Contemporary Perspectives on Organizational Social Networks: Volume 40

Is social network analysis just measures and methods with no theory? We attempt to clarify some confusions, address some previous critiques and controversies surrounding the issues of structure, human agency, endogeneity, tie content, network change, and context, and add a few critiques of our own. We use these issues as an opportunity to discuss the fundamental characteristics of network theory and to provide our thoughts on opportunities for future research in social network analysis.

Traditionally, organizational theory has been a receptacle of methods and mechanisms from network theory. In this paper, we argue that organizational theory can also be an active contributor to network theory’s conceptual development. To that end, we make explicit a theoretical strategy that has only been used informally by network theorists so far, which – following Vaughan (2002) – we refer to as analogical theorizing. Using the basic correspondence between dyadic relationships as the most minimal form of “organization,” we show that processes and mechanisms extracted from various theoretical strands of organizational theory can be mapped onto the dynamics of social relationships. This allows us to build novel theoretical insight as it pertains to issue of relationship emergence, maintenance, and decay in social networks.

This paper proposes an agentic model of new social capital creation in organizations. The core concepts are “making pipes” and “using pipes.” A “pipe” is a metaphor in network theory for the connection between two nodes through which something flows. “Making pipes” means that members of an organization are agents who can build new pipes. “Using pipes” refers to how a new pipe is utilized. Several illustrations and examples are provided. Overall, this model illustrates how human agency drives the creation of social capital and the evolution of networks in organizations.

We propose a framework of constrained agency grounded in the actors’ resources and motivations within their structurally constrained context. Structural positions influence the resources available to actors and color the motivations that shape their actions. Resources equip actors to exert agency, while motivations propel them to do so. We derive a typology of network actions and illustrate how the form of constrained agency through which a particular network action is taken can affect actors’ ensuing structural positions and the nature of the constraints they subsequently face. Our conceptualization of constrained agency identifies new sources of endogenous change in network structure.

Organizations are embedded in multiple interdependent networks comprising different types of relationships, which are managed by different functional units inside each organization. We define tie transfer across networks as the influence of relationships in one network on relationships in another. We argue that the probability of tie transfer will depend on the differences in context in which relationships are formed in two networks, on the past dynamics of relationships in a network into which the tie transfer takes place and on the relative salience of different networks. Our paper develops these conjectures into a relational multiplexity perspective.

I integrate several literatures on how networks relate to firms and markets. While the logic of strong ties (based on mutual goodwill, trust, and commitment) is distinct from the style of interaction classically associated with hierarchical firms and free markets, such ties in fact depend on the legal environments constituted by firm and market. A key role is played by credible commitments to refrain from exercising rights to control others and exit relationships with them. But since such commitments to employees conflict with commitments to contractors, strong ties may be prevalent internally or externally, but not in both simultaneously.

We argue for a broadened approach to brokerage by distinguishing between brokerage emphasizing a particular structural pattern in which two otherwise disconnected alters are connected through a third party (“brokerage structure”) and the social behavior of third parties (“brokerage process”). We explore a processual view of brokerage by examining three fundamental strategic orientations toward brokerage: conduit, tertius gaudens, and tertius iungens that occur in many different forms and combinations. This processual view is especially relevant in increasingly complex and dynamic environments where brokerage behavior is highly varied, intense, and purposeful, and has theoretical implications for studying multiplexity, heterogeneity, and brokerage intensity.

The structural holes to which a person is connected are embedded in a broader organization or market. High status in the broader context signals a reputation that can make a would-be broker more attractive, more likely to engage opportunities to broker, and allay audience concerns about proposed brokerage. The implications are correlation and contingency. We offer illustrative evidence of both implications and conclude that status and structural holes are so closely related in concept and fact that advantage is more clearly revealed when the two network forms are analyzed together as complements defining the hubs in a network.

Weak organizational actors can overcome the consequences of their dependence by securing the control of valuable resources or by embedding dependence relationships into social networks. While these strategies may not eliminate the underlying dependence, they can curtail the ability or the willingness of the stronger party to use power. Embedding strategies, however, can also have unintended consequences. Because the network structures that confer power to the weak are inherently more stable, they can persist beyond the point of being beneficial, trapping weak actors into unsuitable network structures. The power of the weak can thus become the weakness of the strong.

Using Simmel’s external threat–internal cohesion hypothesis, I argue that a group that succeeds in nullifying the threat that it faces will tend to become increasingly fragmented as a consequence. I illustrate this process by drawing on a study of the changing nature of cohesiveness among the leaders of large American corporations from the mid-twentieth century to the present. I use this historical case to develop a series of propositions about the relations among collective action, network structure, and political outcomes.

I draw on psychological and sociological theories of affect to depict the relatively stable set of moods and emotions that an individual experiences in social interactions with a given person (relational affect) as a fundamental engine of social action influencing both how and with whom employees perform assigned tasks. I discuss an approach to define and measure relational affect that complements the typical network approach to affect. I then explore motivational mechanisms through which relational affect influences task tie formation and functioning. I conclude that relational affect contributes directly to individuals’ ability to achieve task goals, and to organizational functioning generally.

Most network research in organizations assumes away the dissociative forces instantiated in negative ties, instead pursuing ties that reflect only associative forces, to the detriment of understanding organizational networks. This essay provides a brief history of negative tie research in organizations; discusses different definitions of negative ties, situating them within the tripartite model of interpersonal attitudes; suggests alternative paths to network dynamics when considering negative ties; covers existing and suggested paths to studying personality antecedents of negative ties; and briefly reviews the research on the consequences of negative ties in organizations and suggestions for future work.

We reformulate regression modeling so that ideas often associated with field theory and social network analysis can be brought to bear at every stage in the computation and interpretation of regression coefficients in studies of organizations. Rather than “transcending” general linear reality, we seek to get more out of it. We formulate a dual to regression modeling based on using the variables to learn about the cases. We illustrate our ideas by applying the new approach to a database of hundreds of violent extremist organizations, focusing on understanding which groups use or pursue unconventional weapons (chemical, biological, radiological, nuclear).

I assess the extent that the egonet method of collecting data permits accurate inferences to the true structural features of the network. This inference problem has three critical components: local-global inference error, ego distortion, and feature inference error. Analysis of four data sets indicates that structural features can be predicted by estimates generated from egonets in some instances, but more often than not they are not good predictors.

I evaluate the accuracy with which respondents report egocentered network data. I find that 72% of the time ego’s assessment of the tie between a pair of alters corresponded with both alters’ report of the tie between themselves, and 87% of the time, ego’s report of the tie between a pair of alters is in agreement with at least one of the alters. I discuss the implications of the findings for the debate over the merits of egocentered network data.

Social networks are not just patterns of interaction and sentiment in the real world; they are also cognitive (re)constructions of social relations, some real, some imagined. Focusing on networks as mental entities, our essay describes a new method that relies on stylized network images to gather quantitative data on how people “see” specific aspects of their social worlds. We discuss the logic of our approach, present several examples of “visual network scales,” discuss some preliminary findings, and identify some of the problems and prospects in this nascent line of work on the phenomenology of social networks.

Bipartite networks (e.g., software developers linked to open-source projects) are common in settings studied by organization scholars. But the structure underlying bipartite networks tends to be overlooked. Commonly, two modes are reduced to one mode for analysis, causing loss of information. We review techniques for projecting 2-modes onto 1-mode and discuss 2-mode measures of clustering. We also address the potential for 2-mode theory development concerning (a) how change in one mode influences change in the other, (b) the question of two types of agency, and (c) how diversity in one mode is a substitute for diversity in the other mode.

Knowledge is critical for employee and firm success. We show that being perceived as prototypical organizational members is a source of prominence in knowledge exchange that operates beyond preexisting communication or affective relationships. Self-categorization processes produce – through depersonalization – a positive attitude among the members which represents an autonomous mechanism of social attraction for knowledge exchange, while social network mechanisms are triggered by interpersonal attraction. Our findings also suggest that including perceived members’ prototypicality can avoid a potentially spurious relationship in assessing the role played by social identity and categorization theory in explaining attitude and behaviors.

This paper provides an analytical theory of appropriateness judgments that introduces structural dimensions in the study of social rationality of organization members. This approach helps explore the coevolution of members’ relative position in structure and normative choices in their organization. Illustration of this approach is based on the study of controversial judicial decisions and dynamics of advice networks in a courthouse where lay judges have to choose between punitive and nonpunitive awards in cases of unfair competition in business. In this case, coevolution is facilitated by an endogenous process of centralization–decentralization–recentralization of advice networks over time, and by use of a procedural “weak legal culture” that helps align and homogenize conflicting normative choices among organization members. It is suggested that this approach to social rationality helps revisit our understanding of social processes, in this case collective learning and secondary socialization in organizations and flexible labor markets.

The distinction between network theories and theories of networks is particularly salient in studying social status because social status is both a consequence and an antecedent of network ties. Status is a consequence of network ties because it is conferred by interdependent acts of deference connecting a sender and a recipient. Status is also an antecedent of network ties because it affects individual preferences for social interaction which produce distinct forms of preferential attachment. A new generation of stochastic actor oriented models (SAOM) for social networks is now available that may help to integrate network theories and theories of networks.

We conceptualize corporate social capital within the context of Chinese guanxi culture. We assert that the formation and mobilization of corporate social capital are culturally and institutionally contextualized. Building upon a relational approach to corporate performance, we examine culture-sensitive properties of Chinese guanxi and compare guanxi social capital with non-guanxi social capital. We then explain why guanxi-based corporate social capital is of growing significance to the Chinese transitional economy in an era of increasing market competition and institutional uncertainty. We conclude by proposing a research agenda about the roles that guanxi-based corporate social capital plays for boosting corporate performance.

Recent labor market research has called into question whether social capital effects are causal, or are spuriously due to the influence of social homophily. This essay adopts the demand-side perspective of organizations to examine the causal status of social capital. In contrast with supply-side approaches, we argue that homophily is a key mechanism by which organizations derive social capital. We develop an approach to bolster inferences about the causal status of social capital, and illustrate these ideas using data from a retail bank.

Online communities form a challenging and still-evolving field for social network research. We highlight two themes that are at the core of social network literature: formative processes and structures, and discuss how these might be relevant in the context of online communities. Processes of tie formation might evolve differently in online communities. Second, we discuss how network structures emerge in different ways than previously studied, and should therefore be interpreted differently.

Long-standing traditions of long-distance collaboration and networking make scholars a good test case for differentiating hype and reality in distributed, networked organizations. Our study of Canadian scholars in the GRAND research networks finds that they function more as connected individuals and less as members of a single bounded work group, often meeting their needs by tapping into diversified, loosely knit networks. Their internet use interpenetrates with in-person contact: the more they use one, the more they use the other. Despite digital networking, local proximity is important for collaboration and seniority for inter-team and interdisciplinary boundary spanning.