Transactive Memory System Coordination Mechanisms in Organizations

Abstract

This article integrates the organizational coordination and transactive memory system (TMS) literatures and provides an empirical study on organizational TMS coordination mechanisms. The study findings, based on 60 interviews in a Japanese manufacturing company, show that organization design (team-based structure, small unit size), human resource management (HRM) practices (recruitment and selection, training, promotion, and reward systems and performance evaluations), and relational interactions (roles, routines) coordinate organizational TMSs. Organization design and HRM practices as more formal coordination mechanisms also support and provide continuity to relational interactions. Contributing to the organizational TMS literature in which employees are described to form TMSs through informal, face-to-face interactions, the findings provide evidence of formal and informal coordination mechanisms and suggest that team-based structures without reinforcing vertical mechanisms are insufficient to coordinate organizational TMSs.

Keywords

coordination human resource management team-based structure transactive memory systems

Researchers have increasingly used transactive memory system (TMS) theory (Wegner, 1986) to examine cognitive processes and outcomes in knowledge-intensive teams (Peltokorpi, 2008). A TMS refers to a shared cognitive system that team members collectively develop to encode, store, and retrieve information in different domains. TMSs enhance task coordination and performance in teams with complex, interdependent tasks, because their members are aware who knows what, are able to specialize in different but compatible information domains, and use each other as external cognitive aids (Lewis, 2003).

TMSs have been extended from teams to organizations as cognitive systems in which distributed information is processed collectively (Anand, Manz, & Glick, 1998; Brauner & Becker, 2006; Jackson, 2012; Jackson & Klobas, 2008; Moreland, 1999; Moreland & Argote, 2003; Peltokorpi, 2004, 2012). Although the organizational TMS literature helps to understand how employees encode, store, and retrieve distributed information in organizations, more conceptual and empirical works have been called for due to simplistic extensions of team-level tenets to organizations (Peltokorpi, 2012). In particular, the organizational TMS literature has paid little attention to coordination—the integration of organizational work under conditions of task complexity and interdependence (Faraj & Xiao, 2006). At best, scholars have described organizational TMSs as informal coordination mechanisms that provide a shared awareness of organizational expertise (Brauner & Becker, 2006; Jackson, 2012). Among other things, these scholars have not taken account that the larger size of social entities increases the reliance on formal coordination mechanisms (Van de Ven, Delbecq, & Koenig, 1976). Because of the importance of both coordination (Okhuysen & Bechky, 2009) and TMS (Ren & Argote, 2011) on organizational effectiveness, a specific focus on organizational TMS coordination mechanisms is warranted.

In this article, I contribute to the TMS literature in two ways. First, I integrate the organizational coordination and TMS literatures to take into account both formal and informal coordination mechanisms as well as a potential trade-off between them. Formal coordination mechanisms (e.g., plans) are pre-designed ways to manage work processes. Informal coordination mechanisms (e.g., mutual adjustments) are lateral interactions through which people manage their own work processes. Drawing on Faraj and Xiao (2006), I define organizational TMS coordination as the integration of organizational TMS processes under conditions of task complexity and interdependence. While high task complexity and interdependence in TMSs suggest a shift from formal toward more informal coordination mechanisms, the increased size of organizations suggests the opposite relationship. Coordination of TMSs solely through informal mechanisms can thus be insufficient in large social entities due to the associated coordination costs (Van de Ven et al., 1976). Formal mechanisms alone might also be insufficient to coordinate complex, interdependent work processes (Faraj & Xiao, 2006). Therefore, I focus on both formal and informal organizational TMS coordination mechanisms.

Second, I provide an empirical study on organizational TMS coordination. Despite the increasing scholarly interest and calls for empirical research (Moreland & Argote, 2003; Ren & Argote, 2011), the organizational TMS literature has remained conceptual. To date, the only empirical study on organizational TMS concludes that “the basic TMS processes and directories [individual perceptions of what others know] appear to be present at an organizational level as well” (Jackson & Klobas, 2008, p. 409). Taking a more critical approach, I explore to what extent TMS tenets derived from teams can be extended to organizations and how organizational TMSs are coordinated. In line with this approach, I define organizational TMSs as “networks of interdependent teams that use each other as external cognitive aids to accomplish shared tasks” (Peltokorpi, 2012, p. 17).

Literature Review

The organizational TMS literature draws on team-level tenets. Thus, I start with a description of team TMS formation and functioning through the encoding, storage, and retrieval phases (Wegner, 1986). During the encoding phase, team members learn who knows what by making, refining, and updating inferences about each other’s expertise domain(s). TMSs can be formed in part because each team member accepts responsibility for certain task expertise domain(s). Once each team member accepts the responsibility as domain expert, other members are able to pass information to them for processing and storage. Teams with complex, interdependent tasks form differentiated TMSs in which each member holds different task information items. This specialization reduces knowledge overlaps, allowing teams to possess a greater amount of task-related information. In efficient differentiated TMSs, specialized task information is credible and well coordinated. Specialization is the level of memory differentiation within a team, coordination is the ability of team members to work together efficiently, and credibility suggests that members trust the designated expert’s task information (Lewis, 2003). Transactive retrieval occurs when team members retrieve task information from domain expert(s). Team members are able to retrieve task information by identifying a relevant domain expert through internally/externally held location information about who knows what. TMS processes facilitate expert coordination in teams (Faraj & Sproull, 2000). For example, by knowing who knows what, team members can find information more easily and assign tasks to those best able to carry them out.

While considering most, if not all, team TMS attributes to be present at the organizational level, scholars have not discussed in detail organizational TMS encoding, storage, and retrieval processes. Instead, following Moreland (1999), scholars often use technological and interpersonal approaches to describe the formation and functioning of organizational TMSs (Peltokorpi, 2012). Through the information and communication technology (ICT)–based technological approach, information is encoded, stored, and retrieved through direct (e.g., e-mail) and indirect interactions (e.g., shared databases). Direct interactions through e-mail allow TMS formation without face-to-face interactions (Moreland, 1999). In indirect interactions, individuals provide their expertise areas and information to communal repositories (e.g., shared databases) and others then access these repositories through ICT (Moreland, 1999). The interpersonal approach facilitates organizational TMSs through social networks that are used to search, locate, and retrieve information. In the search processes, individuals are able to rely on direct-network links (knowing what other people know) and indirect-network links (knowing who other people know). Organizational and work-related aspects that also enhance organizational TMSs include boundary spanners (individuals who form linkages and manage interactions with external parties), team-based organizational structures, (Anand et al., 1998), meetings, roles (the set of expectations and behaviors associated with a particular position), training, job rotation, and newcomer socialization (Moreland, 1999).

Scholars have further differentiated and described interactions between team and organizational TMSs. Moreland (1999) proposed that organizations composed of several subgroups are less cohesive than small teams, decreasing employee motivation to contribute to organizational TMSs. People higher in the organizational hierarchy are also noted to have a more accurate awareness of and better access to distributed expertise (Peltokorpi, 2004). Hierarchical positions and formal task roles also make people sensitive about what kind of information they share with whom (Moreland, 1999). In terms of team and organizational TMS interactions, simultaneous participation of employees in several teams may blur the distinction between team and organizational TMSs (Anand et al., 1998). Job rotation can also have differentiated effects on team and organizational TMSs. While frequent team membership changes enhance organizational TMSs in terms of greater organization-wide network linkages (Moreland & Argote, 2003), they can weaken team TMSs because of incorrect or incomplete expertise awareness (Lewis, Belliveau, Herndon, & Keller, 2007) and the risk of relying on domain expert(s) who are not available in a continuing basis (Moreland & Argote, 2003).

Scholars, however, have overlooked organizational TMS coordination. At best, organizational TMSs have been described to act as informal coordination mechanisms that provide a shared awareness of organizational expertise. For example, Brauner and Becker (2006) proposed, “[M]utual metaknowledge [of who knows what] facilitates transactive processes and the mutual access to each other’s knowledge” (p. 66). Jackson (2012), in turn, argued, “The goal of a TMS approach is to create . . . a directory structure which provides a unified mental model of the firm” (p. 119). These examples suggest that a shared awareness of organizational expertise, which is a part of a TMS, acts as a coordination mechanism. Yet, in terms of the mechanisms that coordinate organizational TMSs, the literature offers little more than informal interpersonal and ICT-mediated interactions. While informal interactions coordinate TMSs in small teams (Lewis, 2003), organizational coordination theories maintain that informal coordination mechanisms in larger social entities need to be accompanied with formal coordination mechanisms.

Organizational Coordination

Since March and Simon’s (1958) description of organizational coordination through formal controls and informal relations, coordination theories have evolved and taken many different forms. Until the 1980s, coordination in the dominant information processing and contingency-based theories revolved around the relationship between various coordination drivers (in particular, task complexity and interdependence) and matching coordination mechanisms. Common in related coordination typologies is a distinction between formal and informal coordination mechanisms. Formal coordination mechanisms include plans, rules (Thompson, 1967), and vertical relations (Galbraith, 1973). Informal coordination mechanisms include mutual adjustments (Thompson, 1967), lateral relations (Galbraith, 1973), and team meetings (Van de Ven et al., 1976). For example, team meetings have high information-processing capability and facilitate interactions among participants in a work process. In contrast, routines have low information-processing capacity and thus reduce the need for interaction among participants. Information processing and contingency-based theories suggest that increased task complexity and interdependence are accompanied by a shift from formal toward more informal coordination mechanisms as well as from hierarchical toward more team-based organizational structures (Galbraith, 1973). In contrast, increased size of work units suggests a shift toward more formal mechanisms and hierarchical structures (Thompson, 1967).

Being less concerned with aligning coordination mechanisms and drivers, scholars have more recently described organizational coordination as informal processes that occur through a network of relationships among employees who perform interdependent tasks. For example, Gittell (2000) defined coordination as “a network of communication and relationship ties among workers” (p. 518). This more spontaneous form of coordination refers to relational coordination (Gittell, 2000). In the relational coordination literature, coordination is carried out through emergent relationships of shared goals, knowledge, and mutual respect. Strong relationships enable employees to embrace their connections with one another and to more effectively coordinate the work in which they are engaged. Examples of relational coordination are boundary spanners (Gittell, 2000), decentralized decision making, self-managed teams, training (Evans & Davis, 2005), roles (Bechky, 2006), and communities of practice (a group of people who share a craft or profession; Faraj & Xiao, 2006). Lateral informal interactions are common in these coordination typologies (Okhuysen & Bechky, 2009).

Organizational TMS Coordination

While informal interactions coordinate TMSs in small teams (Lewis, 2003) and knowledge-intensive work in small organizational units (Faraj & Xiao, 2006), I draw on the related literature to argue that TMS coordination only through informal interactions is not sufficient in large social entities. For example, the law of N-squared states that the number of potential links in a network increases geometrically with the number of people (Krackhardt, 1994). The number of potential links grows so fast that the number of people to which each person could be linked quickly exceeds everyone’s cognitive and communication capacity. In organizations, physical distance also reduces chances for spontaneous, informal interactions (Cross & Sproull, 2004) through which TMSs are coordinated (Wegner, 1986). While facilitating interactions through physical distance, employees are shown to rely more often on people than on ICT to search and retrieve information (Cross & Sproull, 2004). While a computation model further shows that members of larger teams have greater difficulty forming a shared awareness of distributed expertise (Ren, Carley, & Argote, 2006), organizational TMSs appear to be coordinated through informal, face-to-face interactions.

How are organizational TMSs coordinated? The integration of the organizational coordination and TMS literatures suggests that both formal and informal mechanisms coordinate organizational TMSs. For example, formal routines (as recurring patterns of activity among interdependent organizational members) coordinate work by providing a shared template for task completion, bringing employees together and creating a common perspective across organizational units (March & Simon, 1958). Reward interdependence (the extent to which the rewards that accrue to an individual depend on the performance of coworkers) is further argued to motivate employees to share knowledge (Evans & Davis, 2005) and to contribute to organizational TMSs (Peltokorpi, 2012). A study also suggests that formal modes of coordination do not always melt away in favor of more informal modes of coordination (Faraj & Xiao, 2006). Informal coordination mechanisms can be equally important and interact with formal coordination mechanisms because they facilitate (a) information encoding and (b) credibility and trust in information retrieved from domain experts (Faraj & Sproull, 2000).

Summary and Research Questions

While organizations are essentially systems to coordinate work characterized by task complexity and interdependence (March & Simon, 1958; Thompson, 1967), little is yet known about organizational TMS coordination. Taking account that task expertise in TMSs needs to be coordinated (Wegner, 1986), this study focuses on this important but overlooked area in the organizational TMS literature. Specifically, this study seeks answers to two questions: (a) Which mechanisms coordinate organizational TMSs? (b) How do formal and informal coordination mechanisms interact with each other in organizational TMSs?

Method

To answer the research questions, I adopted a qualitative case study method that can be seen as an appropriate approach given the need to develop in-depth understanding of a relatively unexplored area (Yin, 2014). Indeed, to the best of my knowledge, this is the first empirical study on organizational TMS coordination. Case studies are also well suited to create theoretical constructs, propositions, and/or midrange theory (Eisenhardt & Graebner, 2007), and for “how” questions that seek to solve complex interdependencies (Yin, 2014). Some scholars view qualitative case study research as highly descriptive and stress the social construction of reality (e.g., Gephart, 2004). This view is different from the more objective and positivist approach taken by Eisenhardt (1989) and Yin (2014), who are also among the main authorities in management research (Amis & Silk, 2008). In this study, I gear toward the more positivist approach.

Case Company

I conducted the study in Mayekawa Manufacturing Ltd. (hereinafter Mayekawa). Founded in 1924 in Tokyo, Mayekawa has expanded from industrial refrigerators to food processing and extremely low temperature environment systems. Industrial refrigerator and food processing machine sales presented one third of gross income in 2006 and Mayekawa’s share of the world industrial refrigerator market was 30%. The remaining income came from services, such as maintenance. In 2006, Mayekawa employed 3,000 people (2,250 in Japan and 750 overseas) and had offices in 28 countries. All product development is conducted in Japan.

I considered Mayekawa as a suitable case company for this study for two reasons. First, organizational TMSs are identified to exist in knowledge-intensive companies in which interdependent teams collaborate to achieve shared goals (Anand et al., 1998). Mayekawa meets these requirements because tasks conducted in interrelated organizational units and projects are complex and require employees to apply and share their expertise as well as to rely on others’ expertise (Maekawa, 2008). Mayekawa can also be classified as a knowledge-intensive company as it has invented, for example, the world’s first semi-hermetic screw compressor for ammonia refrigerant that achieves 20% energy efficiency in comparison with conventional units (Maekawa, 2008).

Second, Mayekawa has a “team-based” organization design in which financially independent small companies coordinate their activities. In Japan, Mayekawa operations consist of more than 100 small companies—Doppos—designed to meet the needs of a single customer or region (see Figure 1). The word Doppo is an acronym for “dokuritsu houjin” (独立法人 or independent legal entity). Doppos belong to functional (e.g., food and meat) or regional (e.g., Hokkaido) Blocks. Assisting and coordinating interactions among Doppos, Blocks act as the link between Doppos and Zensha (全社 or the whole company). Each Block has its own research unit. Blocks do not represent a higher level of management as Doppos organize Block activities (Maekawa, 2008). As the counsel of the whole company, Zensha decides on resource flows and tackles problems too big for Blocks and Doppos to deal with alone.

Figure 1.

Mayekawa organization structure.

Data Collection

The data were collected mainly through interviews. Following Eisenhardt and Graebner (2007), the interviews were conducted with highly knowledgeable informants. Because individuals higher in the organizational hierarchy have a more comprehensive awareness of organizational TMS processes (Peltokorpi, 2004), one collaborator (a former Mayekawa employee) and I conducted individual interviews with 40 Doppo leaders and 20 Block leaders between 2002 and 2009 in Japan. The collaborator, who provided access to the company, identified and contacted these leaders, who all agreed to participate in this study. I was also introduced to the company by Professor Ikujiro Nonaka. The leaders were from the technical research center (56%), production (35%), and other functions (9%). They had worked on the average for 10 years in Mayekawa. We (the author and collaborator) also took field notes, made two one-day factory visits, observed three meetings (2 hr each), and gathered company documents for data triangulation. The interviews, factory visits, and meeting observations were conducted on separate trips.

We conducted all interviews in person with individual leaders in the workplace area where interviewees could not be overheard. We assured all interviewees full confidentiality. As interviewees may feel alienated and present themselves in a formal manner if unfamiliar language and academic terms are used (Kvale & Brinkman, 2009), we started all interviews by collecting information about the interviewees and continued by asking them questions about TMS-related coordination mechanisms and processes in Mayekawa operations in Japan (see the appendix for guiding interview questions). All interviews were semi-structured to allow the leaders to discuss and elaborate on issues they considered relevant. This flexibility was important to gain information about new, unexplored phenomena (Kvale & Brinkman, 2009). All interviews, ranging from 30 to 120 min, were conducted in Japanese and were digitally recorded and transcribed verbatim.

Data Analysis

I used data analysis computer software NVivo 8 to analyze the interview data through template analysis (King, 2004). In template analysis, researcher(s) produce a list of codes (template) representing themes identified in their textual data. A key feature of template analysis is the hierarchical organization of codes, with groups of similar codes clustered together to produce more general dominant codes. Some of these codes are often defined a priori, but they can be modified and added as researchers interpret the interview data. Template analysis occupies a position between content analysis, where codes are all predetermined, and grounded theory, where there is no a priori definition of codes (King, 2004).

I started the data analysis by coding all interviews into one broad category that covered all TMS and coordination-related topics in order not to lose any relevant information. I then coded the broad category into subcategories. Through this process, involving iteration between the data and relevant literature, I identified three dominant codes based on number of interviewees’ comments: (a) organization design, (b) human resource management (HRM) practices, and (c) relational interactions. In line with the interview evidence and design perspectives of organizational coordination (Galbraith, 1973; Thompson, 1967), I identified and coded organizational design as a coordination mechanism. A Doppo leader’s comment provides an example of this coding category: “The Doppo system helps us to combine expertise.” I further created a coding category—HRM practices—based on the interview evidence. A Block leader’s comment provides an example of this coding category: “During OJT (on-the-job training) new employees learn by watching how experienced employees work.” I also created a coding category—relational interactions—because the interviews provide evidence of informal TMS coordination mechanisms that are more consistent with the relational coordination literature than the information and contingency theory related typologies. A Doppo leader’s comment provides an example of this coding category: “Project managers act as informal intermediaries between Blocks.”

Following template analysis procedures (King, 2004), I continued by identifying and coding more detailed topics under organization design, HRM practices, and relational interactions. These topics and the number of leaders mentioning these topics during the interviews were as follows: organization design (team-based structure [58 leaders], small unit size [56 leaders]), HRM practices (recruitment and selection [35 leaders], training [32 leaders], promotion [39 leaders], reward systems and performance evaluations [51 leaders]), and relational interactions (roles [49 leaders], routines [52 leaders]). Except for the roles and routines, the meanings of these topics were relatively straightforward. Roles refer to the set of expectations and behaviors associated with a particular position (Moreland, 1999). Routines are recurring patterns of activity among interdependent organizational members (March & Simon, 1958). In line with King (2004), I stopped the process after agreeing on all codes and their content with the collaborator.

I made special efforts to enhance the quality of the data analysis. In line with Yin (2014), I used the additional information (company documents, field notes, etc.) gathered to triangulate the interview findings. For example, I manually cross-checked the interview findings with the additional information but did not find significant contradictions. Although coding was based on number of interviewees’ comments, I also report minority views in the “Results” section. Although I cannot reveal leaders’ identity, I distinguish them and their interview excerpts in this article by numbers.

Results

In this section, I elaborate on my findings of organizational TMS coordination by organization design (team-based structure, small unit size), HRM practices (recruitment and selection, training, promotion, and reward systems and performance evaluations), and relational interactions (roles, routines). Table 1 provides examples of the findings.

Table 1.

Organizational TMS Coordination Mechanisms.

Higher order codes	Lower order codes	Examples (quotes)
Organization design	Team-based structure	“There are about 100 Doppos, sorted to several groups called Blocks. I believe that collaboration and specialization in the Doppo system provides a fundamental basis for people to develop their expertise.” (Doppo leader #4) “In contrast to many other [Japanese] companies, Mayekawa encourages lateral interactions. There are active interactions going on within Blocks, such as human networks, planning, processing, discussions, and so on.” (Block leader #5) “The good thing about the Doppo system is that we can conduct business based on our judgment and staff projects with employees with the needed skills and knowledge.” (Doppo leader #19) “In each Block there are dynamic human networks that are used to identify relevant experts and use their expertise in various projects.” (Doppo leader #16)
Organization design	Small unit size	“What small units can do is limited. That is why employees have to share information and work together . . . Employees in my Doppo have specialized skills and knowledge. Other Doppos contact us when they need our support.” (Doppo leader #13) “When a large project comes to us, several Doppos and Blocks get together to make decisions and collaborate . . . This way we create understanding of specific skills and create networks.” (Doppo leader #1) “Each Doppo unit has its own accounts and is treated as an independent company within the Mayekawa corporation . . . a good point of the Doppo system is that people have control on their jobs.” (Doppo leader #17) “Naturally, because of small size, Doppos have to interact with each other. They need to collaborate and learn what other Doppos know.” (Block leader #7)
HRM practices	Recruitment and selection	“When we start to work in other markets, such as construction, we hire mid-career employees with rich experience outside Mayekawa . . . mid-career recruits are often high level engineers.” (Doppo leader #16) “Entry-level recruiting is preferred over mid-career recruiting in part because of the importance give to collaboration and networks at Mayekawa.” (Doppo leader #19) “In new technology areas where Mayekawa doesn’t have experience, we hire mid-career employees with rich experience . . . these employees are hired mainly due to their expertise.” (Block leader #17) “Occasionally we hire appropriate people from outside to become more familiar with new technology and markets.” (Block leader #9)
	Training	“We develop our engineers’ expertise by making them work and learn with each other. The Doppo structure facilitates this learning process.” (Block leader #3)“Employees learn and develop their expertise mainly through experience and advice from more experienced senior employees . . . this way they interact with experienced employees and start to understand project demands, problems, and things like that.” (Doppo leader #18)“Mayekawa training provides young employees [entry-level recruits] a broad understanding of company operations and customer needs. This understanding is gained mainly through learning-by-doing.” (Block leader #3)“New employees are given various kinds of challenging tasks and experiences that help them to extend their knowledge, skills and abilities. This way they became aware of their skills and start to understand how to develop their skills.” (Doppo leader #20)
	Promotion	“Only those who have long experience at the company are chosen as leaders . . . this is important because they have wide collaborative networks.” (Block leader #1)“When some member grows in his or her ability, the leader may leave to give the position to him or her. Old members develop young members to pass their work and positions to them. This practice enhances reciprocal relations and collaborative networks.” (Block leader #6)“Leaders are those who have gradually grown in the workplace. Only work experience enables them to be effective leaders. Through experiences these leaders have built trustful relations with various units that enable them to start new projects.” (Doppo leader #3)“Two important promotion criteria for Doppo leaders are their ability to motivate employees to rely on each other and to work together efficiently.” (Block leader #9).
	Reward systems and performance evaluations	“When employees move to other Doppos, their compensation stays the same . . . due to the standardized compensation system employees are willing to move and collaborate.” (Block leader #2)“There are two kinds of evaluations: one is the individual level and the other unit level. Blocks grade each other. That is why they need to know what and how other units are doing. Without knowing others, they are not able to evaluate each other. Block units learn about each other, for example, during Block-level meetings.” (Block leader #3)“At the Doppo level, some important points in evaluations are profit and loss statements and cooperativeness. In these evaluations, cooperativeness is especially important. It is a factor that glues people together in the Doppo system.” (Doppo leader #3)“To make your grade better, you’d better work with people around you. If we don’t have much contact, they would not have chances to see our work.” (Doppo leader #8)
Relational interactions	Roles	“In the Doppo system, employees get recognition for their expertise knowledge. If one has rich skills, it is easier for him or her to build relations at Mayekawa. That kind of environment motivates employees to develop their expertise.” (Doppo leader #14)“Doppo leaders have to hold members together, understand their roles as experts, understand the expertise of other members, and lead them to the same direction. To complete projects successfully, they have to unite members and their different opinions and make them to work together willingly.” (Doppo leader #11)“In new projects, all members do not have the same starting point. That is why leaders need to give them the direction to go, identify individual abilities, and to build good team work. This way, all team members try to be a specialist in the field he or she was good at. As time passes, expertise areas became more refined.” (Doppo leader #18)“Block leaders have important roles in personnel transfer. They sometimes move employees to other units when they think it is better for them to exercise their abilities to their full extent . . . they try to match tasks with members with the right expertise.” (Block leader #7)
Relational interactions	Routines	“When personnel transfer is based on Doppo leader agreement, we have to inform the Block leader. Doppo leaders can carry out personnel transfer by themselves, yet they have to get the final approval from the Block leader. The practice occurs in all Blocks in Japan.” (Doppo leader #4)“We have regularly held Block meetings at the regional and national level. In these meetings, we exchange information and become familiar with each other’s operations. We also try to find out if there is any chance to work together.” (Block leader #6)“When making personnel changes, two relevant (Doppo) leaders discuss with each other. I don’t know who starts out first, sempai (senior) or kohai (junior), but anyway, one would tell the plan and ask another to give appropriate people to compose a team.” (Doppo leader #11)“You have to let good members go for new Doppos. That is Mayekawa culture. Freedom and self-discipline were established and have been passed on to new employees since Mayekawa was established. I think that freedom and self-discipline are guiding routines in all Doppo units.” (Block leader #1)

Note. TMS = transactive memory system; HRM = human resource management.

Organization Design

The findings suggest that “team-based” Doppo structure acts as a coordination mechanism and provides the base for organizational TMSs through the flexibility along with task specialization and interdependence it creates at all organizational levels. The Doppo structure is flexible, allowing Mayekawa constituents to design and coordinate their work activities through lateral interactions. Indeed, a company document (Mayekawa, 2000) states that “The flexible organization helps us to combine various seemingly contradictory technologies to arrive at a comprehensive customer solution” (p. 1), and that “Expertise in projects can be drawn from two or more Doppos” (p. 15). At the same time, the Doppo structure is also formal, increasing inter-unit task collaboration, interdependence, and specialization. Complex projects give rise to TMSs among interdependent units with complementary expertise. A Doppo leader (#12) explained a Toridasu (chicken whole leg deboning machine) project:

The development of Toridasu was conducted in collaboration with three Doppos. At that time, I was involved in poultry business and selected employees from several Doppos to the project based on their expertise. During the project, the refinement and combination of members’ expertise helped us to build the machine. The expertise developed by these members was used later in other related projects, such as Yeildas (chicken breast deboning machine) and Hamdas-R (pork ham deboning machine).

This individual and other Doppo leaders said that they rely on their informal, experience-based awareness of who knows what to staff projects with relevant experts. In projects, people further rely on informal interactions to coordinate expertise responsibilities and use each other’s expertise. Thus, the cognitive processes as outlined in TMS theory were apparent in inter-unit projects. Because Doppos are financially independent and project profits are shared, both units and people in them also seek to increase their demand by developing and being known by their expertise. Projects are beneficial, because they expose both employees and units to different expertise and help them to develop and update their awareness of distributed expertise.

The findings also suggest some limitations of the Doppo structure as a coordination mechanism. In particular, the Doppo structure supports clustered rather than organization-wide TMSs due to the more frequent interactions between physically close, interrelated units. These clustered interactions make it more difficult to locate relevant task expertise from physically distant Doppo units with less frequent interactions. The Doppo structure limitations also become more apparent as Mayekawa gets larger. A Doppo leader (#19) explained,

When the Doppo system was introduced, there were 800 employees in Mayekawa. Now they count 2,000 only in Japan. It used to be easy to call appropriate people when staffing projects because I used to know almost all employees by their names and faces. But as the company gets bigger, it becomes harder to know all employees.

Small unit size

The findings suggest that small Doppo size (on average 10 to 20 employees) coordinates organizational TMSs by increasing intra-unit and inter-unit task collaboration, interdependence, and specialization. For example, a company document states that Doppos need to cover several functional areas due to financial independence (Maekawa, 2008). Although all Doppos seek to develop and to be known by their own expertise areas, they need to collaborate due to limited internal resources. In this way, TMSs expand beyond Doppo boundaries. As Doppos get bigger and their operations become more diverse, they give rise to new Doppos, usually without top management interventions. A Block leader (#3) reasoned,

New Doppos are organized when a group of motivated employees get together and the functions of units became specialized to unique subjects . . . Because of their small size, these units need to collaborate with other units in related areas.

The interviews also suggest that small Doppo size motivates employees to accept expert responsibilities and to develop their expertise. Doppo leaders (#7 and #11) noted, “The [Doppo] system provides great opportunities for people to grow their abilities” and “A good point of the Doppo system is that employees have self-contained control of their work.” Because employees are able to work semi-permanently in projects and other units, they are often members of both intra-unit and inter-unit TMSs. The findings also suggest that the small unit size and frequent interactions among Doppos foster TMS coordination and reduce knowledge hoarding and free-riding for three reasons. First, project teams are staffed with members with complementary expertise, creating cognitive interdependence among them. Second, interactions among members are efficient, because 80% of Mayekawa employees are engineers and therefore have sufficient base to internalize shared technical information. Third, members tend to share information openly, because non-contributing members have the risk of developing a bad reputation and low future demand for their services.

HRM Practices

The findings suggest that recruitment and selection practices also coordinate organizational TMSs. Similar to many other Japanese manufacturing companies, Mayekawa recruits the bulk of its employees directly from (engineering) schools. In the selection process, emphasis is given to certain personality traits and social skills, such as the ability to learn and collaborate. In addition to reducing turnover, these practices increase the likelihood that entry-level recruits possess the traits and skills that facilitate TMS processes. A Doppo leader (#3) explained,

Most entry-level employees have engineering degrees . . . we place emphasis on their personality, especially the ability to be proactive and work effectively in teams. These recruits therefore have a suitable base to develop their expertise and work in joint projects.

The practices and programs that follow entry-level recruiting, such as training and mentoring, facilitate newcomers’ awareness of organizational expertise. Twelve Doppo leaders also reported that entry-level recruits are more willing to accept and develop new expertise areas than mid-career recruits. Mid-career recruitment at Mayekawa is limited and used mostly to gain in-demand expertise that is not readily available internally. Numerous practices are used to increase collective awareness of mid-career employee expertise. For example, Block leaders often became aware of new employees’ areas of expertise through the hiring process, and Doppo leaders through informal discussions and formal announcements of recruitment decisions in Doppo and Block-level meetings. Doppo leaders disseminate this information to employees in their units. Through these practices, Mayekawa differentiates entry-level recruits, who incrementally develop their expertise and location information, and mid-career recruits, who are hired to act as domain experts. This heavy reliance on entry-level recruitment shows the importance of experience-related expertise awareness and networks on TMSs coordination at Mayekawa. To illustrate, an entry-level recruited Doppo leader (#16) said he needs to know about 60 well-connected Doppo leaders and to use his direct networks to locate domain experts at Mayekawa.

Training

Training coordinates organizational TMSs by expertise development, differentiation, and awareness. All entry-level recruits initially go through a basic training program, after which they receive OJT under the guidance of senior employees, who also act as their informal mentors. The basic training program provides entry-level employees with a broad overview of company operations and chances to develop cohort-based networks. The networks developed tend to be beneficial and long-lasting because of low turnover, strong interdependencies, and cohort identification at Mayekawa. A Block leader (#10) explained,

New recruits attend a three month’s basic training program. After that, half of them will be trained for service . . . when a customer raises a complaint about our machinery, our service reports the problem. Then they will have a meeting with product development members and take the overview to manufacturing. As a result, there is a strong [cohort-based] network among them. There are about 60 to 70 offices and the most of their managers had experienced the service [training] about 20 years ago.

After entry-level training, employees develop their domain expertise and expertise awareness through OJT and information seeking. The findings suggest that interactions during OJT between sempai (senior) and kohai (junior) employees were open and interactive. In the Mayekawa website, an employee explains, “During [the] training period, I often interacted with senior employees in different functions to gain expertise.” These frequent interactions help junior employees to develop their expertise, networks, and awareness of organizational expertise. Junior employees often join projects and visit customers with their sempai. Instead of formal training programs, as a Doppo leader (#19) explained, the emphasis is “for employees to make efforts to find information holders in the company.” Due to their extensive task and location information, sempai are frequently consulted by kohai. A Doppo leader (#9) who has worked more than 30 years at Mayekawa said, “Young workers come to me for help on how to read drawing boards and anything else they don’t know.” These informal interactions and expert inferences provide the basis for social capital that entry-level employees utilize later when seeking expertise and forming new projects.

Promotion

Consensus-based promotion practices coordinate organizational TMSs by aligning employees’ expertise with their ability to identify and utilize organizational expertise. Formal rank in the Mayekawa organizational hierarchy is based on accumulated expertise, collaborative networks, and the ability to employ oneself. Indeed, the interviews suggest that Doppo leaders act more as project initiators, informal mentors, and expertise coordinators than functional managers in hierarchical organizations. A Block leader (#5) explained: “In addition to motivating people to develop strong skills in their units and promoting this expertise to other units, Doppo leaders need to have good understanding of organization-wide expertise. Otherwise, they are not able to be effective managers at Mayekawa”. Block leaders, who can simultaneously act as Doppo leaders, are selected by the top management based on their expertise, internal and external collaborative networks, and personal characteristics. By overlooking and approving all personnel transfers, Block leaders ensure that people with right expertise are transferred to projects and Doppos. They also help Doppo leaders to locate and utilize expertise within/beyond the Block and organizational boundaries. Due to the importance given on location knowledge and social capital, Doppo and Block leaders have long experience and extensive expertise networks at Mayekawa.

Reward systems and performance evaluations

The findings suggest that reward systems and performance evaluation practices coordinate organizational TMSs through more frequent intra-unit and inter-unit interactions and enhanced expertise awareness. By ensuring that wages of transferred employees do not change (Mayekawa, 2000), uniform reward systems coordinate TMSs by reducing the risk associated with joining new Doppos and projects. Performance evaluations facilitate information encoding and retrieval through collaboration and increased intra-unit and inter-unit interactions. All employees, Doppos, and Blocks are evaluated periodically using a five-point grading system. In Doppos, all employees are evaluated by Doppo leaders based on their expertise, collaboration, and contribution. In line with 22 others, a Doppo leader (#2) explained, “Collaboration is a very important factor [in performance evaluations]. It ties employees together in the Doppo system. Without collaboration, we would fall apart and would not be able to conduct our work successfully.”

For the same reason, a large portion of employees’ bi-annual bonus is linked to Doppo and Block performance. A Doppo leader (#1) explained, “Lateral connections are important for units and individuals. Employees and units with less cooperativeness often receive lower evaluations.” A similar logic guides Doppo and Block evaluations. Each Doppo is evaluated based on its profit and innovativeness by related Doppos, and each Block is evaluated by other Blocks based on its performance. Research units, in turn, are evaluated by related Doppos based on their knowledge, skills, and supporting research. A Doppo leader (#11) explained, “Units were graded A, B, N, C, and D, and the ratio of each grade was fixed, so some units had low grades. We want to give good grades to all units but it is not allowed.” This distribution was explained to force all units to collaborate, interact, and share information.

Relational Interactions

Although not formalized, employees have and seek to fulfill certain role expectations at Mayekawa. Because expertise is an important determinant of collective recognition and internal mobility, employees seek and are expected to fulfill their roles as domain experts. A company document states, “One’s position within Mayekawa is determined by being conscious of one’s role within the group” (Maekawa, 2008, p. 29). As further explained by a Doppo leader (#15), task expertise increases internal mobility: “Since I evaluated them as specialists in the field in which I wanted to start, it wasn’t hard to make a nomination list.” Doppo leader (#4) reasoned that cognitive limitations also lead to specialization: “There is a limit to what a single employee can do. That is why we need to have expertise roles.” For Doppo leaders, important roles are to create profitable projects using their networks to draw the needed expertise to carry them out. Most of their time is allocated to HRM-related matters, such as locating experts to new projects. A Doppo leader (#14) said, “When you become a director . . . you realize the limitations of your Doppo and start thinking about which other Doppos can complement your capabilities and try to cooperate to achieve your objectives.”

Instead of seeking to increase the size of their Doppos and having to cope with excess labor during slow periods, Doppo leaders utilize distributed expertise. This way, they facilitate inter-unit TMS formation in Mayekawa. Within Doppos, as explained by a Doppo leader (#4), “Skillful leaders make employees understand the direction to go, build good team work, and encourage them to be specialists in areas they are good at.” Block leaders, in turn, approve personnel transfer and help Doppo leaders to staff new projects with the right expertise. Although personnel transfers are usually decided informally among the relevant parties, a Block leader (#6) explained, “Block leaders give advice and force members to move to other units if they think it is needed.” Taken together, the findings suggest that whereas Block and Doppo leaders have roles as memory depositories of “who knows what” at Mayekawa, employees have roles as domain experts.

Routines

Routines coordinate organizational TMSs by guiding TMS processes. In addition to coordination through informal lateral interactions, my observations and the interviews provide evidence of structured intra-unit and inter-unit interaction patterns. For example, Doppo leaders develop and refine their expertise awareness in regular meetings through established processes of “selling” their ideas to start projects and discussing and agreeing on personnel transfers. The related financial matters are “very businesslike” as a Doppo leader (#6) explained. Doppo leaders also have to get a formal approval from Block leaders for personnel transfers. If Doppo leaders cannot identify relevant experts or all other units are fully occupied, they routinely consult Block leaders. A Doppo leader (#2) explained, “When I need extra workforce and there is none available in my Block, I ask other Blocks. If I still don’t find one, I ask superior managers. They have wider information [on distributed expertise].”

These routines enhance collective awareness of who knows what in Mayekawa and ensure that new projects are staffed with the right expertise. Routines and associated rules further guide expertise differentiation and development. A Doppo leader (#12) explained, “Since project members had developed expertise in mechatronics, they were transferred to other related projects. We let them focus on mechatronics, nothing else. This is our policy to grow their skills.” Mayekawa also has routines that restructure intra-unit and inter-unit cognitive networks. A Block leader (#1) explained, “Excellent Doppo leaders are transferred to badly performing Doppos to rebuild them.” Facilitated by reciprocity expectations, other Doppos participate by transferring employees to these units. Even the company website and internal documents, respectively, state the following: “Mayekawa engineers must collaborate without boundaries of functionality and interests” and “The most important point is how employees with different expertise unify themselves” (Maekawa, 2008, p. 35).

Discussion

This study addressed two questions: (a) Which mechanisms coordinate organizational TMSs? (b) How do formal and informal coordination mechanisms interact with each other in organizational TMSs? The findings suggest that organization design, HRM practices, and relational interactions coordinate organizational TMSs. Furthermore, the findings suggest that organization design and HRM practices as more formal mechanisms enhance and provide continuity to coordination by relational interactions. The findings are used to develop a model of organizational TMS coordination (see Figure 2).

Figure 2.

A model of organizational TMS coordination.

Theoretical Contributions

First, the findings suggest that organization design (team-based structure and small unit size) coordinates organizational TMSs. Although overlooked in the organizational TMS literature, the findings reflect the organizational coordination design theories (Galbraith, 1973; Thompson, 1967) by suggesting that team-based structures coordinate TMSs, which are subject to high task complexity and interdependence (Wegner, 1986). Yet, instead of being coordinated solely through lateral interactions, the findings provide evidence of hybrid structures that consist of lateral and vertical interactions. Thus, in contrast to the bulk of TMS literature, in which people are assumed to have trustful relations and are motivated to form TMSs (Peltokorpi, 2008), the findings suggest that team-based structures without reinforcing vertical mechanisms can be insufficient to coordinate organizational TMSs.

The findings suggest that small unit size coordinates organizational TMSs. Although scholars have already discussed how employee transfer and participation in several teams can blur the distinction between team and organizational TMSs (Anand et al., 1998), this study empirically shows that small unit size acts as an organizational TMS coordination mechanism. The influence of small unit size on organizational TMS coordination can be explained by resource dependency theory (Benson, 1975) that maintains that resource scarcity forms interdependencies among organizational units. The resource dependency can explain why employees participate and contribute to both intra-unit and inter-unit TMSs.

Proposition 1: Organization design (team-based structure, small unit size) facilitates the coordination of organizational TMSs.

Second, the findings suggest that HRM practices (recruitment and selection, training, promotion, and reward systems and performance evaluations) coordinate organizational TMSs. The findings have similarities with commitment-based HRM practices in terms of the following: (a) an internal labor market by promoting from within, (b) team-based rewards, and (c) training and performance appraisal aimed at employee growth and development (Collins & Smith, 2006). In a related vein, entry-level recruitment is found to increase cohort-based interactions in organizations (Williams & O’Reilly, 1998) and reduce turnover that hinders TMSs functioning (Moreland, 1999). Furthermore, OJT is noted to support the awareness of distributed knowledge in organizations (Kang, Morris, & Snell, 2007) and organizational TMSs (Moreland, 1999). Reward interdependence is also argued to enhance coordination among interdependent units (Wageman, 1995) and organizational TMSs (Peltokorpi, 2012).

Proposition 2: HRM practices (recruitment and selection, training, promotion, and reward systems and performance evaluations) facilitate the coordination of organizational TMSs.

Third, the findings suggest that relational interactions (roles, routines) coordinate organizational TMSs. In teams, TMSs are coordinated through shared expertise awareness and expertise role differentiation (Lewis, 2003). The findings suggest that organizational TMSs are coordinated through location and task information role differentiation. Whereas leaders have roles as location information depositories, engineering staff have roles as task information depositories. Furthermore, the findings suggest differences in formal and informal role coordination. That is, leaders relied on their awareness of organizational expertise to staff new projects and to coordinate expertise in them. Project members engaged in informal role-based coordination by refining task allocation based on their awareness of one another’s expertise. In a related manner, established roles are shown to enhance TMS formation in related projects (Lewis et al., 2007).

The findings suggest that routines coordinate organizational TMSs. For example, the routines used to initiate and staff new projects coordinate TMSs by helping leaders to form and refine their awareness of distributed expertise. Despite having narrower bandwidths (March & Simon, 1958), formal routines coordinate organizational TMSs as sources of connections and overlapping understanding among employees in organizations. Formal routines are important because they build connections among relevant but disconnected parties and are less affected by turnover than ostensive routines (Ren & Argote, 2011). However, the findings also suggest that employees develop and maintain ostensive routines that coordinate organizational TMSs through repeated interactions with one another. The ostensive routines have an especially important role in new projects that span existing technological boundaries.

Proposition 3: Relational interactions (roles, routines) facilitate the coordination of organizational TMSs.

Fourth, the findings reflect the information and contingency-based theories (Galbraith, 1973; Thompson, 1967) by suggesting that organization design and HRM practices as more formal mechanisms reinforce and provide continuity to coordination by relational interactions. Indeed, formal mechanisms not only coordinate organizational TMS processes but also provide structures that can be used when coordination through informal mechanisms fails. For example, the findings suggest that uniform promotion practices make it more likely that employees in higher echelons meet the role expectations as boundary spanners and expertise coordinators. These roles are aligned with team-based structures. In a related vein, team leaders in team-based organizations are shown to have roles as boundary spanners, establishing and coordinating team interactions with parties in the external environment (Marrone, 2010). Yet, while team leader roles guide boundary spanning activities, the actual activities are largely based on informal interactions.

Proposition 4a: Relational interactions partially mediate the relationship between organizational design and organizational TMSs.

Proposition 4b: Relational interactions partially mediate the relationship between HRM practices and organizational TMSs.

Fifth, the findings suggest that clustered, network-type TMSs exist in team-based organizations characterized by complex collaborative tasks and interdependent units. In particular, the findings suggest that TMSs are more apparent in clustered intra-unit and inter-unit collaboration rather than homogeneously distributed across units at the organizational level. TMSs are stronger in dense network clusters due to more frequent interactions and opportunities to encode, store, and retrieve information. In these clusters, interactions are conducted largely through reciprocal network ties. Because these clusters, connected by weaker ties, make it possible to encode and retrieve non-complex information, TMSs in organizations can resemble small world networks with high intra-unit network density and short average path lengths of TMS directories among people in different units. Clustering promotes local TMSs, whereas linkages among clusters promote reachability through the network.

Practical Implications

The findings can be used to provide practical implications. First, the findings suggest that an organization that seeks to form organizational TMSs needs to use both formal and informal coordination mechanisms. For example, structuring work around interrelated units with different expertise areas and using HRM practices to facilitate inter-unit interactions can be used to form organizational TMSs. In particular, the findings suggest that job rotation and inter-unit collaborative projects enhance organizational TMS formation and functioning. In addition, the use of interrelated teams, together with a relatively powerful hierarchy, can be beneficial in large organizations. Second, the findings suggest that well-coordinated organizational TMSs provide efficiency and innovation related benefits, such as faster project completion and innovative products. Indeed, the very reason for Mayekawa to adopt the Doppo structure was to speed up project completion and product innovation. The findings further suggest that the Doppo structure together with other coordination mechanisms facilitated several product innovations, such as Toridasu.

Limitations and Suggestions for Future Research

This study has limitations that should be taken into account in future research. First, this and other case studies are weak in terms of external validity. While I expect similar findings to occur in other knowledge-intensive, team-based organizations, more research is needed in other settings. Second, this study provides a managerial account of organizational TMS coordination. Although having a broad awareness of organizational TMS processes (Peltokorpi, 2012), managers are not able to provide an accurate account of employee-level phenomena. Extending interviews to all organizational layers helps to provide a more balanced account of organizational TMS coordination.

Footnotes

Appendix

Acknowledgements

I would like to thank three very helpful anonymous reviewers, Editor Gayle Baugh, and Rebecca Piekkari and Eero Vaara for incisive comments on earlier drafts of this study. I would also like to thank Emiko Tsuyuki for collaboration and Ikujiro Nonaka for introducting me to the case company.

Author’s Note

A previous version of this study was presented at the 25th EGOS (European Group for Organizational Studies) Colloquium in Barcelona, Spain.

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.

Author Biography

Vesa Peltokorpi is an associate professor at Japan advanced Institute of Science and Technology. He has published more than 30 articles in academic journals, such as Journal of Management Studies and Journal of International Business Studies.

References

Amis

J. M.

Silk

M. L.

(2008). The philosophy and politics in qualitative organizational research. Organizational Research Methods, 11, 456-480.

Anand

Manz

C. C.

Glick

W. H.

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

Bechky

B. A.

(2006). Gaffers, gofers, and grips: Role-based coordination in temporary organizations. Organization Science, 17, 3-21.

Benson

J. K.

(1975). The interorganizational network as a political economy. Administrative Science Quarterly, 20, 229-249.

Brauner

Becker

(2006). Beyond knowledge sharing: The management of transactive knowledge systems. Knowledge and Process Management, 13, 62-71.

Collins

C. J.

Smith

K. G.

(2006). Knowledge exchange and combination: The role of human resource practices in the performance of high technology firms. Academy of Management Journal, 49, 544-560.

Cross

Sproull

(2004). More than an answer: Information relationships for actionable knowledge. Organization Science, 15, 446-462.

Eisenhardt

K. M.

(1989). Building theories from case study research. Academy of Management Review, 14, 532-550.

Eisenhardt

K. M.

Graebner

M. E.

(2007). Theory building from cases: Opportunities and challenges. Academy of Management Journal, 50, 25-32.

10.

Evans

W. R.

Davis

W. D.

(2005). High-performance work systems and organizational performance: The mediating role of internal social structure. Journal of Management, 31, 758-775.

11.

Faraj

S. L.

Sproull

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

12.

Faraj

S. L.

Xiao

(2006). Coordination in fast-response organizations. Management Science, 52, 1155-1189.

13.

Galbraith

J. R.

(1973). Designing complex organizations. Reading, MA: Addison-Wesley.

14.

Gephart

R. P.

(2004). Qualitative research and the Academy of Management Journal. Academy of Management Journal, 47, 454-462.

15.

Gittell

J. H.

(2000). Organizing work to support relational coordination. International Journal of Human Resource Management, 11, 517-539.

16.

Jackson

(2012). Transactive directories of organizational memory: Towards a working data model. Information & Management, 49, 118-125.

17.

Jackson

Klobas

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

18.

Kang

S.-C.

Morris

S. S.

Snell

S. A.

(2007). Relational archetypes, organizational learning, and value creation: Extending the human resource architecture. Academy of Management Review, 32, 236-256.

19.

King

(2004). Using templates in the thematic analysis of text. In Cassell

Symon

(Eds.), Essential guide to qualitative methods in organizational research (pp. 256-270). London, England: SAGE.

20.

Krackhardt

(1994). Graph theoretical dimensions of informal organizations. In Carley

K. M. P.

Prietula

(Eds.), Computational organizational theory (pp. 89-111). Hillsdale, NJ: Lawrence Erlbaum.

21.

Kvale

Brinkman

(2009). Interviews: Learning the craft of qualitative interviewing (2nd ed.). Thousand Oaks, CA: SAGE.

22.

Lewis

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

23.

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.

24.

Maekawa

(2008). Basho management: New Japanese philosophy of manufacturing and empowering people. Bloomington, IN: Xlibris Corporation.

25.

March

J. G.

Simon

H. A.

(1958). Organizations. New York, NY: John Wiley.

26.

Marrone

J. A.

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

27.

Mayekawa. (2000). Mayekawa Mfg. Co, Ltd Group: Products, technology, challenges. Tokyo, Japan.

28.

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. 3-32). Mahwah, NJ: Lawrence Erlbaum.

29.

Moreland

R. L.

Argote

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

Mannix

E. A.

(Eds.), Understanding the dynamic organization (pp. 135-162). Mahwah, NJ: Lawrence Erlbaum.

30.

Okhuysen

G. A.

Bechky

B. A.

(2009). Coordination in organizations: An integrative perspective. Academy of Management Annals, 3, 463-502.

31.

Peltokorpi

(2004). Transactive memory directories in small work units. Personnel Review, 33, 446-467.

32.

Peltokorpi

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

33.

Peltokorpi

(2012). Organizational transactive memory systems: Review and extension. European Psychologist, 17, 11-20.

34.

Ren

Argote

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

35.

Ren

Carley

K. M.

Argote

(2006). The contingent effects of transactive memory: When it is more beneficial to know what others know? Management Science, 52, 671-682.

36.

Thompson

J. D.

(1967). Organization in action. Chicago, IL: McGraw-Hill.

37.

Van

Ven

A. H.

Delbecq

A. L.

Koenig

(1976). Determinants of coordination modes within organizations. American Sociological Review, 41, 322-338.

38.

Wageman

(1995). Interdependence and group effectiveness. Administrative Science Quarterly, 40, 145-180.

39.

Wegner

D. M.

(1986). 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-Verlag.

40.

Williams

K. Y.

O’Reilly

C. A.

(1998). Demography and diversity in organizations: A review of 40 years of research. Research in Organizational Behavior, 20, 77-140.

41.

Yin

R. K.

(2014). Case study research: Design and methods (5th ed.). London, England: SAGE.