Cyclical group development and interaction-based leadership emergence in autonomous teams: an integrated model.

Cyclical models of group development may be integrated with an interactive model of leadership emergence in autonomous teams. This paper contextualizes the interaction of leader traits and situational factors in the cycling of an autonomous team within and between the storming, norming and performing phases of group development.

**********

The attention to groups and teams, in both the academic literature and the practical realm, has grown steadily in recent years. In practice, one needs only to scan the classified advertisements in any newspaper to notice the emphasis on team-oriented skills in the workplace. In the literature, from McGrath's (1964) first systematic study of teams to Gladstein's (1984) first model of effectiveness in self managed work teams (SMWT), organizational research all but prophesied this real world emphasis.

Surely, the reorganizations and elimination of management layers that have characterized recent practice have not been intended to cause anarchy, nor have they purported to reduce the level of leadership in these organizations. Organizations that cut formal supervision and management layers are not attempting to reduce the number or quality of their leaders. Rather, companies that are so (re)structured make themselves increasingly reliant upon teams to produce leadership and to be the entities from which organizational leadership emerges (Seers, 2001). We may infer that these organizations, while they are purposefully attempting to reduce the expense of leadership (i.e., bureaucracy), are, perhaps inadvertently, preserving and unfettering the essence of leadership. To the extent that transactional leadership is supplanted by transformational leadership, the ability to study, and even predict, the patterns and phenomena of such leadership emergence has meaningful implications for scholars and executives. Thus, understanding the interaction between team processes and organizational leadership emergence is imperative for researchers and practitioners alike.

This paper examines the interaction between team processes and leadership emergence by drawing upon an update of the familiar Tuckman (1965) model and offering it as the interactive context for the dynamic leadership emergence process. In doing so, it offers the following contributions: (1) extension of the theory that team processes serve as substitutes for hierarchical structure, (2) a better understanding of both the interaction of the phases of group development and the interactions of group constituents, particularly as leadership emerges, (3) the examination of leadership emergence as contextualized in a natural and dynamic frame (4) extension of Osborn, Hunt and Jauch's (2002) work, addressing the embeddedness of leadership within its context.

Group Context

Cohen and Bailey (1997) offer a definition of the team that represents a group of individuals, who are perceived and perceive themselves as an intact social entity, working together toward a common goal or goals, and who manage relationships and exchanges across organizational boundaries. In this paper, I use the terms group and team interchangeably, based upon the assumption that a team is a specific, goal-oriented kind of group. I also refer to followers and members synonymously, based upon the assumption that a follower is a specific kind of group, or team, member.

Even though many, if not most, instructors still present Tuckman's five-stage model of group development, Tuckman himself revisited the original five-component (forming, storming, norming, performing, and adjourning) model and altered it in favor of a model that emphasizes only the storming, norming, and performing phases (Tuckman & Jensen 1977). Further, Bales and Cohen (1979) suggest that the team development process is not necessarily linear as researchers such as Tuckman iterated, but that it is cyclical. They, along with Arrow and McGrath (1995), imply that a modification based upon elements of Tuckman's hypothesis is useful. This cyclical modification of Tuckman's hypothesis involves what Tuckman identified as the two "realms" of group behavior, task-activity behavior and interpersonal behavior (Bales & Cohen 1979) (Arrow & McGrath 1995) (Tuckman & Jensen 1977). Also of note is Tuckman's revisitation of the original five-component (forming, storming, norming, performing, and adjourning) model in favor of the realm-based model that emphasizes only the storming, norming, and perfomaing phases (Tuckman & Jensen 1977). Integrating an interactive model of leadership emergence with the "realms" of this cyclical model will result in a better understanding of both the interaction of the phases of group development and the interactions of group constituents, particularly as leadership emerges in the context of both newly formed and pre-existing autonomous teams.

Further, integrating a cyclical model of autonomous group development with an interactional model of leadership emergence appropriately embeds the leadership emergence process in a natural and dynamic frame. Figure one (below) offers a depiction of this integrated model. In this paper, I suggest that it is the storming phase of group development in which members signal their expectations for leadership, leader behaviors emerge, and a potential leader is allowed influence by followers. The storming phase cycles and recycles through actions of potential leaders, reactions of followers to those actions, reactions of potential leaders to follower reactions, and so on, finally reaching an equilibrium and completing itself. The norming phase sees a pattern of established leader behaviors, referred to as emergent leadership, and the performing phase is a task-oriented one in which relationship oriented leader behaviors are superseded by task oriented leader behaviors. When team composition, the task or another situational factor changes significantly, member expectations for leadership are altered in a return to the storming phase, wherein new expectational signals are sent and the process repeats, or recycles, itself.

This integrated model brings together two processes, one of which is embedded in the other. It is a recognition that the cyclical group development process serves as the context for the interaction-based leadership emergence process.

P1: The interactive process of leadership emergence is embedded in a cyclical group development process.

Networked Context

Cohen and Bailey state that the factors of a team are threefold: structure, process, and psychosocial factors. DeSanctis and Poole (1997) echo the work of Cohen and Bailey. They also argue for three separate dimensions of teams: structure, process, and a more specific psychosocial factor, social identity. Relatedly, and of great importance here, they propose the advent of a new organizational structure, the networked form of organization. This networked form is conceptually somewhere between traditional hierarchical forms, and market structures. It is characterized by fluidity, flexibility, learning, and exchanges across various organizational boundaries. This work regarding such a "post-bureaucratic" organization is quite germane to the study of teams, as DeSanctis and Poole state that the autonomous team is the fundamental substructure of the networked form of organization. We may appropriately conceptualize the process element of a team as actually comprising its structure in the case of the autonomous team or SMWT. As such, the autonomous team actually substitutes for hierarchy (LaMer, 1988) and, by extension, team processes substitute for hierarchical structure. The pattern of interaction (Katz & Kahn, 1978) becomes the structure. Thus, the patterns of teams provide the structural context in which leaders emerge.

The omission of overt contextualization in leadership emergence research has been, in part, supplanted by an implied context of hierarchical structure. Scholars have examined strategic management and leadership issues from the assumption of an in-place hierarchy, leaders who are in management/leadership positions that would exist regardless of who is in those positions. In recent years, however, we have turned our attention toward other organizational forms, beginning with market-driven forms and progressing to the network form of organization (Powell 1990). The implication here is that since leaders, in increasing examples, are not formal, they must be viewed as "informal" and "designated" through network team processes rather than through hierarchical edicts. These process designations are more fluid and reflect "true" leadership as elicited by followers, as opposed to figurehead leadership enacted formally.

This integration also is consistent with, and extends, the concept of shared leadership (Pearce, 2004). Leadership in autonomous teams can be, and often is, shared; in fact, that is largely the point of recycling through the group development/leadership emergence process, where different leaders emerge at different times based upon situations, tasks, and/or group composition. In an autonomous team, not only might single leaders emerge at different points in time, based upon situational or task concerns, but more than one leader may be emergent at any given point in time. For example, one leader may be task-oriented, while another may be focused on interpersonal matters at the same time, such that the two share leadership functions. This paper takes the view that leadership, especially in autonomous teams, can be shared concurrently, rather than always in a "taking turns" mode.