47 / 372
臺大管理論叢
第
26
卷第
2
期
47
individuals to reach one another when sending and receiving information and knowledge.
Second, meaningful communication typically requires some sharing of context between
two parties (Boisot, 1995; Boland and Tenkasi, 1995; Campbell, 1969), reflecting two facets:
shared language/codes and shared narratives (Nahapiet and Ghoshal, 1998). Cognitive social
capital includes both shared language and codes (Arrow, 1974; Cicourel, 1973; Monteverde,
1995) and shared narratives (Orr, 1990). The sharing of a common language can help people
access information. Moreover, language affects individuals’ perceptions (Berger and
Luckmann, 1966; Pondy and Mitroff, 1979), and codes organize data into perceptual
categories and provide a schema for individuals to observe and interpret the environment.
Thus, language filters out non-existent elements and filters in existent ones. In addition,
cognition can arise in either information mode or narrative mode (Bruner, 1990). Information
mode refers to a process of knowledge creation from a rational analysis, while narrative
mode suggests systematic use of narratives such as fairy tales, myths, and legends. Orr
(1990) shows that narratives in the form of stories with seemingly insignificant details can
facilitate exchange of practices and tacit experiences between technicians.
Third, relational social capital refers to aspects of individuals’ particular relationships,
such as respect and friendship, through which individuals fulfill social motives, such as
sociability, approval, and prestige (Nahapiet and Ghoshal, 1998). Relational social capital
consists of four facets: trust (Fukuyama, 1995; Putnam, 1993), norms (Coleman, 1990;
Putnam, 1995), obligations and expectations (Burt, 1992; Coleman, 1990; Granovetter, 1985;
Mauss, 1954), and identity and identification (Hakansson and Snehota, 1995; Merton, 1968).
Trust contains the following four aspects: (1) belief in positive intent to the other party
(Ouchi, 1981; Pascale, 1990; Ring and Van de Ven, 1994), (2) belief in the other party’s
competence and capability (Sako, 1992; Szulanski, 1996), (3) belief in the other party’s
reliability (Giddens, 1990; Ouchi, 1981), and (4) belief in the other party’s perceived
openness (Ouchi, 1981). Previous studies have shown that with high levels of trust,
individuals are more willing to exchange knowledge (Nahapiet, 1996; Ring and Van de Ven,
1992), combine different information (Luhmann, 1979), and exhibit greater openness to the
potential for value creation.
“Norms” mean that the socially defined right to control an action is held not by the
individual but by others (Coleman, 1990), indicating that individuals’ behavior is controlled
by informal rules in the social network. Norms can affect individual behavior of exchanging
and creating knowledge in the organization. For example, Leonard-Barton (1995) found that
norms of interaction were important in the creation of knowledge, including willingness to