46 / 372
發言或緘默：心理安全與自我效能在社會資本影響社群網站使用者知識分享行為上所扮演的中介角色
46
Zander, 1992; Zander and Kogut, 1995) and requires social communities comprising social
capital (Kogut and Zander, 1992, 1996).
In their conceptualization of social capital, Bourdieu (1986, 1993) and Putnam (1995)
include the actual or potential resources that can be accessed through social networks.
Bourdieu (1986) suggests that social capital comprises both the network and the assets that
can be mobilized through the network. Putnam (1995) argues that social capital is not a uni-
dimensional concept. Nahapiet and Ghoshal (1998) argue that social capital can be separated
into three dimensions: structural, cognitive, and relational. These dimensions have two
characteristics in common: (1) they contain some aspect of the social structure, and (2) they
facilitate the actions of individuals within the structure (Coleman, 1990). Moreover, all three
dimensions of social capital can increase the efficacy of action (Nahapiet and Ghoshal,
1998). For example, loose structural social capital can increase the diffusion of knowledge
(Burt, 1992), while higher levels of trust in relational social capital can reduce the need for
costly monitoring processes (Putnam, 1993). We next discuss how these three social capital
dimensions influence psychological safety and self-efficacy.
First, structural social capital refers to the social system and the networks within it. It
mirrors the impersonal configuration of linkage between individuals and the social system
and also shows the pattern of connections between actors (e.g., who individuals reach and
how they reach them). Structural social capital contains three facets (Nahapiet and Ghoshal,
1998): (1) the presence (or absence) of network ties between actors (Feld, 1991; Wasserman
and Faust, 1994); (2) network configuration (Krackhardt, 1994) or morphology (Tichy,
Tushman, and Fombrun, 1979), which describes the patterns of linkages; and (3)
appropriable organizations, which provide accessibility of knowledge exchange to members.
Because networked relationships represent a valuable resource in the form of “information,”
network ties provide the channel to access the resources, decrease the search time for
resources, and increase the available opportunity to receive the resources. Network
configuration, or the density, connectivity, and hierarchy of network ties (i.e., how people
link to each other in the social network), can also affect the transmission of resource. These
three elements vastly affect the flexibility of information exchange. For example, a dense
network is efficient in the transmission of resources but inefficient in the spread of diverse
information, while weak social network ties facilitate information search but impede
information transfer (Hansen, 1996). Last, social capital (e.g., ties, norms, trust) developed in
one social setting can be transferred to another setting, such as the family, religious
affiliations, and work (Fukuyama, 1995). These appropriable organizations give access to