generic propecia

Social Networks

Social networks—structures of relationships linking social actors—are omnipresent in contemporary society. People often obtain information about such things as job opportunities, housing, and medical care through interpersonal contacts rather than from formal sources such as the mass media. Networks provide emotional support in times of crisis as well as instrumental aid such as help with household tasks. Identities are constituted by locations in networks; opinions are formed and decisions are made in light of information and conformity pressures that flow through network linkages. Also, social networks are important channels through which both infectious diseases and innovations are diffused.

Ties among individuals in social networks give rise to important larger-scale social patterns. Levels of socioeconomic or ethnoreligious segregation in a society, for example, reflect the degree to which social ties such as marriage and friendship are confined to sets of persons with a common social status or heritage. Such characteristics become salient as markers of differentiation to the degree that they serve as bases for the formation of intimate social relationships. Networks that link individuals to supraindividual units such as work organizations and voluntary associations may serve as modes of integration or separation. Thus, the study of networks contributes to the linking of micro and macro levels of analysis in sociology.

Many macro-level social phenomena can be understood as networks. Increasingly, systems of production for goods (such as automobiles) and services (such as care for the severely mentally ill) are located in multiorganizational fields consisting of separate but interdependent units linked by contingent cooperation rather than in self-contained units administered through elaborate formalized sets of rules. Innovations in corporate strategy and governance diffuse through overlapping boards of directors and other interorganizational structures. Patterns of consensus and cleavage in community and national politics are shaped by alliances and conflicts in networks of governmental agencies, interest groups, and party organizations. In international relations, network ties among nation-states and international organizations define geopolitical alignments.

Emphasis on social networks grew as a result of substantive observations about contemporary society. Many early twentieth-century observers posited that large-scale transformations associated with industrialization—especially urbanization, bureaucratization, and the development of mass media— led to a ‘‘mass society’’ of atomized individuals in which formal, special-purpose ties supplanted diffuse interpersonal relations.

Several lines of research, however, pointed to the continuing vitality of social ties. Industrial sociologists found that informal structures were crucial to the day-to-day functioning of work organization. Indeed, workplace social networks came to be seen as a solution to the inflexibility and excess formalization of bureaucracies and as important incentives for (or impediments to) the performance of individual employees. Urban sociologists found that friendship, neighboring, and informal assistance remained prominent in large cities, although technological innovations in transportation and communication reduced the extent to which the formation and maintenance of those social ties were constrained by spatial considerations. Rural-to-urban and international migrants were not rootless citizens in a normless society but tended to settle in districts populated by persons from their places of origin. In contrast to the expectations of theories predicting protest and activism among those marginal and peripheral to society, researchers found that those active in social movements tended to be drawn from among the persons best integrated into communities, and social ties proved to be important channels through which new members were recruited to social movements.

A social network perspective highlights the interdependence among social actors. This extends beyond the competitive interdependence of actors competing for shares of a stock of scarce resources to encompass obligations and commitments that accumulate as a result of past social interaction. Individual action is embedded in, and therefore continually affected by, preexisting ties between specific actors (Granovetter 1985). In some theories, individuals are viewed as largely passive recipients of environmental pressure; in this structural emphasis, social networks constitute constraints that limit an actor’s discretion. An alternative view makes more room for individual agency, viewing networks as structures of opportunity or social resources. Assuming a context of constrained voluntarism, it treats individuals as proactive, self-interested agents who use networks to manipulate outcomes to their advantage (Haines 1988; Emirbayer and Goodwin 1994).

Social scientists have used the term ‘‘social network’’ metaphorically for some time. Beginning in the 1970s, however, scholarly attention to the analytic development of the social network approach increased. It is now seen as a distinct specialty within several social science disciplines, especially sociology and anthropology, and has many adherents in professional schools, particularly those of business administration and public health. Mathematicians and statisticians have participated in this work, especially the development of novel techniques for studying relations between interdependent social units. These methods are distinct from the conventional ones used to study relations between variables within presumably autonomous units. The theoretical assumptions of formal network models are often implicit (Granovetter 1979), and a distinct ‘‘network theory’’ has not developed. Contemporary studies of social networks instead draw on diverse sociological and social psychological theories.


Some foundations of a methodology for studying networks of social relations were laid in Moreno’s Who Shall Survive? (1934). Moreno coined the term ‘‘sociometry’’ to refer to methods for describing group structures and individual positions within them. His work focused on the affinities and disaffinities of individuals for one another, and his ‘‘sociometric test’’ accordingly stressed affective choices and rejections. These network data were mapped in ‘‘sociograms’’ in which persons were located at different points, with lines indicating the connections between them. Such graphic representations are also common in contemporary network analysis (Figure 1). On the basis of their locations in sociometric networks of affect, individuals were classified as attractive, isolated, rejected, and so forth.

A Social Network Diagram

Figure 1: A Social Network Diagram

Jennings collaborated with Moreno in developing sociometric methods; her (1943) work reported studies of attractiveness and emotional expansiveness. Sociometry as practiced by Moreno and Jennings had an applied component: They attempted to use sociometric measurements as a basis for rearranging groups to enhance both group functioning and individual creativity.

Displays such as sociograms are extremely useful as visualization devices but do not facilitate formal analysis. Forsyth and Katz (1946) and Luce and Perry (1949) were among the first to attempt to surmount the limitations of sociograms by representing networks in the form of matrices. They argued that this would reduce the subjectivity of statements about sociometric structure and allow objective identification of chains, groups, and cliques in network data. In later studies, the application of graph theory (Harary et al. 1965) to the analysis of networks has extended those early efforts.

A second set of influences on the development of social network analyses has emanated from the fieldwork of social anthropologists in complex societies. Those analysts observed that the categorical concepts of a structural-functional approach were insufficient for the study of societies in which not all behavior was regulated by ‘‘corporate groups’’—institutions of kinship, community, or work. Barnes (1954) is credited with the first use of the term ‘‘social network’’ to refer to a set of existing social relationships as distinct from cultural prescriptions about the construction of such ties. In its initial formulations, the concept was used to refer to informal or extrainstitutional links, but it was soon noted that formalized relations and groups also could be analyzed as networks of interactions.

Classic studies in the anthropological tradition display the same variability in theoretical orientations seen in present-day work: Bott (1957) treated social networks as sources of norms prescribing an appropriate allocation of tasks between spouses, while Boissevain (1974) stressed the potential use of networks by maneuvering, selfinterested entrepreneurs.

Structural Properties of Networks

Social networks are studied from many standpoints, including the sociocentric and egocentric perspectives. Researchers with a sociocentric orientation examine complete networks of actors and relations, studying the global properties of a network and characterizing the position of any given actor by reference to all the others. Egocentric network analysis takes the perspective of individual actors and the ‘‘personal networks’’ surrounding them, focusing on the local structure of each actor’s interpersonal environment.

Figure 1 depicts a complete network with fourteen actors, with each one represented as a circle or square. Lines connecting those points indicate relationships between pairs of actors, such as communication ties and expressions of affect. An actor’s ‘‘first-order’’ egocentric network consists of the other actors to which it is linked and the relationships among them; actor G’s egocentric network includes C, D, and F, while that of H consists of B, L, and N. The ‘‘second-order’’ zone of a focal actor’s egocentric network includes those to which the actor is linked via one intermediary; the second-order zone of G’s network consists of actor A, while that of H includes both A and I.

Arrows indicate directionality; for example, actors A and E are involved in a reciprocal relationship, while A and C are linked by an asymmetric tie in which the relationship flows in only one direction. Indirect relationships join actors to one another through intermediaries: for instance, H and I are indirectly linked through B.

The density of a network reflects the overall intensity of connectedness among actors. In Figure 1, sixteen of the ninety-one distinct pairs of actors have direct relationships, and so the network has a density of 18 percent. There are, however, wide variations in the density of the egocentric networks surrounding actors: G is in a very dense or closely knit locality, while H is in a sparse or loosely knit region.

Diagrams of more complicated networks might use lines of different thicknesses to represent relationships of varying intensity or different types of lines to show distinct relationships, such as providing emotional support versus giving instrumental aid. Different types of actors can be represented by different symbols; the actors in circles in Figure 1 might be men, while those in squares might be women. When a relationship typically joins actors that have the same attributes or statuses, as in Figure 1, it is said to display homophily.

Social ties, especially those involving positive sentiment, often create transitive configurations. In Figure 1, for example, transitivity implies that by virtue of the links between C and D and between C and F, there will also be a tie between D and F. Strong tendencies toward transitivity create closure and tend to fragment a group into distinct cliques: The mutually interconnected set of four actors (C, D, F, and G) provides an illustration. Relationships that do not conform to the transitivity principle are important sources of integration between otherwise separate parts of a social structure (Granovetter 1973): in Figure 1, the I–B and A–B ties exemplify bridges of this kind.

Actors may be central in a network because they are involved in many relationships, are in a controlling position between other pairs of actors, or are comparatively close to others (Freeman 1979). Sociometry referred to central actors as ‘‘sociometric stars.’’ In Figure 1, actors A and B occupy relatively central locations. Those such as E, J, and L are peripheral within the network. Actor M is not involved in any relationships and is said to be an isolate.

Two actors are said to occupy the same position in a social network when they have profiles of relationships to other actors that are identical in a particular way (Borgatti and Everett 1992); these actors are therefore substitutable for each other from an observer’s standpoint. Two actors having the same relationships to the same others are structurally equivalent. In Figure 1, actors J and K both have a reciprocal relationship with actor I but no direct relationship with each other or any of the other actors. J and K are thus structurally equivalent, as are three other pairs: C and D, F and G, and L and N. Actors are said to be role-equivalent when they have the same types of relationships to the same types of other actors. Actors L and K are roleequivalent, but not structurally equivalent; although they occupy similarly peripheral locations in the system of relations, L’s direct link is to H, while K’s is to I.

Additional patterns often appear in networks that involve more than one type of social relationship. Exchange patterns arise when a flow of one type in one direction is directly or indirectly reciprocated by a flow of a second type. A multiplex pattern occurs when the relationship between two actors consists of two or more distinct strands, such as kinship and emotional support; a uniplex relationship has only one strand.

Concepts and methods for identifying subgroups within networks have drawn much attention. The two dominant approaches focus on cohesion and equivalence as grouping principles. Techniques that emphasize cohesion locate subsets of densely interconnected actors; fully connected cliques are a limiting case. Blockmodel analysis (White et al. 1976) and related positional methods group equivalent actors together. This yields considerable flexibility. Social positions can be defined by the common ties of actors to outsiders rather than by actors’ links to one another; for example, those in ‘‘broker’’ roles occupy mediating locations between other positions but are not necessarily linked to other brokers. Moreover, positional approaches are not confined to a single type of tie; they can identify subgroups on the basis of patterns in multistranded relations rather than focusing on a single type of tie extracted from its context, as approaches resting on cohesion generally must. Thus, positional analyses of international trade relations may examine flows of raw materials and flows of processed goods simultaneously.

The Structuring of Networks

Rational choice explanations of the formation of network ties are based on exchange theory: Social relations form when actors depend on one another for resources. Related behaviorist accounts stress a reinforcement history. An exemplar is Blau’s (1955) description of the exchange of advice for expressions of deference among coworkers in a bureaucracy. The terms of exchange and thus the relative power of the actors involved depend on the number of alternative actors who control different resources, the extent of unity among those with a given resource, and the parties’ relative interests in outcomes controlled by others (Cook 1982; Burt 1980).

Exchange-theoretic reasoning provides a basis for some commonly observed micro-level network patterns. To maintain autonomy and avoid power disadvantages, actors should avoid asymmetric relations in favor of reciprocal ones. Multiplex ties in which relations of solidarity overlie instrumental exchange links can protect people against exploitation by actors in positions of power. These ideas have played an important role in resource dependence theories of interorganizational relations. Such theories suggest, for example, that interdependent organizations tend to form network ties such as long-term contracts, joint ventures, interlocking directors, and mergers (Pfeffer 1987).

Some attempts to explain tendencies toward homophily are preference-driven accounts in which people actively seek out similar associates. From this viewpoint, communication is easier if people share implicit premises regarding interaction and trustworthiness in the face of uncertainty is enhanced if partners can assume that they have shared interests and predispositions.

Other lines of theorizing about the sources of homophily stress the structure of opportunities for association. Feld (1981) observes that most relationships arise within ‘‘foci’’ of association such as families, neighborhoods, workplaces, and voluntary associations. When segregating processes create foci composed of persons with similar attributes, they create systematic biases toward homophily. Blau’s macrosociological theory (Blau and Schwartz 1984) postulates that networks vary with opportunities for association. Blau shows, without any assumptions about preferences, that intergroup relations—the converse of homophilous ones— are more likely for small than for large groups, when there is great inequality or heterogeneity in a population instead of little, and when different characteristics (e.g., socioeconomic status, race, ethnicity) that structure the formation of social ties are intersecting (or uncorrelated) rather than consolidated. The implications of these factors for intergroup relations depend on the degree to which they penetrate into substructures such as Feld’s foci: Distributional effects on intergroup association are strongest when heterogeneity, inequality, and intersection lie within rather than between substructures.

Transitivity has been most intensively studied for relations of positive sentiment; there the theoretical case for it rests on balance theories that posit pressures toward cognitive consistency (Davis and Leinhardt 1972). Tendencies toward closure also may result from increased opportunities for contact resulting from the copresence of actors in a triad, or an actor in contact with two others may facilitate or serve as a guarantor for a venture involving the other two. Principles of expectation states theory predict transitivity in dominance relations (Fararo and Skvoretz 1986).

Consequences of Networks

A network perspective lends itself to the construction of theories at multiple levels. Contextual theories that examine the effects of an actor’s position in a network on achievement, well-being, and other individual-level outcomes are common, and extensive research literatures have developed around some of them. Network entrepreneurship, diffusion and influence, social support, and power are among the most prominent themes in such works. Group-level theorizing about the properties of complete networks is less typical, although there have been important efforts in this direction. As Coleman (1990) stresses, group-level theorizing is a demanding enterprise.

Granovetter’s (1973) discussion of ‘‘weak ties’’ has been a fruitful source of ideas for network analysts. Although Granovetter defines tie strength in terms of dyadic content (intimacy, intensity, exchange of services and time commitments), the ‘‘strength of weak ties’’ arises from their location within a network structure. Weak ties are less subject than strong ones to the transitivity pressures that induce closure and thus are more likely to be bridges that join distinct subgroups, thereby serving as channels for integration and diffusion. Different contextual theories positing network effects may stress the virtues of either strong or weak ties.

The purposive use of networks as individuallevel ‘‘social capital’’ is a prominent theme in writing about network effects. Much research has been done on how networks facilitate or impede instrumental actions, particularly job seeking. Granovetter (1995) stressed the informational advantages of wide-ranging networks composed of many weak ties. He reasoned that such networks are likely to connect actors to diverse information sources that provide novel information and access to powerful others; by contrast, persons in a densely connected clique are apt to have similar social standing and know similar things. Other writing in this vein emphasizes the content rather than the form of networks. Lin (1990) contends that networks composed of highly ranked contacts are the most advantageous to an actor. Weak ties may facilitate access to social resources, but the aid such contacts can provide reflects their power and influence rather than the type of channel that links them to the actor.

Burt (1992), developed influential ideas about network entrepreneurship. Like Granovetter, Burt stressed the benefits of loosely knit networks for obtaining information quickly. Structural holes refer to the absence of connections between an actor’s contacts in a sparse network; thus, in Figure 1 there are holes separating actor B’s contacts (A, I, and H). Beyond providing actors with the ability to acquire information sooner than competitors can, Burt argues that networks rich in structural holes confer control benefits. Actors at the middle of networks that bridge many holes gain autonomy and leverage over others by virtue of their unique interstitial positions. They are able to place their contacts in competition with one another, avoid excessive dependence, and negotiate favorable bargains. Empirical applications in diverse settings lend support to these ideas: Structural holes have proved advantageous in studies of competition for promotions and bonuses in firms and studies of comparative profit margins in manufacturing industries.

Studies of the role of networks in diffusion and influence processes echo debates over cohesion and equivalence as bases for defining network subgroups (Marsden and Friedkin 1993). This work posits that networks provide actors with a basis for constructing reference groups and social comparisons (Erickson 1988). Approaches that stress the socializing potential of cohesive ties contend that people look to close associates for guidance toward appropriate attitudes or conduct in conditions of uncertainty. Network closure thus is viewed as a source of locally defined norms, as persons in dense networks respond to consistent conformity pressures from their strong ties.

Burt (1987) suggests an important alternative process of diffusion or influence, arguing that in seeking normative direction, actors look not to their close contacts but to their competitors. They engage in a process of role taking, examining the views held or the actions taken by structural peers who occupy similar positions within a system of relations. Conformity to norms here is less a matter of social pressure from the environment than the result of an actor’s mimicry of others (DiMaggio and Powell 1983).

An extensive research literature on the sociology of health and illness draws a connection between social networks, exposure to stressors, the availability of social support, and well-being (House et al. 1988; Thoits 1995). Social networks are ‘‘structural’’ elements that may facilitate access to supportive contacts but also may expose someone to additional stressors, conflicts, and demands; hence, the quality as well as the structure of ties within networks must be considered. A ‘‘direct effect’’ view states that receiving social support enhances health and well-being in all conditions. The ‘‘buffering hypothesis’’ suggests a conditional effect in which those with supportive networks have less severe responses to stressful life events.

Several mechanisms have been suggested to account for the way in which aspects of social networks translate into social support and measures of physical and mental well-being. Some of these suggestions have a Durkheimian emphasis, reasoning that people integrated into dense networks— that is, strongly tied to a number of partners who are strongly tied to each other—have better-defined social identities, stronger senses of internal control, and more positive self evaluations, which in turn may lead to the use of more effective coping strategies under stress. Collins (1988) suggests that network density indicates that an individual is integrated into the ‘‘interaction rituals’’ of a solidary group, a situation that produces moral sentiments and energies that enhance well-being.

It is relatively well established that the availability of a strong tie or confidant has health-promoting effects (Thoits 1995). Theories stressing the availability of support resources imply that the size of a personal network is linked to physical and psychological health. Large, diverse networks are thought to facilitate adjustment to change. Contacts in one’s network may offer companionship as well as both instrumental and emotional assistance; they can both provide health information and focus attention on it. Social contacts also may be sources of regulation and control, providing social feedback on one’s behavior and performance, discouraging harmful behaviors such as substance abuse, and encouraging beneficial ones such as adherence to treatment protocols. It is increasingly recognized that providing social support can be burdensome and that stress as well as aid may emanate from relationships within social networks.

Network theorists have given substantial attention to the connection between an actor’s location in a social network and social standing or prestige. Freeman’s (1979) discussion of the conceptual foundations of centrality measures focused on processes in communication networks. He observed that distinct measures are sensitive to communication activity, the capacity to control the communications of others, and the capacity to avoid the controlling actions of others. Many empirical studies have documented an association between centrality and manifestations of power or influence.

Exchange-theoretic approaches observe that actors with predominantly favorable exchange ratios with others are central within networks of dependency relations and hence acquire power (Cook 1982). No universal principle leads to that connection, however; instead, it depends on a particular form of exchange. For Cook this is ‘‘positively connected’’ or ‘‘productive’’ exchange in which actors must combine diverse resources to be successful; hence, the use of one exchange relation tends to encourage the use of others, and advantages emerge for those in intermediary ‘‘broker’’ positions. Willer (1992) terms these ‘‘flow’’ networks. For Coleman (1990), the requisite conditions include resource transferability or fungibility: For systems of social exchange to develop fully, actors must be able to transfer not only control over resources but also the right to further transfer such control.

Under other conditions of exchange, network positions have different consequences for power. Among these are what Cook (1982) calls ‘‘negatively connected’’ exchange networks and Willer (1992) terms ‘‘restricted’’ exchange networks. In these networks, resources do not flow through positions and the use of one relation rules out or makes less likely the use of others. Matching systems such as marriage and dating markets exemplify negatively connected or restricted networks. In these conditions no special advantages accrue to an actor in a central position; instead, power is concentrated in actors who can exclude others from exchanges (Markovsky et al. 1988). The capacity to exclude others can depend both on the nuances of a network’s structure and on the incentives or rules governing exchange in any given situation. Depending on such subtleties, there may be a nonlinear association between centrality and power; a network also may decompose into small substructures as some potential exchange relations fall into disuse.

Theorizing about network effects at the level of aggregates is less extensive than is that about networks as contexts for individual action. One line of work examines the implications of structural biases toward homophily or transitivity for the spread of diseases and innovations (Morris 1993). Another considers how the centralization or dispersion of networks shapes differences in system performance or effectiveness (Laumann and Knoke 1987). Still another deals with networks and collective action: Social density has been viewed as an infrastructure that allows latent ‘‘interest groups’’ to overcome social dilemmas (Marwell et al. 1988). Granovetter (1973) warns against excessive closure, however: Once formed, a coalition or ‘‘collective actor’’ may lack the social connections required for ultimate political success.

Some theorists have developed the notion of social capital as a collective rather than an individual property. Coleman (1990) suggests that closed social networks can create trust and enforce strong norms that facilitate collective action. Actors in densely interconnected systems can expect to encounter one another frequently in the future, and a reputation for trustworthiness therefore becomes valuable. Moreover, frequent communication among densely linked actors means that information about a failure to honor obligations diffuses quickly and that sanctions can be applied rapidly. Coleman asserts that strong norms of trust are essential to the creation of ‘‘social credit’’: aid or assistance contributed in exchange for future compensation. Preexisting social networks thus are valuable as sources of social capital to the extent that they are appropriable for new purposes.

Different varieties of network theory make contrasting observations about network density. Density can integrate an actor into a subculture, provide a well-defined social identity, create and enforce norms, and promote the production of collective goods it simultaneously may subject one to conformity pressures and limit the diversity of one’s affiliations. Reconciling these seemingly conflicting effects of density represents a challenge in the contemporary study of networks. Understanding the manner in which networks channel and block transactions among the elements of society will remain a crucial and intriguing task for twentyfirst- century sociology.

Further Literature

A number of review articles and collections treat aspects of network analysis in greater depth. Mitchell (1974) provides a useful review of the anthropological approach. Wellman (1983) reviews basic principles from a sociological perspective. Review articles on substantive applications appear in Wasserman and Galaskiewicz (1994). The journal Social Networks (1978–present), edited by Linton Freeman, presents methodological advances and substantive studies. Marsden (1990) surveys the literature on measurement, and Wasserman and Faust (1994) provide a comprehensive review of analytic methods.

Related Links

This Aricle was Written by

This Article was Published in
Second Edition
A Book by

University of Washington, Seattle


Managing Editor
University of Kansas, Lawrence Copyright 2010 - 2012 © All Rights Reserved
  Home | About | Contact | Links