The Problem of Knowledge Sharing

Knowledge sharing is found in most knowledge-management activities, from capturing and localizing knowledge, transferring knowledge and technology, to knowledge creation. The intent in capturing and localizing knowledge in the organization is to enhance the potential for linking and integrating knowledge dispersed throughout the organization, and to improve the efficiency of doing so (Davenport and Prusak, 1996; Choo, 1998). Knowledge and best-practice transfer within and between organizations is not a one-way activity; it is a sharing process, involving trial and error, feedback, and mutual adjustment (Szulanski, 1996; Powell, 1998; Kaeser, 2001). While the early literature on knowledge and technology transfer between and within organizations hinged on a simple model of information communication, new contributions in this area understand knowledge and technology transfer as a model of knowledge sharing and local knowledge (re)-creation (for a review, see von Krogh and Koehne, 1998). In theories of knowledge creation and innovation, the sharing of tacit knowledge among participants in an innovation process precedes the articulation of new concepts, the appraisal and justification of these concepts, and product prototyping (Leonard and Sensiper, 1998; von Krogh, Ichijo, and Nonaka, 2000; Nonaka, von Krogh, and Voelpel, 2006).

Often, apprentice training is used as an example of sharing, especially of tacit knowledge. Sharing tacit knowledge entails self-observation, reflection, and immersion in the routines of the master, as much as it does observation and imitative learning by the apprentice (Polanyi and Prosch, 1975). It also involves transformation of masters’ cognition, as the training of apprentices confronts masters with new experiences that are tightly connected to, if not caused by, the performance of their own routines. Additionally, apprentices might bring a ‘fresh’ view to the tasks, which may help masters to perform them differently (Perkins, 1981). Masters’ reflections on their own routines are indistinguishable from their observation of apprentices’ attempts to re-create those routines at comparable standards of performance. Of course, it is this indistinctness of observation and reflection that could be said to be the very essence of being a ‘master.’ In this system of master and apprentice, masters’ routines will change as a result of observation and reflection but to what extent, and in which direction, is dependent on individual masters’ interest and willingness to adopt a shared identity with the apprentice (see Cohen, 1985). Some authors claim that good apprenticeship training can only be achieved when masters’ routines include the observation and reflection that comes with training (Dreyfus and Dreyfus, 1986).

This example of tacit knowledge sharing reveals an important insight. If knowledge sharing is knowledge re-creation, as outlined, it could be construed as a sequenced collective action problem rather than the communication of information, representation of tasks, and procedural knowledge (see Gaechter, von Krogh, and Haefliger, 2010). Sharing, in its simplest form, needs a knowledge giver, e.g. master, and knowledge receiver or ‘re-creator,’ e.g. apprentice, and a sequence of activities. The giver must be interested in transmitting knowledge, and the receiver must be interested in applying it; in both cases, cognition and action are altered. The receiver becomes a giver, who, by combining the new knowledge with gained experiences, rewards the giver with knowledge; and the giver in turn becomes the receiver.3 As Merleau-Ponty (1995) remarked, individual learning often aims at producing a ‘maximum grip’ in a situation. Working in isolation, individuals direct their perception and positioning in relation to a task so that they can optimally control task execution and performance. This implies ignoring the presence of others during task performance. However, in a master–apprentice relationship, the ‘other’ becomes an integral part of task performance, cognition, and action simply because there is a master or apprentice observing, commenting, helping, or maybe even ignoring the person performing the tasks.4

Knowledge and interest are inter-dependent,5 so interests should have an effect on knowledge sharing. For the purpose of this discussion, we use a broad understanding of interests based on Schutz (1970); a specific motivational focus determining an individual’s definition of the situation as well as his particular rules of conduct, objective, or intent.6

As the literature on cognitive psychology shows, individual interest influences individuals’ learning in a specific situation (Bruner, 1979), for example, how much of a particular subject matter they are willing to study and absorb. Stories people tell of their expertise and experience also reveal that their interests are varied and manifold and tightly connected to their own areas of tacit and explicit knowledge, history of action, and task performance (von Krogh, Ichijo, and Nonaka, 2000). Organizations create opportunities for people to realize their individual interests through learning; organizational task systems allow individuals to invest in learning and build expertise that is considered more or less costly, valuable, and even legitimate (see also discussion in Gourlay, 2006).

Were it not for the variation in individual interest throughout an organization, knowledge sharing would be a straightforward activity, without much shirking, guile, or resistance. In an organization populated only by altruists, who are interested beyond all else in helping others to learn, knowledge sharing would not represent a major issue to researchers concerned with organizational efficiency and innovation (Collard, 1978). However, empirical studies in many fields, for example Bruno Latour’s work on how science proceeds as a collective enterprise, show that there are diverse and distributed interests behind knowledge production, barriers to knowledge sharing and people who find ways to teach less than they learn from others (e.g. Latour, 1987, 1993). Paul Carlile (2002) has shown that people’s long-term investments in areas of expertise, for example in engineering disciplines, make them reluctant to share knowledge with representatives from other areas, and they tend to be very conscious of ‘boundaries’ and diverse interests, which separate their work practices from those of other disciplines. Darrah’s (1990) ethnography of a high-tech manufacturing firm is also interesting in this respect. Fearing the loss of authority, manufacturing engineers who had full information about product design and specifications were reluctant to share this with front-line workers. As a (negative) result, the firm’s attempts to lower quality costs by training led to more frustration than skill enhancement among the workers. Recent laboratory experiments have also confirmed that sharing explicit knowledge might easily break down with even the slightest change in the (economic) incentives affecting people’s interests. We developed a game-theoretic model showing that explicit knowledge sharing is an activity with multiple equilibria (Gaechter, von Krogh, and Haefliger, 2010). Since there are many outcomes of knowledge sharing, we conducted an experiment involving 228 undergraduates. While we found more knowledge sharing than initially predicted, we also found that knowledge sharing depends on social preferences, for example, fairness. When people feel others benefit unfairly from what they share, they may be more reluctant to cooperate; this effect may be even more pronounced for tacit than for explicit knowledge. These studies, which pay explicit attention to people’s diverse and distributed interests, conclude that knowledge sharing is a highly fragile and uncertain activity.7

A theory that explains why and under what conditions the collective act of knowledge sharing occurs, must take into account a variety of interests and personality traits, including self-interest and individuals who seek to maximize their own utility (see Ostrom, 2000).8 Given the sequencing of activities and self-interest, free riding on others’ knowledge is also of some concern (e.g. Olson, 1967). A free rider chooses to receive a higher payoff for defecting than for cooperating, even though contributors get a higher payoff if they all cooperate (Olson, 1967; Hardin, 1968; Mannix, 1991). Here, knowledge free-riding is taken to mean that an individual learns but hides the learning process from the knowledge giver (Gaechter, von Krogh, and Haefliger, 2010). Sometimes, the free rider will be able to obtain the same, or even more, benefits from the knowledge as a contributor (see also von Hippel and von Krogh, 2001). This raises the question, if people are prepared to give knowledge, why should others be interested in reciprocating?

It should be noted that diverse interests in the organization do not necessarily have a negative effect on organizational outcomes: think of innovation. Since knowledge and interest are connected, in the sense that individual interests influence an individual’s learning, it stands to reason that diverse interests give rise to the miscellany and creativity organizations need to extract value from sharing knowledge. Moreover, when the ‘mean level’ of knowledge in a group is not sufficient for solving a particular problem, diverse interests may well facilitate knowledge sharing (e.g. Oliver, 1993). For example, a team of (self-interested) engineers might realize that the firm has not been efficient in adopting a new manufacturing technology, because they lack the basic knowledge needed for its implementation. Yet, engineering pride or a concern with building human capital makes each individual engineer interested in acquiring the new knowledge. This might be highly complex and demanding, requiring the dedicated effort of several people studying and acting together. Even if individually each thinks this new technology has potential for very different reasons, selfish or otherwise, the only way to realize individual interests is to learn jointly, especially under time pressure.

Given distributed and diverse interests, there are several ways to theorize why, and under what circumstances, people share knowledge in organizations. Two fruitful areas are concerned with agency and the community.