, pp. 189-219.
Rommetveit, R. (1980), "On 'meanings' of acts and what is meant by what is said in a pluralistic social world", in Brenner, M. (Ed.), The Structure of Action, Blackwell and Mott, Oxford, pp. 108-49.
Ross, L., Greene, D. and House, P. (1977), "The false consensus phenomenon: an attributional bias in self-perception and social perception processes", Journal of Experimental Social Psychology, Vol. 13, pp. 279-301.
Sambamurthy, V. and Poole, M.S. (1992), "The effects of variations in capabilities of GDSS on management of conflict in groups", Information Systems Research, September, pp. 224-51.
Scheer, A.-W. (1992), Architecture of Integrated Information Systems: Foundations of Enterprise-modeling, Springer-Verlag, Berlin.
Senge, P. (1990), The Fifth Discipline, Doubleday, New York, NY.
Starbuck, W.H. (1992), "Learning by knowledge-intensive firms", Journal of Management Studies, Vol. 29 No. 6, pp. 713-40.
Steedman, M.J. and Johnson-Laird, P.N. (1980), "The productions of sentences, utterances and speech acts: have computers anything to say?", in Butterworth, B. (Ed.), Language Productions: Speech and Talk, Academic Press, London.
Tam, K.Y. (1990), "Automated construction of knowledge-bases from examples", Information Systems Research, June, pp. 144-67.
Thach, L. and Woodman, R. (1994), "Organizational change and information technology: managing on the edge of cyberspace", Organizational Dynamics, Summer, pp. 30-46.
Thachankary, T. (1992), "Organizations as texts: hermeneutics as a model for understanding organizational change", in Pasmore, W.A. and Woodman, R.W. (Eds), Research in Organizational Change and Development, Vol. 6, JAI Press, Greenwich, CT, pp. 197-233.
Tyran, C.K., Dennis, A.R., Vogel, D.R. and Nunamaker, J.F. Jr (1992), "The applications of electronic meeting technology to support strategic management", MIS Quarterly, September, pp. 313-34.
Weick, K.E. (1990), "Cognitive processes in organizations", in Cummings, L.L. and Staw, B.M. (Eds), Information and Cognition in Organizations, JAI Press, Greenwich, CT.
Wittgenstein, L. (1922/1961), Tractatus Logico-philosophicus, Routledge & Kegan Paul, London (first English edition with a translation, 1922).
Zucker, L.G. (1986), "Production of trust: institutional sources of economic structure", in Staw, B.M. and Cummings, L.L. (Eds), Research in Organizational Behavior, Vol. 8, JAI Press, Greenwich, CT.