Within- and Between-Team Coordination via Transactive Memory Systems and Boundary Spanning

Abstract

In this article, we suggest that the transactive memory system (TMS) and boundary-spanning literatures are useful for understanding how individuals in team-based collectives can be structured to improve within- and between-team coordination. We argue that such coordination can be facilitated—or thwarted—by boundary-spanning behaviors and patterns of knowledge exchange within and between teams. Our theorizing explains how an existing team TMS can offset the within-team coordination burdens typically associated with boundary spanning and we offer predictions about how these factors interrelate to affect TMS and coordination over time. Finally, our theory underscores significant implications and provides insights for how management practices might improve coordination within and between teams.

Keywords

transactive memory coordination boundary spanning social networks

Introduction

In our increasingly dynamic society, organizations are faced with the challenge of remaining efficient while developing effective collaborative strategies to enhance organizational performance. We observe an increased emphasis on the need for organizations to simultaneously coordinate multiple internal activities and knowledge flows to innovate and perform at high levels (Edmondson & Nembhard, 2009).

Recent research has highlighted the challenges of coordinating activities and knowledge flows within and between teams. For example, the synchronization of teams’ activities and outputs is essential for coordinated action, but task demands and timelines might vary widely across different teams. Furthermore, team members must consistently strive to address both internal team tasks and boundary-spanning activities across teams, presenting trade-offs between accomplishing within-team tasks and between-team coordination. Our conceptualization argues that the transactive memory system (TMS) construct is useful for understanding not only within-team coordination but also between-team coordination. A TMS is a form of collective cognition that allows for efficient information processing and learning in teams (Lewis, Lange, & Gillis, 2005; Moreland & Argote, 2003). Although most of TMS research has emphasized the team-level of analysis, some research investigates the role of TMS in promoting coordination at higher levels of analysis. For example, Peltokorpi (2012, 2014) argued that an organization-level TMS can operate when certain conditions are in place, such as formal human resources practices that motivate and encourage inter-team interactions.

In this article, we delve more deeply into explanations for between-team coordination, which we argue can be facilitated or thwarted by boundary-spanning behaviors and patterns of knowledge exchange within and between teams. In so doing, we offer a generalizable account of the role of TMS in organizations that resolves some inconsistencies in the literature. For example, boundary-spanning research suggests that individuals who are functional generalists, rather than specialists, are best positioned to execute boundary-spanning activities since such activities increase the cognitive load on individuals serving in these roles (Ancona & Caldwell, 1992; Joshi, Pandey, & Han, 2009). However, TMS research demonstrates a performance advantage for teams composed of members who possess specialized knowledge (Lewis & Herndon, 2011), in part because members can expend their cognitive efforts on deepening, rather than broadening, their expertise (Lewis, Belliveau, Herndon, & Keller, 2007). How might organizations optimize both within- and between-team coordination, given these competing arguments? Second, while social network research suggests that boundary spanners who are centrally positioned in a dense within-team network are better able to coordinate knowledge exchanges with others outside of the team (Cross & Cummings, 2004; Ibarra, 1993), recent research on TMS and within-team knowledge exchange networks (Lee, Bachrach, & Lewis, 2014) shows that TMS depends not on dense knowledge exchange networks (which was found to negatively affect TMS development), but instead on particular patterns of knowledge exchanges that form network structures called transitive triads. We explain how the position of the boundary spanner in the team knowledge exchange network impacts within- and between-team coordination. Finally, we explore the effects of time and associated feedback mechanisms on TMS and boundary spanning, identifying the conditions under which coordination can be sustained over time.

Our theorizing applies to teams and collectives engaged in highly interdependent tasks that require a large amount of specialized knowledge (expertise) to perform well. Examples of such collectives include product development or design teams, cross-functional teams, and project teams—each of which can benefit from the development, application, sharing, and integration of knowledge from different domains. We consider three levels of analysis: the individual level (characteristics of the boundary spanner), the team level (team-level TMS), and the “between team” level (the linkage between two teams). We propose that these levels interact, to affect both within- and between-team coordination.

We examine the following research questions:

Research Question 1: How might teams both maintain high TMS and incorporate useful knowledge from outside the team via inter-team boundary spanners?

Research Question 2: What is the ideal position for boundary spanners such that they can effectively facilitate between-team coordination while at the same time contribute to the coordination inside their own teams?

Research Question 3: Given the benefits of maintaining a high TMS, how do the boundary spanner’s knowledge and position contribute to TMS development over time?

The rest of the article proceeds as follows. First, we provide some background about the theories that underpin our arguments about TMS and boundary spanning. Next, we address the aforementioned research questions and develop testable propositions that further explore our theorizing. Finally, we conclude with a discussion of the implications of our conceptualization and highlight opportunities for future research.

Theoretical Background

TMSs

A TMS is a collective cognitive system used for learning, remembering, and communicating information (Lewis & Herndon, 2011; Moreland, 1999). Originally described by Wegner (1987) to explain how people in close relationships implicitly divide responsibilities for knowledge in different but complementary domains, the TMS concept has gained acceptance in the teams literature as a powerful form of team cognition (Ren & Argote, 2011). A TMS develops as team members learn about “who knows what” and come to rely on other members of the team to learn, remember, and communicate information in different areas of knowledge. This division of cognitive labor benefits both team processes and performance, by increasing the quality and depth of the knowledge available for the team’s tasks and by ensuring that critical knowledge is retained and not forgotten. Knowing “who knows what” in the team affords members access to knowledge possessed by others, reduces the cognitive burden on any one member to learn in multiple knowledge areas, and decreases redundant search and learning, while at the same time increasing the corpus of team knowledge available for accomplishing team tasks.

Early research on team TMS by Liang, Moreland, and Argote (1995) identified three interrelated behaviors that can be observed when a TMS exists in a team. First, when there is a TMS, members specialize in learning, remembering, and communicating knowledge from distinct expertise domains. Second, members rely on what they understand about others’ knowledge to decide about what to learn, remember, and communicate—typically, in an area not already “claimed” by another member. This reliance on others is important in members’ decision to specialize—if a member does not have confidence in another’s expertise that member might choose to learn in an overlapping rather than distinctive knowledge area. However, the power of a TMS is rooted in the team’s ability to apply a greater amount of task-relevant knowledge to the team’s tasks, which is reduced when members cannot confidently rely on other members to share the cognitive burden for learning, remembering, and communicating task-relevant knowledge. The third behavior observed in teams with a TMS is smooth and coordinated activity. Coordinated action and communication depends on members providing task-relevant information to the team when it is needed, and members’ knowing where to find information. Moreover, as members come to agree about “who knows what,” it increases the efficiency and efficacy of information search and allocation, thereby enabling the team to apply a large amount of task-relevant knowledge to their tasks. When all three behaviors are observed in concert, one can infer that it is because a TMS is operating in the team (Lewis, 2003; Lewis & Herndon, 2011; Liang et al., 1995).

Empirical research on TMS in teams has produced strong evidence that teams with a TMS perform at higher levels, and that the higher performance can be attributed to the distributed and specialized knowledge that teams enjoy when they divide the cognitive labor for their tasks (Anderson & Lewis, 2014; Liang et al., 1995). The performance effects of TMS are robust across many types of teams and settings, including product development teams (Akgün, Byrne, Keskin, Lynn, & Imamoglu, 2005) air traffic control teams (Smith-Jentsch, Kraiger, Cannon-Bowers, & Salas, 2009), top management teams (Rau, 2006), consulting teams (Lewis, 2004), day care groups (Peltokorpi & Manka, 2008), anesthesia teams (Michinov, Olivier-Chiron, Rusch, & Chiron, 2008), software development teams (Faraj & Sproull, 2000), on-going student teams (Lee et al., 2014), and emergency response teams (Majchrzak, Jarvenpaa, & Hollingshead, 2007).

Patterns of knowledge exchange that affect TMS

Recent research linking social network research and knowledge exchange provides new insights about how the pattern of knowledge exchanges between team members affects team processes and performance (e.g., Dokko, Kane, & Tortoriello, 2014; Lee et al., 2014; Tortoriello, Reagans, & McEvily, 2012). Although a complete review of this literature is beyond the scope of this article, we emphasize here the specific patterns of knowledge flows that form structures called transitive triads, which have been shown to lead to TMS (Lee et al., 2014; Palazzolo, 2005). A triad structure is one in which three actors are linked by ties, which in the context of our arguments, are defined by knowledge flows between those actors. A special type of triad structure—a transitive triad¹—appears to be important to the development and operation of a TMS. For example, Palazzolo (2005) found that in mature organizational teams, the transitive triad structure was more prevalent than would be expected by chance, suggesting that TMS may be related to particular knowledge flow patterns in the form of transitive triads.

We emphasize transitive triad network structures in our theorizing for several reasons. First, there is consensus among social networks researchers that triads—network structures that link three people—are the “minimal size for group phenomena to manifest” (Goh, Krackhardt, Weingart, & Koh, 2014, p. 474). Early theorizing by Simmel (1950) and Heider (1946, 1958) and subsequent social network research (e.g., Burt, 1992; Krackhardt, 1998; Tortoriello & Krackhardt, 2010) suggest that a third person, who is tied to two others, is in a unique position to coordinate information flow and facilitate the integration of knowledge. Lee et al. (2014) used these arguments to show that transitive triad structures in teams are associated with the emergence of TMS, in part because the triadic structure allows members to better identify the member-experts in the team.

Second, transitive triads help members develop a shared understanding of “who knows what” within the team. Transitive triads have long been recognized in the social network literature as promoting efficient and balanced patterns of relations between actors (Heider, 1946; Wasserman & Faust, 1994). People in triadic relationships are also more likely to come to agreement about “information about the social network itself” (Krackhardt & Kilduff, 2002, p. 281). Krackhardt and Kilduff’s (2002) empirical study of cultural agreement in organizations indeed found that triadic relations, rather than dyadic relations, were associated with higher levels of agreement among organizational members about the organization’s culture. Lee et al. (2014) used similar arguments to show how certain patterns of knowledge exchange can establish and reinforce team members’ shared understanding about the location of expertise in the team. Results from that study confirmed that TMS was more likely to emerge in teams in which knowledge exchanges between two members were facilitated and regulated by a third member—a transitive triad structure. A greater number of transitive triads in a team’s knowledge exchange network predicted TMS development and subsequent team performance. The same study examined the role of other network characteristics (e.g., network density, centralization, reciprocity) in the emergence of TMS, finding that the number of transitive triads in a team’s knowledge exchange network explained TMS and subsequent team performance above and beyond these other network features. For these reasons, we emphasize transitive triads in our investigation of the role for boundary spanners in coordinating knowledge flows.

The role of time on TMS

Researchers have acknowledged the temporal nature of TMS emergence and change, although empirical examinations of the role of time on TMS are fairly rare. Brandon and Hollingshead (2004) illustrated the dynamic nature of the TMS construct, exploring both linear and cyclical aspects of TMS development over time. They conducted a longitudinal assessment of TMS to better capture the continually changing context of work teams, rather than assuming a linear progression and using static measures as prior studies had done. Lewis (2004), in a longitudinal study of knowledge-worker teams, found that TMS emergence and development over time (through communication interactions) were positively related to team performance and viability. A computational model by Anderson and Lewis (2014) explored the effects of time on TMS and team productivity by simulating how individual (i.e., specialized learning in a domain) and collective learning (i.e., a shared understanding of who knows what) interact over time. Their simulations not only identified how a TMS might evolve over time under certain conditions but also produced new insights about the factors that promote or disrupt TMS in teams. In the current article, we attempt to further articulate the role of time on boundary-spanning activities and the evolution of TMS within teams.

TMS at the Organizational Level

Although TMS has been studied extensively in small-team settings, a few studies have explored its functionality at higher levels of analysis (cf. Anand, Manz, & Glick, 1998; Jackson & Klobas, 2008; Nevo & Wand, 2005; Peltokorpi, 2014). Peltokorpi (2012) provided a review of studies examining TMS at the organization level, arguing that organizational TMS might best be understood in terms of knowledge exchange networks within and between teams. He posits that information and communication technology (Anand et al., 1998; Nevo & Wand, 2005) and interpersonal knowledge management approaches—such as network linkages (Moreland, 1999)—can support the operation of organizational TMS. In a subsequent qualitative study on organizational TMS, Peltokorpi (2014) investigated how an organizational-level TMS may depend on the existence of both formal and informal coordination mechanisms. His findings show that an organization’s design (e.g., team-based structure), formal mechanisms (e.g., human resource management practices), and employee roles and routines are relevant to the operation of an organizational TMS. Although these important studies helped to further understand how the TMS literature has extended to incorporate contexts beyond the team level of analysis, the studies only briefly acknowledge how the pattern of knowledge exchanges might influence coordination across the organization and how boundary spanners might facilitate knowledge exchange.

We concur with the conclusions drawn by Jackson and Klobas (2008), who conducted a case study examining TMS processes (directory updating, knowledge allocation, information retrieval) in an organizational context, that organizational TMS remains an idealized aggregate of team-level TMS. These and other researchers have argued that organizational TMS and the processes or structures that support it need additional conceptual development (Anand et al., 1998; Moreland & Argote, 2003; Nevo & Wand, 2005; Peltokorpi, 2008), with opportunities to enhance understanding through related theoretical perspectives. The current article aims to advance our understanding of how TMS might operate beyond the team level, by articulating how boundary-spanning activities between teams interact with within-team TMS to affect coordination not only within teams but also between teams.

Boundary Spanning Activity Within and Between Teams

Boundary spanners serve a critical role on teams; they have long been explored for their efforts in coordinating interdependent tasks within the team and managing connections external to the team (Ancona & Caldwell, 1992; Hansen, 1999; Marrone, Tesluk, & Carson, 2007). Moreover, the literature on boundary spanning raises some interesting questions when integrated with TMS research. For instance, in a study looking at the relationship between TMS and performance in mature continuing groups in a rapidly expanding retail company, Austin (2003) suggested that being aware of team members’ external relationships—their boundary-spanning activity outside the team—can improve the team’s TMS because it provides external access to relevant knowledge that team members may not possess.

Team boundary spanning is commonly identified as the combination of activities taken by team members to establish linkages with external stakeholders—external to the team, not necessarily to the organization (Marrone, 2010). These boundary-spanning activities may be especially necessary in complex, uncertain environments where the coordination of specialized knowledge is particularly valuable (Faraj & Yan, 2009). The utilization of boundary spanners has repeatedly been acknowledged as an effective approach for sharing knowledge within and between teams, and mitigating the coordination challenges that exist (Joshi et al., 2009; Peltokorpi, 2014). For example, boundary-spanning activities have been suggested to drive inter-team learning and performance (DeChurch & Mathieu, 2009) and influence coordination in more complex team formations (Davison & Hollenbeck, 2012). More generally, boundary spanners are positioned to regulate information flow within and between teams, with communication activities focused on informing the team about the environment within which it is embedded (Ancona and Caldwell, 1988, 1990; Somech & Khalaili, 2014). Boundary spanners may also act to protect the team’s resources as well as build the team identity to cultivate sustained member commitment to the team’s task (Faraj & Yan, 2009).

Despite the potential role of boundary spanning in affecting knowledge flows between teams, many questions remain unanswered in the literature. For example, how do boundary spanners balance the demands of engaging meaningfully within their own teams while at the same time coordinating across teams? How does the boundary spanner’s position within the team influence coordination within and across teams? What sort of knowledge must the boundary spanner possess, for boundary-spanning activities to be fruitful?

The literature offers contradictory advice on these questions. For example, extant research suggests that a moderate amount of boundary-spanning activity is associated with the highest levels of team effectiveness (Gibson & Dibble, 2013). Marrone et al. (2007) observed that higher levels of team boundary spanning led to reduced role overload, and thus advocate for embedding boundary-spanning tasks within the responsibilities of all team members, while others (Davison, Hollenbeck, Barnes, Sleesman, & Ilgen, 2012) note that interaction among all team members in more complex team arrangements is detrimental to performance. Some scholars (Ancona & Caldwell, 1992; Joshi et al., 2009) have advocated for factors such as functional and tenure diversity to be considered in identifying the boundary spanners. Marrone (2010) also identified alternative boundary-spanning configurations, including the use of experts, rotating members, full and part-time members, and core and peripheral members. Limited empirical research, however, has explored these configurations.

What has been fairly consistent in the literature is that boundary spanners are necessary for facilitating knowledge exchange both within and between teams (Hansen, 2002; Joshi et al., 2009; Peltokorpi, 2014; Tushman & Scanlan, 1981). For example, Davison et al. (2012) advocated for more formal boundary spanning and coordination structures in complex team formations, noting that the extant literature has not yet illuminated what these coordination mechanisms are and how they function.

In the next section that expands on our conceptual arguments, we integrate the TMS and boundary-spanning literatures to inform how the boundary spanner’s knowledge and position can impact within- and between-team coordination of knowledge flows, and we present propositions that further enhance our understanding of the boundary spanner’s role. We also consider the impact that boundary spanner knowledge and position in the knowledge exchange network might have on TMS over time, and subsequently, within- and between-team coordination. Adding a temporal dimension to our theorizing, we expect that boundary-spanning characteristics will affect the maturation of TMS, which in turn affects within- and between-team coordination of knowledge flows, and vice versa. The nature of this “feedback loop” suggests that short- or immediate-term effects may differ from long-term effects of boundary spanning and TMS on coordination. Moreover, some effects may generate virtuous or vicious cycles that affect the potential for coordination.

Conceptual Development

Boundary Spanner Knowledge

A key argument present in the boundary-spanning literature is that activities performed by individuals who are functional generalists, rather than specialists, are best suited to boundary-spanning roles (Ancona & Caldwell, 1992; Joshi et al., 2009). This argument rests on the idea that boundary spanning is itself an effortful activity that increases the cognitive load on individuals in these roles. A boundary spanner is responsible not only for coordinating with units outside of the team but also for ensuring that external knowledge is passed on to members of the team and integrated into the team’s internal activities (Ancona & Caldwell, 1992; Faraj & Yan, 2009; Marrone et al., 2007; Somech & Khalaili, 2014). With a higher cognitive load, a boundary spanner is therefore less able to develop deep specialized knowledge that would otherwise be advantageous to the team of which the boundary spanner is a member. This point is antithetical to the literature on TMS, which stresses the advantages of developing specialized—rather than generalized—knowledge (Lewis & Herndon, 2011).

We imagine that within a team, variety in the individual knowledge that each team member possesses can be described as existing along a continuum, with one end of the continuum characterized by members possessing disparate knowledge and skills with no overlapping territories, and the other end reflecting a substantial amount of overlap in members’ knowledge or skill sets. We consider the space in between these extremes rather than the polar ends of this continuum. For the sake of exposition, we use the terms generalist and specialist to capture the type of individual skill set that the boundary spanner possesses. A generalist boundary spanner possesses individual knowledge and skills that overlap with other members’ knowledge inside and/or outside the team. The “boundary spanner as generalist” argument aligns with the current boundary-spanning literature that underscores the important role that boundary spanners play in managing information flow (Ancona & Caldwell, 1992; Marrone, 2010). We define a specialist boundary spanner as one that possesses task-relevant expertise that is distinctly different from other team members’ knowledge.

We theorize that the division of cognitive labor that is produced with a TMS can offset some of the cognitive burdens typically associated with boundary spanning. Members of a team with a strong TMS have a clear awareness of “who knows what” and are able to quickly access relevant knowledge from other members. A strong TMS also facilitates the dissemination of new knowledge entering the group, which can be quickly allocated to the member expert who is most likely to remember and communicate that knowledge when it is needed for team tasks (Wegner, 1995). These coordination behaviors, which are observed in groups with a TMS (Lewis & Herndon, 2011), are the very activities that typically fall to the boundary spanner to accomplish—the dissemination of external information within the team, its allocation within the team, and the retrieval of team knowledge that can be passed to units external to the team (Ancona & Caldwell, 1992; Faraj & Yan, 2009). When the boundary spanner’s team has a strong TMS, however, the burden of learning, remembering, and communicating information to other teammates is distributed across team members (Lewis et al., 2007). With a reduced cognitive load and relying on the team’s improved information flow and coordination, the boundary spanner is more likely to have the cognitive slack to develop knowledge that is more specialized than generalized.

If the team has a weak TMS, however, the boundary spanner must take on additional coordination responsibilities to ensure that the appropriate information is being transferred from—and received by—the teams linked by boundary spanners. In this case, the type of knowledge possessed by the boundary spanner may not materially affect between-team coordination in the short term. If the boundary spanner has specialized knowledge and the team’s TMS is weak, then between team coordination is compromised because the weak TMS impedes access by external units to high quality knowledge. If the boundary spanner has generalized knowledge and the team’s TMS is weak, then any communication or comprehension benefits that the generalized boundary spanner knowledge might otherwise produce will be of little value because the weak TMS limits the boundary spanner’s ability to identify, disseminate, and retrieve knowledge efficiently inside the team. Although a strong TMS may allow the boundary spanner to develop specialized knowledge that is useful within the team, a weak TMS offers no such opportunity. Indeed, a weak TMS is likely to undermine between-team coordination, independent from the boundary spanner’s knowledge.

The paragraphs above suggest that a strong within-team TMS may provide a boundary spanner with cognitive slack to develop more specialized than generalized knowledge—that is, specialized knowledge that adds value to the team’s knowledge base. Although some past research suggests that between-team coordination may be compromised when a boundary spanner expends efforts on developing his or her own expertise rather than on coordinating between teams, we contend that this tradeoff is not as relevant when the boundary spanner’s team has a strong TMS. A strong TMS offsets the within-team coordination demands that would ordinarily fall to the boundary spanner, allowing even a specialized boundary spanner to simultaneously attend to between-team coordination. Therefore, we propose that when a team has a strong TMS

Proposition 1: Coordination of knowledge flows within and between teams is higher when a boundary spanner possesses more specialized, than generalized, knowledge.

Effects of boundary spanner knowledge over time

Proposition 1 posits that between-team coordination of activities and knowledge flows can be achieved even when a boundary spanner possesses specialized knowledge, as long as the team also has a strong TMS. We note that if the team has a weak TMS, then between-team coordination may be compromised regardless of whether the boundary spanner possesses specialized or generalized knowledge. When considering the role of time, however, the type of boundary spanner knowledge may matter more profoundly. If the boundary spanner possesses more generalized than specialized knowledge, it might ease communications between teams in the short term (as is typically advocated in boundary-spanning research). However, at the same time, generalized boundary spanner knowledge will weaken a team’s TMS, which depends on a clear and well-understood division of cognitive labor along the lines of expertise. A weakened TMS will reduce a boundary spanner’s ability to access the “right” information when it is needed by another team, or to deliver (allocate) external information to the team member who is most able to use and remember that information. When left unchecked, blurry member-exchange linkages will also make the job of the boundary spanner more difficult, resulting in a longer information search process that impedes coordination both within and between teams.

Together, the arguments above imply that a boundary spanner with more generalized knowledge will eventually weaken a team’s TMS, and consequently reduce within-team coordination (because member-expertise linkages are more blurred) and reduce the efficacy of between-team coordination (because of reduced access to useful knowledge). Moreover, without the within-team coordination that a strong TMS affords, the boundary spanner’s cognitive load will be greater, limiting the boundary spanner’s ability to exchange knowledge between teams. Therefore, we propose

Proposition 2: When a boundary spanner possesses generalized rather than specialized knowledge, the boundary spanner’s team TMS will be weakened over time, eventually leading to a reduced capacity for within- and between-team coordination.

Boundary Spanner Position in the Knowledge Network

Social network research has shown that an individual’s network position can influence knowledge sharing opportunities and access to new knowledge (Wasserman & Faust, 1994). For example, individuals with many ties or large egocentric networks have access to knowledge possessed by others. Such individuals are ideally situated to accumulate work-related knowledge (Burt, 1992; Sparrowe, Liden, Wayne, & Kraimer, 2001; Tsai, 2001), and may be ideally positioned to coordinate knowledge exchanges with other individuals external to the team (Cross & Cummings, 2004; Ibarra, 1993). Being in such a position also makes an individual attractive to colleagues that are external to the team because of their extensive access to knowledge (Cross & Cummings, 2004). Yet, they assert a need for future research to further explore the relationship between network characteristics and boundary spanner roles. Indeed, if a boundary spanner is positioned centrally in a knowledge exchange network, the boundary spanner must manage a potentially large number of knowledge exchange interactions, leading to an increased cognitive load and reduced cognitive capacity to address their own task-related work.

Recent research on TMS and within-team knowledge exchange networks offers some new insights about how the network position of a boundary spanner may affect knowledge exchange, within and between teams. We noted earlier that research by Palazzolo (2005) and Lee et al. (2014) suggest that efficient knowledge exchange does not depend on dense knowledge exchange networks, but rather on the presence of transitive triad structures in the team’s knowledge exchange network. When members are embedded in transitive triad structures, they are better able to learn about “who knows what,” and better able to disseminate knowledge to, and retrieve knowledge from, the right member-experts. If a boundary spanner is embedded in one or more of the team’s transitive triads, the boundary spanner would similarly benefit from these efficient patterns of knowledge exchange. In sum, when team members are embedded in transitive triad structures, it improves the coordination of knowledge flows within the team. Notably, this occurs with less effort being exerted by the boundary spanner to carry out dissemination, coordination, and retrieval activities.

We expect that between-team coordination is also improved when a boundary spanner is embedded in a transitive triad in the team’s knowledge exchange network. The improved knowledge exchange process and effective access to—and dissemination of—information that results from being embedded in a transitive triad will help the boundary spanner to engage effectively with other teams, providing a higher payoff for exchanging knowledge with units outside the team. Notably, Austin’s (2003) findings that boundary spanning improves (internal) team TMS suggest that within-team and between-team coordination can indeed be symbiotic. Furthermore, as we discussed in Proposition 1, when a boundary spanner can rely on a strong TMS (which develops from patterns of knowledge exchange forming transitive triad structures—Lee et al., 2014), the boundary spanner is more likely to have the cognitive slack available to deepen his or her expertise in a functional area. This enhances the quality of between-team collaborations that the boundary spanner can engage in because the boundary spanner is interacting as both a team coordinator as well as a content expert.

Alternatively, if a boundary spanner is not embedded in a transitive triad, the boundary spanner will have reduced opportunities to interact or exchange knowledge with other members. This will prevent the boundary spanner from learning about or updating his or her understanding of “who knows what” in the team, and consequently limit the boundary spanner’s ability to allocate pertinent knowledge obtained from outside the team to the appropriate member-experts (Lee et al., 2014). Moreover, an isolated boundary spanner could not benefit from the efficient knowledge exchanges typical of transitive triad structures, forcing the boundary spanner to exert additional effort and cognitive resources to coordinate with each member of the team. This in turn reduces the available resources that the boundary spanner could otherwise devote to between-team coordination. Therefore, we propose

Proposition 3: Coordination within and between teams is higher when a boundary spanner is embedded in, versus isolated from, transitive triads in a team’s knowledge exchange network.

Effects of boundary spanner position over time

In Proposition 3, we argue that a boundary spanner who is embedded within a transitive triad is ideally situated to manage knowledge exchanges while still developing specialized expertise (Lee et al., 2014). The boundary spanner’s embeddedness enables him or her to remain cognizant about who knows what and ensures that knowledge is disseminated to—and gathered from—the appropriate team expert. This results in a strong TMS, which is likely to develop over time as long as the boundary spanner remains in a transitive triad structure in the team’s knowledge exchange network.

In contrast, if a boundary spanner is not embedded in a transitive triad (i.e., is in a peripheral position), the boundary spanner must engage in dyadic exchanges with team experts to share relevant knowledge received from outside the team. Social network theory and research shows that dyadic exchanges, while creating dense networks, are less efficient for exchanging knowledge and for developing shared understandings of the knowledge exchange network than are triadic exchanges (Krackhardt & Kilduff, 2002; Lee et al., 2014; Simmel, 1950). Dyadic exchanges are more apt to lead to redundant learning efforts by members of the team and prolong information search (Burt, 2000). Over time, dyadic exchanges will tend to weaken TMS (Lee et al., 2014) and therefore, the team’s capacity for within-team coordination.

Between-team coordination will also be compromised when a boundary spanner is in a peripheral position in the team knowledge exchange network, because that position limits access to high quality and relevant knowledge within the team. Without easy access to knowledge, the boundary spanner is less able to provide other teams with the information that they require or prefer. Over time, this will diminish the strength and associated benefit of the knowledge exchange relationship between teams. Therefore, we propose

Proposition 4: When a boundary spanner is isolated from, rather than embedded in, transitive triads in the team’s knowledge exchange network, the boundary spanner’s team TMS will be weakened over time, eventually leading to a reduced capacity for within- and between-team coordination.

The model below reflects the predictions described in the above propositions (see Figure 1).

Figure 1.

Proposed conceptual model.

Temporal cycles

The above propositions describe how within- and between-team coordination are affected over time by TMS and boundary spanner characteristics. These temporal effects can evolve into virtuous or vicious cycles that perpetuate either positive or negative outcomes, respectively. A virtuous cycle that maintains or improves within- and between-team coordination over time is most likely to develop when a boundary spanner with specialized knowledge is positioned in a transitive triad in the team’s knowledge exchange network. These boundary spanner characteristics will strengthen the TMS over time (Lee et al., 2014), improving within-team coordination. Between-team coordination will also improve, as the strong TMS relieves within-team coordination burdens for the boundary spanner, leaving cognitive resources available for between team boundary spanning. With better-coordinated knowledge exchanges, both within and between teams, members will refine their understandings of member-expertise associations, because either the knowledge exchanges reinforce existing perceptions of who knows what, or they will update prior perceptions so that member-expertise associations become more accurate and more shared (Wegner, 1987). This, in turn, strengthens TMS and coordination within and between teams, perpetuating the virtuous cycle.

Proposition 5: A virtuous cycle that improves within- and between-team coordination is likely to occur when the boundary spanner possesses specialized knowledge and is embedded within a transitive triad in the team’s knowledge exchange network.

Cycles that perpetuate negative, rather than positive, effects may also result from the interplay between boundary-spanning characteristics and TMS. One such vicious cycle would be initiated when the boundary spanner possesses general, rather than specialized knowledge. When a boundary spanner possesses generalized knowledge, it weakens the team TMS—which is strongest when members’ knowledge is specialized and complementary—over the long term. As Tortoriello et al. (2012) noted, individuals that disperse information in a boundary-spanning capacity, without developing a functional skill, will ultimately be of less value to the team. Subsequently, the boundary spanner is unlikely to abandon generalized learning in favor of specialized learning, because the weaker TMS requires the boundary spanner to expend cognitive resources on coordinating knowledge exchanges both within and between teams. This perpetuates the negative effects of generalized boundary spanner knowledge on TMS and on coordination.

A vicious cycle is also likely to develop when a boundary spanner’s position in the team’s knowledge exchange network is peripheral, rather than core. With a peripheral position, outside of efficient transitive triads, the boundary spanner will have limited access to pertinent team member information, such as member expertise. This impacts the team’s ability to develop a shared understanding of who knows what and obstructs efficient allocation of information and responsibility to the appropriate member (Moreland, 1999). Over time, the boundary spanner’s isolated position in the team’s knowledge exchange network and the resulting weaker TMS would imply that any updates to member expertise within the team would not be easily accessible to the boundary spanner. Information entering the team via the boundary spanner might be rich, but the application of such information would be dampened by the lack of connectivity that the boundary spanner has with team members. When useful external knowledge is not effectively disseminated and used by the other members of the team, it will reinforce the boundary spanner’s peripheral position outside of transitive triad structures, eventually undermining the TMS over time, reducing coordination inside the team, and undermining the efficacy of boundary spanning between teams. Therefore, we propose

Proposition 6: A vicious cycle that diminishes within- and between-team coordination is likely to occur when the boundary spanner becomes less specialized over time or when the boundary spanner becomes more peripheral over time in the team’s knowledge exchange network.

Discussion and Future Research

In this article, we suggest that the TMS and boundary-spanning literatures are useful for understanding how individuals in team-based collectives can be structured to improve within- and between-team coordination. Boundary spanning research has called for more studies that explore the factors that influence the efficacy of varying team compositions, while TMS research has focused predominantly on within-team, rather than between-team coordination. We argue that by considering how TMS and boundary spanner characteristics interact, we can better understand the conditions under which coordination within and between teams can be achieved. Our theorizing applies to teams and collectives engaged in highly interdependent tasks that require a large amount of specialized knowledge (expertise) to perform well. Such teams benefit from the development, application, sharing, and integration of knowledge from different domains. In this article, we explain how teams can efficiently incorporate knowledge from outside the team via boundary spanners, how characteristics of the boundary spanner’s knowledge and position in the within-team knowledge exchange network affect both TMS and coordination, and how the interplay among these factors may unfold over time.

Although the benefits of boundary spanning have been documented, the literature does not provide a clear reckoning about the type of knowledge the boundary spanner should possess for boundary-spanning activities to be fruitful, or the ideal position of the boundary spanner for facilitating coordination within and across teams. We assert that recent findings from the TMS literature can be used to explore these gaps. Specifically, insights about the structure of knowledge exchanges that produce a TMS (Lee et al., 2014; Palazzolo, 2005) suggest that the degree to which a boundary spanner is embedded in transitive triads in the knowledge exchange network will affect the boundary spanner’s ability to coordinate inflows and outflows of knowledge between teams. We explain how an existing team TMS can offset the within-team coordination burdens typically associated with boundary spanning. When a team has a strong TMS, responsibility for coordinating knowledge exchanges within the group is spread across members, leaving the boundary spanner with additional cognitive resources to pursue specialized learning that benefits the team without damaging within- and between-team coordination.

Finally, we offer predictions about how these factors interrelate to affect TMS and coordination over time. A virtuous cycle is likely when (a) the team has a strong TMS, (b) the boundary spanner is embedded in the team’s knowledge exchange network, and (c) the boundary spanner contributes to the team TMS with specialized and complementary knowledge. These factors are likely to maintain and improve both within- and between-team coordination. Other conditions, however, are very likely to devolve into vicious cycles that reduce coordination and undermine boundary-spanning activities. When a boundary spanner’s position in the team knowledge exchange network is peripheral rather than core (embedded in transitive triad structures), or when the team’s TMS is weak, these factors will tend to exacerbate problems with both within- and between-team coordination over time.

A team may be able to take action to disrupt the vicious cycle described above. For instance, it is possible that the boundary spanner’s position could change over time, such that the resultant vicious cycle is altered. A boundary spanner with generalized expertise in a peripheral position could become more connected and embedded within the team’s knowledge exchange network—perhaps by participating in a small task with two members who are already actively exchanging knowledge (e.g., by forming a new transitive triad structure). Once embedded in a new or existing transitive triad, the boundary spanner is likely to be influenced by other members of the triad to conform (Heider, 1958; Krackhardt & Kilduff, 2002)—for example, by further developing distinctive and specialized knowledge that helps the team.

The temporal predictions imply that short-term effects of boundary spanning and TMS may be quite different from effects in the long term. We explain how short-term conditions can either reinforce positive effects in a virtuous cycle, or perpetuate negative effects that would not be evident from “snapshots” of data gathered at one point in time. For example, in the short term, a boundary spanner’s generalized knowledge could seem helpful, because broad knowledge may help the boundary spanner understand and communicate about knowledge within and between teams. In the longer term, however, generalized boundary spanner knowledge will weaken the team’s TMS. Perpetuating a negative cycle, a weaker TMS blurs any distinctive lines of expertise, encouraging members to develop redundant or overlapping knowledge because members are unaware that other members possess particular expertise or because members do not find other members’ expertise credible. As the team’s knowledge becomes less differentiated, the TMS is unlikely to recover (Lewis & Herndon, 2011).

Another example comes from our propositions regarding the position of the boundary spanner in the team’s knowledge exchange network. In the short term, it might appear that boundary spanners who are in central positions in a dense knowledge exchange network would be best able to retrieve and disseminate knowledge—indeed, the boundary-spanning literature supports this claim (i.e., Cross & Cummings, 2004; Ibarra, 1993). In the long term, however, the frequent dyadic exchanges typical of centralized and dense knowledge networks can weaken TMS (Lee et al., 2014), eventually limiting a boundary spanner’s ability to recognize, allocate, and disseminate knowledge within and between teams. As our arguments make clear, more sparse patterns of knowledge exchange that are characteristic of transitive triad structures are more likely to maintain or strengthen TMS. When embedded in transitive triads, the boundary spanner’s improved access to members’ expertise benefits within-team coordination and allows more useful information to be exchanged with the external units linked by boundary-spanning ties.

Practical Contributions

Given organizations’ ongoing dependence on teams across various settings (Allison & Shuffler, 2014; Davison et al., 2012; Hoegl, Weinkauf, & Gemuenden, 2004), our theory underscores significant implications for management practices within teams. Managers and team members can improve coordination within and between teams by increasing team member awareness of the TMS construct and reinforcing the TMS creation processes—learning about who knows what in the team, and learning about how teams can coordinate knowledge flows to collectively produce knowledge and work products.

Managers can enhance TMS development within teams by training members together (Liang et al., 1995) and by facilitating discussions about which members possess what expertise and which members are good at what activities (Moreland & Myaskovsky, 2000). We suggest ways that boundary spanners within such teams can be more effectively utilized. First, we offer recommendations for how boundary spanners can best contribute to team coordination, depending on the strength of TMS that exists in the teams. The nature of the boundary spanner’s expertise—either specialized or generalized—will affect the boundary spanner’s ability to link between teams, and this ability is affected by the maturity of TMS. This suggests that training interventions aimed at developing generalist boundary spanners may be misplaced, if the boundary spanner’s team has a strong TMS. Furthermore, we show that over time, being a generalist can actually hinder TMS development and the coordination of knowledge flow. Our theorizing suggests that boundary spanners can maintain specialized knowledge when knowledge flows within teams are already effective.

Effective team coordination is also impacted by the position of the boundary spanner. The boundary spanner’s classification as a core (embedded in a transitive triad) or peripheral member will affect the speed and quality of information that enters or exits the team, which will consequently influence the strength of the TMS that the team develops. Practically, managers can strategically develop triadic structures to maintain team coordination (Lee et al., 2014). In considering the impact of positioning over time, we assert that being a peripheral member is detrimental to TMS development, not only because boundary spanners may possess inaccurate or untimely information for external connections but also because boundary spanners must engage in dyadic exchanges for within-team knowledge sharing, which are less efficient than triadic exchanges and may ultimately weaken the team’s TMS.

Relationships are evolving all the time as the ability of team members and teams to more accurately learn who knows what is impacted by ongoing team and member adjustments. Our theorizing provides some guidance about how to manage such adjustments. For example, in the case of planned or unplanned turnover of members, managers will do well to consider the network positions of departing and replacement members. If a departing team member is already in a transitive triad, then the individuals linked to the departing team member are likely to have the most information and develop new triadic structures, enabling them to link within and between other teams, thus limiting the amount of information that is lost. In addition, a manager or a member-designee can deliberately manage knowledge sharing in and between teams by establishing him- or herself within a triad that would link to team members within the knowledge exchange network. Although transitive triads can develop naturally, managers may wish to formalize transitive triad structures to ensure efficiencies and reduce redundancies in information processing activities (Lee et al., 2014).

Limitations and Future Research Directions

The arguments we provide in this article are limited by certain assumptions. First, our theorizing assumes that a focal and target team linked by a boundary-spanning tie are mirror images of each other. That is, we do not discuss whether or not the other team’s boundary spanner knowledge, position, or TMS strength would affect processes or outcomes for the focal team (and vice versa). We note that there may be some interesting interdependencies in exploring the interplay between the focal and target teams. What happens when a weak TMS team spans to a strong TMS team, or when a specialist/core boundary spanner exchanges knowledge with a generalist/peripheral boundary spanner? One thought is that, regardless of the position (embedded or isolated) of the team’s boundary spanner, if that individual is connected to an isolated member on another team, the information obtained by the focal team could be compromised by the external isolated member, which will negatively impact the TMS that develops. An example of this might be two teams within a product development division—Marketing and Engineering—that interact with each other to develop and launch a smart device. Suppose the Marketing team obtains customer feedback about a design flaw in the product that must be addressed by Engineering in the next product release. If that information is passed to a boundary spanner who is isolated from, rather than embedded within the Engineering knowledge exchange network, that information is less likely to be incorporated by the Engineering team. This could be an interesting avenue to explore further.

A second limitation is that we do not consider complex team arrangements that one might observe in so-called “teams of teams,” or multi-team systems (MTS; Marks, DeChurch, Mathieu, Panzer, & Alonso, 2005). As has been noted in the MTS literature, inter-team coordination in these complex arrangements can be affected by a host of factors, including goal incompatibilities, variations in interdependence across different teams, boundary spanning that extends outside of the focal organization, and situations in which individuals work in multiple teams simultaneously (O’Leary, Mortensen, & Woolley, 2011). We note, however, that our more narrow focus does have implications for the MTS and organizational TMS literatures. First, our theorizing highlights the importance of transitive triadic network structures (as opposed to network density or centrality) in promoting coordination within and between teams. Second, we leverage TMS research to explain circumstances under which the cognitive load that burdens boundary spanners might be lightened, so that a boundary spanner can better contribute to task-relevant team work. Our chosen focus on the individual level of analysis and the micro-processes involved in the development and operation of a team-level TMS revealed new insights about individual-level and team-level processes that the MTS and organizational TMS literature does not currently address.

This article sought to utilize the TMS and boundary-spanning literatures to conceptualize how boundary-spanning knowledge and position can affect within- and between-team coordination. Although our arguments are founded in strong theory and are supported by some empirical research, our predictions must be considered tentative until empirical research can be conducted. Future research can empirically test these propositions, and also draw comparisons with other studies that incorporate coordination strategies supporting TMS development both within and between teams. For example, such studies could conduct simulations to clarify boundary conditions and provide more deterministic insights on conflicting arguments present in the literature, such as the ideal number of boundary spanners that a team should have. We acknowledge the challenges of coordinating activities and knowledge flows that organizations are increasingly facing, and we hope that our approach to addressing these challenges through a reconciliation of insights from recent TMS and boundary literature encourages future research that extends this notable contribution.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Associate Editor: Lucy Gilson

Notes

Author Biographies

Joy Olabisi, PhD, is an associate professor within the Management Department at the Saunders College of Business, Rochester Institute of Technology. She earned her PhD in Engineering Management from the Industrial & Operations Engineering Department at the University of Michigan. Her current research interests include understanding how organizations learn and manage knowledge, especially through the effective utilization of teams. She is also interested in social entrepreneurial initiatives within emerging markets, with a social embeddedness and networks emphasis. She has published articles in outlets such as Group & Organization Management, Journal of Management Studies, and Journal of Business Ethics.

Kyle Lewis, PhD, is the Chair and professor of Technology Management Program at University of California, Santa Barbara. She holds a PhD in Organizational Behavior from the R. H. Smith School of Business at the University of Maryland. Her research examines how organizations leverage individual and collective knowledge. She studies the performance of teams, especially those teams engaged in knowledge work such as professional services, new product development, and project-based tasks. She has published articles in Management Science, Academy of Management Review, Academy of Management Journal, Organization Science, Journal of Management, and Journal of Applied Psychology.

References

Akgün

A. E.

Byrne

Keskin

Lynn

G. S.

Imamoglu

S. Z.

(2005). Knowledge networks in new product development projects: A transactive memory perspective. Information & Management, 42, 1105-1120.

Allison

B. B.

Shuffler

M. L.

(2014). Getting the “I” out of multiteam systems: A case study from the financial services industry. In Shuffler

B. B

Rico

Salas

(Eds.), Pushing the boundaries: Multiteam systems in research and practice (pp. 185-203). Bingley, UK: Emerald Group Publishing Limited.

Anand

Manz

C. C.

Glick

W. H.

(1998). An organizational memory approach to information management. Academy of Management Review, 23, 796-809.

Ancona

D. G.

Caldwell

(1990). Beyond boundary spanning: Managing external dependence in product development teams. The Journal of High Technology Management Research, 1(2), 119-135.

Ancona

D. G.

Caldwell

D. F.

(1988). Beyond task and maintenance: Defining external functions in groups. Group & Organization Studies, 13, 468-494.

Ancona

D. G.

Caldwell

D. F.

(1992). Bridging the boundary: External activity and performance and absorptive capacity on business unit innovation and performance. Academy of Management Journal, 44, 996-1004.

Anderson

E. G.

Jr. Lewis

(2014). A dynamic model of individual and collective learning amid disruption. Organization Science, 25, 356-376.

Austin

J. R.

(2003). Transactive memory in organizational groups: The effects of content, consensus, specialization, and accuracy on group performance. Journal of Applied Psychology, 88, 866-878.

Brandon

D. P.

Hollingshead

A. B.

(2004). Transactive memory systems in organizations: Matching tasks, expertise, and people. Organization Science, 15, 633-644.

10.

Burt

R. S.

(1992). Structural holes: The social structure of competition. Cambridge, MA: Harvard University Press.

11.

Burt

R. S.

(2000). The network structure of social capital. Research in Organizational Behavior, 22, 345-423.

12.

Cross

Cummings

J. N.

(2004). Tie and network correlates of individual performance in knowledge-intensive work. Academy of Management Journal, 47, 928-937.

13.

Davison

R. B.

Hollenbeck

J. R.

(2012). Boundary spanning in the domain of multiteam systems. In Zaccaro

S. J.

Marks

M. A.

DeChurch

L. A.

(Eds.), Multiteam systems: An organizational form for dynamic and complex environments (pp. 323-362). New York, NY: Taylor & Francis.

14.

Davison

R. B.

Hollenbeck

J. R.

Barnes

C. M.

Sleesman

D. J.

Ilgen

D. R.

(2012). Coordinated action in multiteam systems. Journal of Applied Psychology, 97, 808-824.

15.

DeChurch

L. A.

Mathieu

J. E.

(2009). Thinking in terms of multiteam systems. In Salas

Goodwin

Burke

C. S.

(Eds.), Team effectiveness in complex organizations: Cross-disciplinary perspectives and approaches (pp. 267-292). New York, NY: Routledge/Taylor & Francis Group.

16.

Dokko

Kane

A. A.

Tortoriello

(2014). One of us or one of my friends: How social identity and tie strength shape the creative generativity of boundary-spanning ties. Organization Studies, 35, 703-726.

17.

Edmondson

A. C.

Nembhard

I. M.

(2009). Product development and learning in project teams: The challenges are the benefits. Journal of Product Innovation Management, 26, 123-138.

18.

Faraj

Sproull

(2000). Coordinating expertise in software development teams. Management Science, 46, 1554-1568.

19.

Faraj

Yan

(2009). Boundary work in knowledge teams. Journal of Applied Psychology, 94, 604-617.

20.

Gibson

C. B.

Dibble

(2013). Excess may do harm: Investigating the effect of team external environment on external activities in teams. Organization Science, 24, 697-715.

21.

Goh

K. T.

Krackhardt

Weingart

L. R.

Koh

T. K.

(2014). The role of Simmelian friendship ties on retaliation within triads. Small Group Research, 45, 471-505.

22.

Hansen

M. T.

(1999). The search-transfer problem: The role of weak ties in sharing knowledge across organization subunits. Administrative Science Quarterly, 44(1), 82-111.

23.

Hansen

M. T.

(2002). Knowledge networks: Explaining effective knowledge sharing in multiunit companies. Organization Science, 13, 232-248.

24.

Heider

(1946). Attitudes and cognitive organization. The Journal of Psychology, 21, 107-112.

25.

Heider

(1958). The psychology of interpersonal relations. Hillsdale, NJ: Lawrence Erlbaum.

26.

Hoegl

Weinkauf

Gemuenden

H. G.

(2004). Interteam coordination, project commitment, and teamwork in multiteam R&D projects: A longitudinal study. Organization Science, 15(1), 38-55.

27.

Ibarra

(1993). Network centrality, power, and innovation involvement: Determinants of technical and administrative roles. Academy of Management Journal, 36, 471-501.

28.

Jackson

Klobas

(2008). Transactive memory systems in organizations: Implications for knowledge directories. Decision Support Systems, 44, 409-424.

29.

Joshi

Pandey

Han

G. H.

(2009). Bracketing team boundary spanning: An examination of task-based, team-level, and contextual antecedents. Journal of Organizational Behavior, 30, 731-759.

30.

Krackhardt

(1998). Simmelian ties: Super, strong, and sticky. In Kramer

Neale

(Eds.), Power and influence in organizations (pp. 21-38). Thousand Oaks, CA: Sage.

31.

Krackhardt

Kilduff

(2002). Structure, culture and Simmelian ties in entrepreneurial firms. Social Networks, 24, 279-290.

32.

Lee

J. Y.

Bachrach

D. G.

Lewis

(2014). Social network ties, transactive memory, and performance in groups. Organization Science, 25, 951-967.

33.

Lewis

(2003). Measuring transactive memory systems in the field: scale development and validation. Journal of Applied Psychology, 88, 587-604.

34.

Lewis

(2004). Knowledge and performance in knowledge-worker teams: A longitudinal study of transactive memory systems. Management Science, 50, 1519-1533.

35.

Lewis

Belliveau

Herndon

Keller

(2007). Group cognition, membership change, and performance: Investigating the benefits and detriments of collective knowledge. Organizational Behavior and Human Decision Processes, 103, 159-178.

36.

Lewis

Herndon

(2011). Transactive memory systems: Current issues and future research directions. Organization Science, 22, 1254-1265.

37.

Lewis

Lange

Gillis

(2005). Transactive memory systems, learning, and learning transfer. Organization Science, 16, 581-598.

38.

Liang

D. W.

Moreland

Argote

(1995). Group versus individual training and group performance: The mediating role of transactive memory. Personality and Social Psychology Bulletin, 21, 384-393.

39.

Majchrzak

Jarvenpaa

S. L.

Hollingshead

A. B.

(2007). Coordinating expertise among emergent groups responding to disasters. Organization Science, 18(1), 147-161.

40.

Marks

M. A.

DeChurch

L. A.

Mathieu

J. E.

Panzer

F. J.

Alonso

(2005). Teamwork in multiteam systems. Journal of Applied Psychology, 90, 964-971.

41.

Marrone

J. A.

(2010). Team boundary spanning: A multilevel review of past research and proposals for the future. Journal of Management, 36, 911-940.

42.

Marrone

J. A.

Tesluk

P. E.

Carson

J. B.

(2007). A multilevel investigation of antecedents and consequences of team member boundary-spanning behavior. Academy of Management Journal, 50, 1423-1439.

43.

Michinov

Olivier-Chiron

Rusch

Chiron

(2008). Influence of transactive memory on perceived performance, job satisfaction and identification in anaesthesia teams. British Journal of Anaesthesia, 100, 327-332.

44.

Moreland

R. L.

(1999). Transactive memory: Learning who knows what in work groups and organizations. In Thompson

L. L.

Levine

J. M.

Messick

D. M.

(Eds.), Shared cognition in organizations: The management of knowledge (pp. 3-31). Mahwah, NJ: Lawrence Erlbaum.

45.

Moreland

R. L.

Argote

(2003). Transactive memory in dynamic organizations. In Peterson

R. S.

Mannix

(Eds.), Leading and managing people in the dynamic organization (pp. 135-162). Mahwah, NJ: Lawrence Erlbaum.

46.

Moreland

R. L.

Myaskovsky

(2000). Exploring the performance benefits of group training: Transactive memory or improved communication? Organizational Behavior and Human Decision Processes, 82, 117-133.

47.

Nevo

Wand

(2005). Organizational memory information systems: A transactive memory approach. Decision Support Systems, 39, 549-562.

48.

O’Leary

M. B.

Mortensen

Woolley

A. W.

(2011). Multiple team membership: A theoretical model of its effects on productivity and learning for individuals and teams. Academy of Management Review, 36, 461-478.

49.

Palazzolo

E. T.

(2005). Organizing for information retrieval in transactive memory systems. Communication Research, 32, 726-761.

50.

Peltokorpi

(2008). Transactive memory systems. Review of General Psychology, 12, 378-394.

51.

Peltokorpi

(2012). Organizational transactive memory systems. European Psychologist, 17, 11-20.

52.

Peltokorpi

(2014). Transactive memory system coordination mechanisms in organizations: An exploratory case study. Group & Organization Management, 39, 444-471.

53.

Peltokorpi

Manka

M. L.

(2008). Antecedents and the performance outcome of transactive memory in daycare work groups. European Psychologist, 13, 103-113.

54.

Rau

(2006). Top management team transactive memory, information gathering, and perceptual accuracy. Journal of Business Research, 59, 416-424.

55.

Ren

Argote

(2011). Transactive memory systems 1985–2010: An integrative framework of key dimensions, antecedents, and consequences. Academy of Management Annals, 5, 189-229.

56.

Simmel

(1950). The sociology of Georg Simmel. New York, NY: Free Press.

57.

Smith-Jentsch

K. A.

Kraiger

Cannon-Bowers

J. A.

Salas

(2009). Do familiar teammates request and accept more backup? Transactive memory in air traffic control. Human Factors: The Journal of the Human Factors and Ergonomics Society, 51, 181-192.

58.

Somech

Khalaili

(2014). Team boundary activity: Its mediating role in the relationship between structural conditions and team innovation. Group & Organization Management, 39, 274-299.

59.

Sparrowe

R. T.

Liden

R. C.

Wayne

S. J.

Kraimer

M. L.

(2001). Social networks and the performance of individuals and groups. Academy of Management Journal, 44, 316-325.

60.

Tortoriello

Krackhardt

(2010). Activating cross-boundary knowledge: The role of Simmelian ties in the generation of innovations. Academy of Management Journal, 53, 167-181.

61.

Tortoriello

Reagans

McEvily

(2012). Bridging the knowledge gap: The influence of strong ties, network cohesion, and network range on the transfer of knowledge between organizational units. Organization Science, 23, 1024-1039.

62.

Tsai

(2001). Knowledge transfer in intraorganizational networks: Effects of network position and absorptive capacity on business unit innovation and performance. Academy of Management Journal, 44, 996-1004.

63.

Tushman

M. L.

Scanlan

T. J.

(1981). Boundary spanning individuals: Their role in information transfer and their antecedents. Academy of Management Journal, 24, 289-305.

64.

Wasserman

Faust

(1994). Social network analysis: Methods and applications (Vol. 8). Cambridge, UK: Cambridge University Press.

65.

Wegner

D. M.

(1987). Transactive memory: A contemporary analysis of the group mind. In Mullen

Goethals

G. R.

(Eds.), Theories of group behavior (pp. 185-208). New York, NY: Springer.

66.

Wegner

D. M.

(1995). A computer network model of human transactive memory. Social Cognition, 13, 319-339.