Locating Distributed Leadership


Email: r.thorpe@lubs.leeds.ac.uk


This special issue addresses a number of the key themes that have been surfacing from the literature on distributed leadership (DL) for some time. Together with those papers selected to be included in this special issue, the authors set out both to explore and contribute to a number of the current academic debates in relation to DL, while at the same time examining the extent to which research on DL has permeated the management field. The paper examines a number of key concepts, ideas and themes in relation to DL and, in so doing, highlights the insights offered through new contributions and interpretations. The paper offers a means by which forms of DL might be conceptualized to be better incorporated into researchers' scholarship and research, and a framework is presented which considers a number of different dimensions of DL, how it may be planned, and how it may emerge, together with how it may or may not align with other organizational activities and aspects.


The significance of this paper lies in the claims made for a view of leadership that takes account of the structural changes taking place in many organizations. Distributed leadership (DL) is the very antithesis to the preoccupation of most Western writers about leadership, with the position reified in single individuals – usually those at the top of an organization (Harris 2009). Over the last 50 years, much of the research has focused on organizational forms of a hierarchical nature; wittingly or not, this reinforces the view that knowledge and direction trickle down from a notional ‘top’ of the organization, and that leaders as individuals (depicted as being ‘at the apex of the triangle’ or ‘on top of the hill’) set the tone and make all the key decisions. These leaders are seen as acting as figureheads, for others (depicted as ‘followers’) to look up to and to follow (MacBeath 1998; Pearce and Manz 2005). Over this period, there has been a decided lack of research evidence (Thorpe et al. 2007) for single individuals having the kind of dramatic impact on organizational performance that is so often claimed. An exception is in the work done by Pfeffer and Sutton (2006, p. 162); their findings have been repeated in other studies going back over many years that the evidence for individual influence was, under most conditions, modest at best.

What we know about organizations is that the speed of external change – whether operational, strategic or technological – is now rapid (e.g. Dunning 2008; Lüscher and Lewis 2008; Ojasalo 2008). This produces an imperative within organizations for them to be able to respond ever more quickly and adaptively. The kind of responses made have been evident for some considerable time, from structural adaptation to a whole range of workforce strategies, which include job design, motivation and management style, rewarded remuneration and employee assurances (Walton 1986). As for organizational structures, we have seen these evolve into forms that can cope with the ambiguity and tensions which rapid change brings (Whittington and Mayer 2002). These organizational forms have included flatter structures, matrix structures and ever more widely linked network structures, all of which reflect, in their varying ways, the limitations of top-down models and the limitations of leadership when the unit of analysis is a single individual. Where organizations have become increasingly project- or knowledge-based (Lindkvist 2004), where they involve professional work (Vermak and Weggeman 1999) or where innovation occurs through knowledge of intensive exchange processes within networks (Hallikas et al. 2009), leadership is now moving to a form that is able to cope with collective endeavour, where individuals can contribute to the establishment and development of a common purpose (a common vision). This literature draws attention to the evidence showing a lack of coherent and consistent understanding of what leadership is; at the same time, it highlights the ‘silences’ that continue to exist in the research literature. We believe that the time is right for academics to break out of their current mindset as represented by that literature (produced by a research agenda that focuses almost exclusively on individuals), and to give attention to the collective forms of leadership that are the focus of this special issue and which we characterize as distributed.

Distributed leadership

Distributed leadership is receiving increasing attention in practitioner and academic literature that focuses on a number of practice-based social science disciplines, but we sense the absence of an understanding of its importance in the largest of the applied social sciences – management and business studies. In some sectors, for example, higher education, where the concept of DL has received some notice, there has already been a backlash against the term, as it is seen as part of a rhetorical ploy by organizations (the higher education sector) to avoid consulting their staff (Gosling et al. 2009). Yet, the broader term ‘leadership’ remains an ‘eternally contested concept’ (Grint 2005a), and a universally accepted definition of DL remains elusive. However, in our view, the papers in this review provide the best overview of current developments to date. Perhaps, though, we need to ask why this is the case and, given the complexities already perceived in the study of leadership, we would want to investigate this particular approach – a ‘distributed’ approach.

At a practice level, many managers and organizations continue to strive to achieve increased success, efficiency, flexibility, competitiveness, ability to survive and long-term sustainability. Against such pressures, anything that offers the possibility of managers achieving better outcomes merits interest, if for no other reason than practicality and pragmatism. At an academic level, we have seen (as the papers contained in this special issue demonstrate) that with the attention given to DL comes a need to examine what it means, how it is interpreted, how robust it is as a management concept, and what it might add to current understanding of organizational leadership.

The current interest in this topic is, in part, a reflection of a growing disenchantment (felt by practitioners and academics alike) with solely individualistic notions of leadership, heroic or otherwise (Gronn 2002). Part of the reason for this can be explained by academic preoccupation with methodological individualism (Schumpeter 1909) when considering organizational leadership, and failure to conduct studies sufficient to embrace wider cultural, historical and social aspects of context: issues to which we return later. The result, as we have seen in our own systematic review of the literature (Thorpe et al. 2007), is something of a research cul-de-sac, with a continuous reinvention of capabilities and competences that becomes dated as soon as the ink has dried. Furthermore, the extant literature on DL is largely descriptive and normative rather than critical. For instance, there have been early attempts to provide individual leaders with a toolbox of skills in DL (McBeth 2008), so adopting a new style that can easily come to be seen as a form of empowerment, participation or engagement. Such approaches characterize a movement which has sought, over the last 20 years, to persuade others to take more responsibility to exert discretionary effort on behalf of the organization.

Definitions of distributed leadership

At one level, we have addressed the question of why we should be interested: because it is hinted at in certain contexts, but not identified and discussed anything like clearly enough, except in education. Furthermore, while other people (for example, consultants) are interested, as yet there is little evidence of different approaches being discussed in an integrated and coherent way. Thus, DL appears to be a relatively unexplored concept which has both academic and practical potential. Although many writers (e.g. Bennett et al. 2003; Harris 2003) have linked the ideas it contains to concepts such as participation, empowerment, engagement and delegation (in principle at least), it may constitute something very different. At the same time, DL is also of interest because of what is not known about the concept and why it is absent in some contexts, while receiving attention in others. So we would like to be able to address basic issues such as: what DL is; how distinct it is from other management and leadership approaches; how it operates; where it can be applied; and how we should develop our understanding of the concept. Crucially, DL is considered as a social phenomenon with a context integral to its understanding and, indeed, constitutive of the practice of leadership, concerned with thinking and actions in situ. The focus therefore is on conjoint actions rather than role or position. It is the way in which leading is enacted in the performance of tasks that is important (Spillane et al. 2001). In addressing these issues, we consider DL in all its forms, as further papers in this issue also note. Thus, co-leadership (Heenan and Bennis 1999), shared leadership (Fletcher and Käufer 2003; Pearce and Sims 2000) and self-managed teams (Barker 1993), among others, are all seen to represent forms of DL. For our purposes and drawing on a range of approaches, we choose to define DL as a variety of configurations which emerge from the exercise of influence that produces interdependent and conjoint action.

Current context and a proposed framework

The papers that make up this issue highlight two strong contextual factors that, we believe, merit examination in relation to DL. The first of these is the apparent increase of individualism (as opposed to collectivism) in Western economies (e.g. Bauman 2001; Beck and Beck-Gernsheim 2002; Hofstede 2001; Holt and DeVore 2005), which reinforce the individualistic interpretation of leadership even when faced by a contrasting example of what we could see as collective effort or conjoint action. To some extent, notions of DL may appear counter to individualism; so we may ask why the interest in this view of leadership should be developing in this context, where it could be seen as paradoxical. Secondly, and perhaps in part answer to the previous question, there is an increasing counter to the development of globalization, as evidenced by an increased attention on what is happening and what should happen at local levels in individual organizations and individual communities (e.g. Kirst Ashman 2008; Lipsky 2010). Perhaps DL offers a means of exercising localized and contingent influence.

The issue of a culture of individualism is important and may go unacknowledged in much Western exploration and examination of leadership. This is reflected, as some commentaries suggest (e.g. Gronn 2002), in the twentieth-century preoccupation with leadership research and theory development focusing on the characteristics, behaviours, styles and outcomes of the work of individual agents as leaders. Similarly, Senge (2002, p. 22) argues that the prevailing individualistic culture decidedly restricts our view of organizational leadership:

Our traditional view of leaders – as people who set the direction, make the key decisions and energize the troops – is deeply rooted in an individualistic and non-systemic worldview. Especially in the West, leaders are heroes – great men (and occasionally women) who rise to the fore in times of crisis. So long as such myths prevail, they reinforce a focus on short-term events and charismatic heroes rather than on systemic forces and collective learning. (Italics in original)

Brown and Hosking (1986) and Gronn (2002) – and Gibb (1947) before them – sought to redirect efforts towards leadership practice as a social process. The focus could shift towards actions and the performance of tasks whereby leadership is stretched over people and situations (Spillane et al. 2001). Distributed leadership is concerned with how influence is exerted in such situations to achieve successful outcomes. It therefore becomes more difficult to make longstanding attributions of leadership power as a characteristic of a single agent. Instead, the construct of leadership is considered as organization- or inter-organization-wide and as a shared attribute or, as presented by Lambert (2002, p. 1), ‘the professional work of everyone’. In addition to professional work, we ought to add any situation where work requires the discretionary application of skills and understanding in conjoint action with others, making command and control notions of leadership, as exemplified by the triangular image of a vertically structured Victorian factory, inappropriate (Ross et al. 2005)

Educational research on DL began to appear in the early 2000s. For example, using evidence from research on school improvement projects, Harris and Muijs (2003) suggested that, although an autocratic leadership style with a single head was prevalent while a school faced special measures, leadership needed to be shared in order to continue progress. This allows the idea of ‘teacher leadership’ to be advanced as a manifestation or realization of DL in action, where, through collaboration and collegiate ways of working, all teachers can take the lead and, in so doing, advance a school's capacity for change and development. The work of Spillane (Spillane et al. 2001; Spillane 2006) on DL relates largely to schools in the US, where principals, assistant principals, curriculum or subject specialists, mentoring peers and others could all be identified as taking on leadership responsibilities. Such emerging evidence was augmented in the systematic view by Bell et al. (2002), which set the question: ‘What is the evidence of the impact of school head-teachers and principals on student/pupil outcomes?’ A search of databases yielded references to over 100,000 papers and books, and 27 studies were eventually selected for review. Of these, eight showed that school leaders had an impact, to varying degrees, on student outcomes, and these were selected for analysis. The results confirmed that leadership was important to school success, but the impact of head-teachers upon student attainment was indirect and mediated by others. Significantly, teachers' work, school organization and relationships with parents and the wider community were identified as key intermediate factors. Thus, leadership was not just a feature of the head-teacher or senior management; it was distributed among staff and others, where it proved to impact directly on student learning outcomes.

Since then, especially in education research, there has been emerging evidence that a positive relationship between DL and outcomes (Leithwood et al. 2007) and normative guidance can be provided (Harris 2008). Against this evidence, Gronn (2009) has suggested more recently that the term ‘distributed’ might seem ‘anomalous’, since leadership could occur within a variety of practice situations involving teams and networks within and between organization units, as well as a more simple person-plus pattern (Spillane 2006). To encourage consideration of these ‘hybrid’ forms of leadership, Gronn (2009) suggests broadening the term ‘DL’ to ‘leadership configuration’. In this respect, the development of interest in DL constitutes an important example of a shift in focus from one on leaders to one on leadership (Grint 2005a).

From leaders to leadership . . . 

Much leadership research is about what individual leaders do or ought to do – in a top-down way (e.g. Bass and Riggio 2008; Kouzes and Posner 2007), where leaders are the main originators and directors of leadership efforts. This notion of ‘vertical leadership’ (Yukl 1999), as exercised by an individual leader who acts to influence followers through the principle of ‘unitary command’, does much to simplify matters in terms of what to study, who to identify as leaders and who to develop. While interest can also be extended to dyads or leader–follower pair exchanges, such as the influential ‘leader–member exchange’ theory (Graen and Uhl-Bien 1995), individually conceived leadership is central to most conceptions of ‘transformational’ or ‘charismatic’ leadership. However, while normative leadership research sees ‘the leader as consistent essence, a centred subject with a particular orientation’ (Alvesson and Sveningsson 2003, p. 961), recognition that individuals are not ‘the autonomous, self-determining individual(s) with a secure unitary identity [at] the centre of the social universe’ (Alvesson and Deetz 2000, p. 98) has long given rise to efforts, albeit seldom reported, to move beyond such a narrow focus. As long ago as 1948, at the height of studies in the US that gave so much value to behaviour models of leadership, Benne and Sheats (1948, p. 41) had already attempted to obfuscate the distinction between leader and follower by suggesting that ‘groups may operate with various degrees of diffusion of “leadership” functions among group members or of concentration of such functions in one member or a few members’. Instead, leadership needed to emphasize ‘multilaterally shared responsibility’, relying on the skills of group members to enact a range of 19 task and maintenance roles. Cecil Gibb, probably the first writer to employ the term ‘DL’, also focused on groups. In the 1940s at the University of Sydney, and then at the University of Illinois (where he completed his PhD), Gibb was part of a research process which helped to show a need for models of leadership that moved beyond a preoccupation with individuals' traits. He argued that leadership, rather than being seen as a fixed attribute of an individual, was better considered as a function of group aims and values, and the techniques available to the group for reaching its goal (Gibb 1947). Continuing this theme into the 1950s, the same author argued that ‘leadership is probably best conceived as a group quality, as a set of functions which must be carried out by the group’ (Gibb 1954, p. 844); as a result, there was a ‘tendency for leadership to pass from one individual to another as the situation changes’ (p. 902). Leadership therefore could be concentrated and focused where one person's influence dominates, or dispersed or distributed based on mutuality and reciprocal influence of different participants. Gronn (2002, 2009) worked with this contrast in two ways: first, to restore or rediscover the concept of DL as a unit of analysis and, secondly, to suggest a need to consider a mix of focused–distributed hybrid formations.

Brookes (2008, p. 6), in his examination of public leadership, echoes and elaborates this point on ‘vertical’ leadership, though his work uses the concept of ‘vertical’ differently from the way it is used in relation to individual leadership. He notes the limited literature on ‘collective leadership’, which he sees as having two elements: ‘horizontal (shared between organizations); and vertical (distributed throughout each organization)’. While the horizontal dimension is named ‘shared’ and the vertical ‘distributed’, he uses these labels interchangeably; in our interpretation, they both form aspects of DL.

Gronn (2002) suggests that aggregation represents a ‘minimalist’ version of DL, and that it does provide a degree of attraction to those in leader positions, in so far as responsibility for performance can be shared among several others. Spillane (2006) refers to this view of DL as the ‘leader-plus aspect’ which can begin with co-leadership (Heenan and Bennis 1999) or leadership couples (Gronn and Hamilton 2004) as a partnership between leaders, but can also embrace many individuals who can be required to take a leader role, whether formally recognized or not. Gronn (2002) sees a switch in the unit of analysis from an aggregation of individuals towards concertive action; it is this view of DL that has gained most consideration in recent years, with its concern for leadership practice and the stretching of practice over multiple leaders (Spillane 2006). Gronn provides three forms of concertive action:

  • 1spontaneous collaboration where leadership practice arises in response to particular problems and requirements: interactions between different participants, with different skills and backgrounds, bring about a solution or a way of moving on; the collaboration may occur fleetingly, disbanding when the problem is solved or requirement met, but can also set in motion further opportunities for joint working
  • 2intuitive working relations based on the close understanding of at least two people which emerges through practice: interdependence and reliance on each other create a ‘shared role space’ (Gronn 2002, p. 627), with mutual trust being a key factor
  • 3institutionalized practice, where learning from (1) and (2) is formalized to some degree.

In addition to forms of concerted action, we must consider the different functions of leadership. In education, for example, Harris (2003, p. 78) notes four dimensions of the role of teacher leader:

  • 1Brokering: the way in which staff translate the principles of school improvement into practice in classrooms and other locations within the school. Links between all are secured, and opportunities for learning and development are seized and maximized. We can see this as leadership of practice.
  • 2Participating: staff are empowered and given ownership of particular changes or developments. Everyone feels equipped with a part to play in change. Collaboration is sought, and work is directed towards a collective goal which everyone has taken part in setting. This can be seen as leadership of the school.
  • 3Mediating: everyone is a potential source of expertise and key information. It is possible to draw upon additional resources and expertise if necessary and to seek personal assistance. This we can see as leadership/management of resources.
  • 4Relationships: there is mutual learning through close relationships between staff. Learning is the source of school improvement. Professional learning and development are distinctive within the school. There is work to build the capacity of everyone to help manage the school. This we can see as leadership of learning.

Lawler (2007) notes the functions of leadership in a different context – social work – which might add a fifth dimension: leading the profession. This includes: leadership as promoting the profession itself; leadership as encouraging and developing staff effectiveness; leadership of inter-professional activities; and leadership to counterbalance the dominance of managerialist values. Thus, further examination of DL needs clear acknowledgement of context and purpose of leadership. In our review of public-sector and small and medium-sized enterprise (SME) leadership (Thorpe et al. 2007), we note the paucity of empirical evidence for leadership in those contexts and conclude that, again in those contexts, there was little investigation of the concept of leadership as a separate and distinct entity. In that research, leadership was generally considered together with other organizational factors and processes, rather than as a single focus. To provide a basis for discussion, a number of composite themes were developed, with the concept of leadership being discussed variously in relation to:

  • • learning/change/culture/innovation
  • • leadership development/succession/turnover
  • • performance/excellence/motivation/service
  • • characteristics/collaboration/decision-making and risk.

All these composite themes have resonance in the developing discussions on DL.

To some degree, it has become possible for those in formal leadership roles to consider DL as a normative stance on leadership, with a requirement to learn how leadership works in their organization. Harris (2006) provides four dimensions for normative DL:

  • 1The representational dimension, which provides recognition for lateral and cross-boundary collaboration as new forms of organizing emerge: thus, partnerships, networks and federations all imply less vertical/top-down leadership based on hierarchical positions.
  • 2The illustrative dimension, which is a reflection of the requirement for allocation of tasks of responsibility to others by expanding leadership teams and sharing of responsibilities.
  • 3The descriptive dimension, which is concerned with finding out what DL ‘looks like’. If it is believed, it can be seen, as Weick (1979) might have suggested. This dimension is a challenge to those seeking a simple formula and programmes that verge on the idea of nominated leaders as distributors. Instead, the formula becomes ‘seek and ye shall find’, within departments, teams, groups, projects and learning programmes such as action learning sets.
  • 4The predictive possibilities for DL to improve outcomes and enhance an organization's capacity for development and change: here, Harris cites a range of studies from education research to support a positive correlation (e.g. Graetz 2000; Morrisey 2000).

The development of thinking, away from individual leaders and focusing on leadership as a process, opens the way to consideration of a wider range of issues: to the different functions of leadership; to the different configurations of concerted action; to relational issues of leadership; to the development and emergence of leadership in practice; and to the contexts in which leadership occurs. The next section provides a framework within which to consider these various issues and to explore the dimensions of DL.

Dimensions of distributed leadership

So, while the literature on DL is developing, much of it is still normative in nature: the number of critical articles on the topic is very small to date, as several contributors to this special issue highlight. In our discussion here, and indeed through this volume, we wish to extend the contributions to the more analytical and critical literature on DL. With that as the backdrop, we also offer here a framework, developing the ideas of Leithwood et al. (2007) and Iles and Feng (2011), to stimulate further debate and to indicate areas of further research. We propose to plot the development of thinking on DL along two dimensions, as indicated in Figure 1.

Figure 1.

Dimensions of distributed leadership. (Adapted from Leithwood et al. (2007)) and Iles and Feng (2011))

The first of these dimensions is the continuum between what we call ‘planned’ activity at one end to ‘emergent’ activity at the other, similar to the approach taken in other areas of management (e.g. planned and emergent models of organizational change –Burnes 2009). It represents the variation of possibilities of the two forms of DL presented by Gibb (1954, 1968). The first is a simple recognition that each person can commit an act of leadership. The sum of leadership in a unit such as a group/team or organization is the aggregation of such acts. It recognizes that there may be a nominated leader but that the overall leadership effort is shared through the accumulation of contributions. As noted earlier, Gronn (2002) suggests that aggregation is a ‘minimalist’ version of DL, but that the sharing of responsibility for performance is attractive to those in leadership positions. To this extent, its distribution can be formally structured and planned.

The second form of DL presented by Gibb recognized emergent and holistic patterns of leadership, including that of a conjoint agency, occurring in a variety of multi-member group settings: for example, in groups there was a ‘tendency for leadership to pass from one individual to another as the situation changes’ (Gibb 1954, p. 902). This was later developed into a critique of the mainstream view of a separation between leaders and followers, with the finding that ‘leaders and followers frequently exchange roles and observation has shown that the most active followers often initiate acts of leading’ (Gibb 1968, p. 252). Through interaction between different participants, the practice of leadership becomes stretched across a variety of contexts (Spillane et al. 2001). There can be little planning for such practices, and there has to be some degree of good fortune arising spontaneously with appreciation of the possibilities, through tacit understanding, intuition and trust between the parties (Gabarro 1987). Thus, our first axis is planned–emergent.

We can lay this axis across another, adapted from Leithwood et al. (2007) and Harris (2006), which we call the continuum between ‘aligned’ activity (where people's/divisions' activities are already seen as having some common aspects, provided in some cases by organizational structure and strategy) and ‘non-aligned’ (where people may be unaware of or unintentionally ignore the activities of others and the potential benefits of sharing aims and interests). This variation also connects to the suggestion by Gronn (2000) and Spillane et al. (2004) that the theoretical roots of DL lie in the field of distributed cognition and activity theory, particularly the version presented by Engestrőm (1987), where the performance of tasks is (to a greater or lesser extent) aligned with the overall purpose or object of the collective activity of the organization. Alignment adds unity and coherence to the work carried out, leading to definable outcomes; but individuals, groups and teams, as they complete particular actions, may (to a greater or lesser extent) align their work with the object, hence the dimension varying between alignment and non-alignment.

We can name these quadrants for discussion purposes. Thus, the aligned–planned quadrant we call ‘Classical DL’; the misaligned–planned is ‘Mis-planned DL’; the emergent–aligned ‘Emergent DL’ and the misaligned–emergent ‘Chaotic DL’.

Thus, we have four quadrants to provide a framework for discussions of DL. The top-left quadrant, Classical DL, will be familiar to most of us, as this is the quadrant where classical views of management might be located: the manager/leader has a clear focus, powerful control and existing structural arrangements through which to operate. Here, there is an intention to develop patterns of DL within the organization, working with normative ideas of DL practice as indicated by Harris (2006), Spillane (2006) and various toolboxes of skills and techniques (McBeth 2008). In this process, sufficient attention has been given to the organization's strategy, structure and culture, together with recruitment and development of appropriate staff. Thus, DL can operate effectively in supporting the achievement of organizational goals. The research agenda informed by this quadrant includes examination of the understanding of DL in a particular organizational context, the criteria for organizational performance in relation to this, and the process that aids the effective implementation of DL, providing case examples of good practice, which can be replicated by others. For example, a CEO who espouses ‘leadership for all’ might talk of empowerment of teams, local decision-making and communication. However, the traditional hierarchy remains, and such teams, however ostensibly autonomous, continue to defer to single individuals. This is the quadrant of DL as the distributing leader. We represent this as the classical organizational triangle where a top-down approach to DL can be planned and brought into operation. However, as Harris (2008) identified, it is insufficient, although quite possibly effective in times of crisis (Grint 2005b). It is in the other quadrants where further examination might be more interesting and more fruitful, both practically and academically.

The top-right quadrant, Mis-planned DL, illustrates organizations where there is the intention to use DL, but where either structures become inappropriate for the purpose of the organization, or people within the structures remain ambivalent about such a move, preferring to set their own goals within local units. A failure to establish effective coordination and consensus between tasks and functions produces less than optimal results. The leadership concern is with inappropriate organization forms, designs and structures, but also with a failure to persuade or inspire others. For example, despite intentions to distribute leadership, people tend to look up to find their leader at the apex of their organization (MacBeath 1998) and, in so doing, position themselves as inferior (Watzlawick et al. 1967). Cultural and historical context supports the definition of a relationship, perhaps contested, between staff and appointed leaders. Distributed leadership is seen as just another form or fad. Research topics suggested by this quadrant would include identifying and examining factors that hindered the effective development of DL, and the nature of the misalignment: for example, whether it is the result of an inappropriate structure, inadequate coordination mechanisms, tensions between organizational and divisional goals or lack of appropriate communication systems or managerial skills such as the disjuncture between what leaders say and what they do.

The bottom-left quadrant, Emergent DL, recognizes the realities of day-to-day organizational life where the everyday cannot be designed with certainty, and where contradictions and disturbances to work occur (Blackler et al. 1999), often spontaneously but also informally, beyond the sight of managers in senior positions. Gronn (2009, p. 383) points to the varied pattern or mix of leadership configurations that can arise; these can include individuals but also pairs, teams, groups and networks, all of which encompass practices and situations where the various types of ‘hybrid’ distribution will occur. With careful attunement to the flow of events, an atmosphere of trust and support aligns structure with cultures to provide an underpinning direction for high performance. For those appointed to leader positions, there is an understanding of DL and also a desire to encourage participation and engagement. In some cases, there is a rejection of leadership responsibilities, or long-held values and reinforced norms that protect particular interests and preserve valued identities that resist unification. The conundrum which exists when organizations are regarded as traditional hierarchies disappears as the reality emerges that no one individual can understand, account for or take direct control of activity as it emerges day by day. There is a link here also to the way in which research has so often been conducted on leadership. This is characterized by a focus on individuals and the pursuit of rational universal truths. Few studies, as we note later, have been conducted using social constructionist methodologies and perspectives that take account of a variety of organizational stakeholders, in an attempt to see the world through the eyes of groups rather than individuals. To understand the workings of leadership and to improve understanding of the ways in which DL might be taking place within organizations, not only does the way in which we conduct research need to change, but also the focus of what we research needs to be adjusted. For example, research needs to focus on human and social capital development, and on those appointed in the process (Day 2000). There is also space for finely tuned case studies of different configurations of leadership, requiring a multi-voiced and multi-layered approach where influence can be exerted by anyone present, at any time.

The bottom-right quadrant, Chaotic DL, illustrates a situation where elements of DL may develop locally but in relatively haphazard ways, with a focus on local contexts and goals and without sufficient attention to operations in other parts of the organization. The result is a somewhat anarchic process of DL – it may appear and even be effective at the local level, but only in the short term, and it does not complement (and indeed may interfere with) organizational functioning and the development of DL at the wider level. At the level of teams within networks, there is apparent coherence within each team, but information and expertise remain local and unavailable for sharing (Friedrich et al. 2009). Anarchy may spread in the sense of outright rejection of DL, perceived as more shedding of responsibility by those who are paid to bear it. Research issues developing from this quadrant include an examination of different elements of the organizational system, decision-making and communication systems, and the process of engagement and participation at all levels within an organization and across boundaries between organizations and networks.

The diagram has a normative element in that, for DL to be effective within any organizational context, it needs to be seen as occurring in the left-hand, aligned quadrants. So to some extent it provides a guide as to how DL might be developing in any one context, as well as a means of structuring an examination of (and helping diagnose problems in) that development. To that degree, we see the use of these quadrants as a means of getting behind the rhetoric of DL to examine its development in practical settings. In that sense, it may have application both for organizational practitioners – managers and staff as they engage with DL – and also for academics, as a framework for a deeper examination of DL. It is important to emphasize, though, that we have laid out what might appear to be a static framework, enabling categorization of DL and helping to structure and focus future research. Nevertheless, the framework can be seen to represent the development of DL in a more dynamic way. Planned/aligned approaches can become misaligned over time. Similarly, apparently emergent/misaligned approaches may, over time and with adaptation, become more aligned. Returning to Gronn's (2002) forms of concerted action noted earlier (of spontaneous collaboration, intuitive working relations and institutionalized practice), we might see the first two of these as emergent, while the third is planned. However, successful emergent/aligned DL practice may become institutionalized over time. Thus, DL in practice is not necessarily static and should be expected to change and develop in response to learning within the organization and to changing influences in the organization's environment. It is important therefore for those involved in future empirical research into, and theorizing about, DL to recognize and to explore its dynamic nature. Furthermore, a recognition of the dynamic nature of DL allows account to be taken of important contextual factors – cultural, historical and social – and their influence on the development of DL in any particular context.

The framework is intended to provide a means of locating different approaches and themes in the development of thinking about DL and to help develop an informed critique of those approaches and themes. As we move on to the introduction to the papers in this special issue, we emphasize the perspectives that an examination of DL permits: namely, that views of leadership which consider the individual leader as the prime ‘unit of analysis’ are restrictive and do not recognize fully the dynamics and interrelationships involved in the leadership process. Nor do they recognize the potential contribution of all those involved in that process. The development of DL enables us to consider in greater detail the relationship aspects of leadership and the changing contexts within which leadership occurs, with its various social and cultural influences and interpretations. With these themes in mind, we turn now to an introduction to the papers which comprise this issue.

Introducing the papers

The papers that constitute this special issue have certain common elements, not least the recognition of the descriptive or normative approaches that typify the developing literature on DL. Each paper has its own particular focus, critique and indications of where further energies might be expended. Considered together, they indicate a number of themes which merit further attention and discussion if understanding and application of DL is to be expanded. Many of the papers argue the importance of context. Most recognize that existing empirical work is largely to be found in the context of education. Consideration could also usefully be given to examining DL in other contexts of public service and beyond, in more commercial environments.

Several contributors (Bolden; Cope et al.) see DL as an emergent phenomenon, noting that it might be most usefully viewed in its different configurations. Further consideration might usefully be given to the spread of DL across and among different groupings, from DL within dyads to more extensive DL across and beyond the boundaries of formal organizations. Several papers emphasize the growing recognition of DL and leadership more generally, as a relational process. Context again is a key issue here, given the unique ways and locations in which relationships develop, and the particular and unique leadership challenges faced according to the situation. Edwards, in particular, highlights the possibilities of considering DL within a broader concept of ‘community’, and the different ways in which this concept itself might be interpreted and applied. Fitzsimons et al. develop the relational aspect of DL in some depth, proposing further theoretical development to help clarify the often muddied waters – and to provide a basis for the research and practice – of DL. On the basis of the above contributions, the dynamics of work relations and of organization communities would prove a useful theme for further research and theory development.

The first paper, by Richard Bolden, is well placed to lead out this special issue, reviewing as it does the academic and empirical literature on the concept of DL, identifying its origins, key arguments and areas for further work. Consideration is given to the similarities and differences between DL and related concepts, including ‘shared’, ‘collective’, ‘collaborative’, ‘emergent’, ‘co-’ and ‘democratic’ leadership. Findings indicate that, while there may be some common theoretical bases, the relative usage of these concepts varies over time, between countries and between sectors. Of particular note is the way in which DL is shown to have attracted a rapid growth in interest since 2000, but research on the topic has largely remained restricted to the field of school education, more so in UK-based than in US-based academics. It reveals how several scholars emphasize the fact that, in order to be ‘distributed’, leadership need not necessarily be widely ‘shared’ or ‘democratic’; however, in order to be effective, there is a need to balance different ‘hybrid’ configurations of practice. The paper highlights a number of areas for further attention, including three factors that relate to the context of much current work on DL power and politics, boundaries and ethics and diversity. Three methodological and developmental challenges are raised, relating to ontology, research methods and leadership development, and reward and recognition. Bolden concludes by making a plea that normative perspectives dominating the literature need to be supported by more critical perspectives and by those recognizing the rhetorical significance of DL in (re)constructing leader identities and mobilizing collective engagement.

The second paper, by Jason Cope, Stephen Kempster and Ken Parry, focuses on SMEs, particularly the manner in which entrepreneurial ventures are effectively led by small teams rather than by individuals. Notwithstanding, the extant literature on SME leadership indicates a relational process characterized as leadership by heroic individuals, and (more recently) by DL within a team context. Implicit theories of leadership tend to explain much of the expectation of heroic individual leadership in the SME context. The overlap between this style of leadership and distributed team leadership is that of ‘blended’ leadership. It is contended that authentic entrepreneurial leadership is best examined through the lens of blended leadership, and that blended leadership is effective at building a successful entrepreneurial team. A future research direction is posited.

The third paper, by Graeme Currie and Andy Lockett, examines the concept of DL and its enactment in a quite different context: health and social care. The paper begins with an analysis of the literature relating to DL, and in doing so synthesizes the different conceptualizations of DL, employing Gronn's (2002) dimensions of concertive action and conjoint agency. It then moves on to examine why governments have promoted DL in the public sector, arguing that they have done so as a means of reviving poorly performing public-service organizations. The authors then narrow their focus to examine the context of health and social care and, in so doing, argue that the complexity of professional and policy institutions often makes it difficult to enact a strong (or pure) form of DL, as conceived by Gronn (2002). This paper shows how leadership in health and social care contexts creates a paradox for the operationalization of DL.

The fourth paper, by Gareth Edwards, continues this public-sector theme by reviewing the literature on community and relating various concepts of community to DL. In the paper, he suggests that the investigation and theoretical discussion of DL with these concepts of community marks out the potential for developing a more context-rich understanding of leadership in organizations and society. The paper begins by reviewing the literature on DL, noting a shift from clichéd ideals to more structured frameworks. It then highlights the need to further contextualize notions of DL, before going on to highlight a number of concepts relating to community which are useful for theoretical reflection, research and practice. These concepts are symbolism, a sense of belonging, a sense of community, individualism, values and ethics, language, dialect and discourse, liminality and friendship. The paper also develops a discussion on postmodern views of community and the importance of recognizing multiple belonging and multiple identities. From the review, the paper develops ten propositions for theory, research and practice regarding DL and concepts of community. It also highlights practical implications from the review by discussing the importance of considering these in management and leadership development initiatives, and in managerial practice in general.

The final paper, by Declan Fitzsimons, Kim Turnbill James and David Denyer, argues for the importance and distinctiveness of shared leadership and of DL as a field of study. It offers a historical review and thematic synthesis, and aims to identify, evaluate and interpret prior research in this field, inductively categorizing the seminal theoretical and empirical contributions. The review reveals a field which lacks cohesion and has no single agreed ontological or epistemological paradigm; it then explores the theoretical antecedents of the field, highlighting three distinct strands of literature emanating from different research traditions. From this literature, the paper highlights the nuances, ambiguities and contradictions of shared and distributed leadership. Two distinct conceptions of shared and distributed leadership are identified from this synthesis (relational–entity and relational–processual), and a third possible model (relational–systemic) is developed after considering the limitations of the first two. In this way, the paper raises fundamental issues about how we should think about leadership in terms of research, practice and development.