1) Define discursive leadership. Describe two aspects of Chrislip & O’Malley’s leadership for the common good that aligns with a discursive approach to leadership. (250 word max) 2) How does Actor Network Theory help account for non-human dimensions of leadership? (250 word max) 3) Member categorization or the disciplinary power of Discourse, what is a more helpful method to understand how leadership was exercised (or not) on the issue racism on campus this semester? (500 word max) 4) According to Fairhurst, the individual leaders and follower are no longer the most important units of analysis. If you were going to study leadership in the Tuesday Alabama Senate race, what would you analyze? In other words, what is a discursive unit of analysis that would be helpful to understand how leadership played a role in Doug Jones being elected to the Senate? (500 word max) 5.) You arrive at restaurant in the private room on Thursday evening at 7pm for the course celebration. When you enter the room, you notice that all of your classmates are sitting silently. You also notice 4 other people are in seated the room. Your instructor is nowhere to be found. Upon taking your seat, the person sitting at the head of table says: “I’m Ron Heifetz, here with my colleagues Gail Fairhurst, Ed O’Malley, and David Chrislip. We understand that you have been discussing leadership communication for a few months. My guess is there’s still much for us to learn about exercising leadership. So…where should we begin?” 21 seconds of silence later someone in the class clears their throat and intervenes. (700 word max) Directions: Create a conversation between the class and four authors that surfaces the most important leadership concept/practice from our course readings. Guidelines: Indicate by name who is talking throughout the dialogue Include all four authors in the conversation and at a minimum four people in the class Make sure it is clear why the leadership concept is the most important (not everyone in the conversations needs to agree) Table of Contents Preface .......................................................................................................................................................... 3 1 Two Traditions ........................................................................................................................................... 5 Defining Leadership .................................................................................................................................. 7 Defining Discourse .................................................................................................................................... 8 The Case for Discursive Leadership .......................................................................................................... 9 DISCOURSE VERSUS MENTAL THEATER .......................................................................................... 9 DECENTERED SUBJECTS/THIN ACTORS VERSUS ESSENCES........................................................... 10 ENCOMPASSING VERSUS DUALISTIC CONCEPTIONS OF POWER AND INFLUENCE ...................... 11 REFLEXIVE AGENCY VERSUS UNTHEORIZED/EXAGGERATED AGENCY ......................................... 12 TEXTUAL, CON-TEXTUAL VERSUS THE VARIABLE ANALYTIC ......................................................... 13 The Path Forward.................................................................................................................................... 15 NOTES...................................................................................................................................................... 16 2 Sequence and Temporal Form ................................................................................................................. 18 Act, Interact, Double Interact ................................................................................................................. 19 Turn Taking, Adjacency Pairs .................................................................................................................. 21 Narrative Schemas, Episodes .................................................................................................................. 24 Scripts...................................................................................................................................................... 27 Script Formulations ................................................................................................................................. 29 A Backward Glance-Final Thoughts......................................................................................................... 31 NOTES...................................................................................................................................................... 32 3 Membership Categorization .................................................................................................................... 33 Membership Categorization Defined...................................................................................................... 34 Categories and Organizational Coordination .......................................................................................... 34 Categories and Organizational Role/Identity.......................................................................................... 35 Categories, Sensemaking, and Meaning Management .......................................................................... 37 Categories and Social Structuring ........................................................................................................... 40 Categories in Task Structuring ................................................................................................................ 41 A Backward Glance-Final Thoughts......................................................................................................... 44 NOTES...................................................................................................................................................... 45 4 Disciplinary Power.................................................................................................................................... 45 A Primer on Foucault .............................................................................................................................. 46 Discipline and Surveillance in Performance Management Technologies ............................................... 50 PERFORMANCE APPRAISAL ........................................................................................................... 50 360-DEGREE FEEDBACK ................................................................................................................ 51 EXECUTIVE COACHING .................................................................................................................. 51 Regulating Executive Coaching 'Psychological' Discourses ........................................................... 52 Targeted Objects of Coaching Discourses: .................................................................................... 53 Performance Management Governmentality......................................................................................... 54 A Backward Glance-Final Thoughts......................................................................................................... 55 NOTES...................................................................................................................................................... 56 5 Self-Identities, Interpretative Repertoires ............................................................................................... 57 The 'Self' in Leadership Psychology and Discursive Leadership.............................................................. 57 Interpretative Repertoires and Subject Positioning ............................................................................... 63 A Backward Glance-Final Thoughts......................................................................................................... 66 NOTES...................................................................................................................................................... 68 6 Narrative Logics........................................................................................................................................ 68 Leader-Member Exchange Theory .......................................................................................................... 68 The Narrative Basis of LMX ..................................................................................................................... 69 NARRATIVES AND STORIES: CONSTRUCTING LMX ....................................................................... 72 NARRATIVE RESOURCES FOR THE LMX ......................................................................................... 74 A Backward Glance-Final Thoughts......................................................................................................... 76 NOTES...................................................................................................................................................... 77 7 Material Mediations................................................................................................................................. 77 Materialist Critiques of Discourse Analysis ............................................................................................. 78 Actor-Network Theory ............................................................................................................................ 79 Rudy Giuliani and September 11, 2001 .................................................................................................. 82 BACKGROUND AND TEXT .............................................................................................................. 83 GIULIANI MACROACTED AT A TIME OF GREAT PERSONAL RISK ................................................... 84 IN HYBRID FASHION, GIULIANI ASSOCIATED HIMSELF WITH THE CHARISMATIC CENTRAL AND SACRED.......................................................................................................................................... 85 GIULIAN I'S AUTHORING AND NETWORKING CONSTRUCTED AN EMOTIONAL SCAFFOLDING ... 87 GIULIANl'S CHARISMA BENEFITED ENORMOUSLY FROM IMMEDIATE SOCIAL SHAPING ANT's .. 87 A Backward Glance-Final Thoughts......................................................................................................... 89 NOTES...................................................................................................................................................... 89 8 Praxis and More Conversation ................................................................................................................. 90 Discursive Leadership and Praxis ............................................................................................................ 90 Discursive Leadership and Leadership Psychology ................................................................................. 94 Discursive Scholars ........................................................................................................................ 99 Preface T his much I know about leadership: It is a topic for the ages. Discussions of leadership date back to Plato and the early Greeks, but also Chinese and Egyptian societies. It was a topic during the Renaissance with Machiavelli's The Prince, which survives as a reference today. Proceeding onward through the turn of the twentieth century, it emerged in 'great man' theories, marking the start of serious scholarship that continues to the present. Such scholarship now joins a business press eager to dispense sage advice to hungry leaders. What is it about leadership that sustains this kind of interest? Bass ( 1981) asserted that leadership is a universal human phenomenon, the templates for which are supplied by parenthood. If true, it should be no surprise that we find leadership in a host of society's collectives-business and governmental organizations to be sure, but also remote African villages, sports teams, and Girl Scout troops. Few agree on a definition. Leadership scholars are famous for their inability to agree on a definition of leadership, leading some analysts to remark that there are as many definitions as there are leadership scholars (Bass, 1981; Fiedler, 1971; Rost, 1991). However, there are good reasons for this inconsistency. Leadership occurs amidst a tremendous amount of situational variability, and it has that elusive 'eye of the beholder' quality. Some will make sense of complex conditions by arriving at an attribution of leadership that others would vehemently contest (think George W. Bush, the 43rd president of the United States, and his handling of the war in Iraq). Yet, Meindl (1995) suggests that our attributions are romanticized in this regard because too often we see leadership as the cause of organizational success or failure when a more complex explanation is in order. Even so, one person's leadership is another's tyranny or ineptitude. Organizational leadership was once the sole province of men. While the concept of leadership has been around for some time, the serious study of leadership is about 100 years old. As mentioned, it began with the turn of the twentieth century 'great man' school of thought, which led social scientists to look for those characteristics and traits (such as intelligence, dominance, height, and so forth) that differentiated leaders from non-leaders. Organizational leadership was considered the sole province of men until women began to enter the workforce in large numbers in the 1970s in other than low power positions (Kanter, 1977). Since then, gender differences in leadership have ranked among the hot topics in both the academic and business press as well as in countless discussions at watercoolers and boardrooms in organizations worldwide (Buzzanell, 2000; Collinson, 1988; Kanter, 1977; Reardon, 1995). Leadership psychologists have supplied important foundational work in leadership studies. 1 Their early trait theories gave way to the study of leader behavior styles, famously captured in the Ohio State leadership studies, which examined initiating structure and consideration as two dimensions of leader behavior (Stogdill & Coons, 1957). Contingency theories followed, such as Fiedler's (1971) emphasizing leader-member relations, task structure, and a leader's position power as determinants of the type of leader effectiveness. Leader-member exchange (LMX) theory subsequently adopted an exclusive relational focus, where high versus low quality leader-member relationships differed in terms of the resources exchanged and outcomes delivered (Dansereau, Graen, & Haga, 1975; Graen & Scandura, 1987). At about the same time, neocharismatic leadership theories arrived on the scene, emphasizing leaders' charisma, vision, and the ability to inspire followers well beyond the terms of their employment contract (Bass, 1985; Conger, 1989; Shamir, House, & Arthur, 1993). Also, the information-processing school of leadership began to study implicit leadership theories and the role of cognition in the enactment and attribution of leadership behavior (Hanges, Lord, & Dickson, 2000; Lord & Maher, 1991). LMX, neo-charisma, and implicit leadership theories continue to this day, as authentic leadership (Gardner, Avolio, Luthans, May, & Walumbwa, 2005), spiritual leadership (Reave, 2005), and leadership in teambased organizations (Day, Gronn, & Salas, 2006) assume the newcomer roles. As good scientists, leadership psychologists have challenged their own theories, methods, and findings over time (House & Aditya, 1997; Lowe & Gardner, 2000). Interestingly, much of the criticism points to the socially constructed nature of leadership (Calder, 1977; Lord & Brown, 2004; Meindl, 1995), a perspective that, if taken seriously, has the potential to both challenge and complement leadership psychology at a foundational level. I do not mean to imply that psychologists are uninterested or unwilling to pursue a socially constructed view of leadership, nor do I wish to diminish their contributions to this topic in any way. I only wish to observe that their concerns for the individual and psychological rather consistently outweigh their concerns for the social and cultural. I argue that both sets of concerns must be entertained in equal strengths in order to understand a socially constructed world. Thankfully, a body of theory and research directly applicable to the social, linguistic, and cultural aspects of leadership has been accumulating. I call this work discursive leadership because of its focus on organizational discourse, both as language use in social interaction and the view of Discourse made popular by Michel Foucault. In his view, Discourse is a system of thought and a way of talking about a subject that together supplies the necessary linguistic resources for communicating actors. Foucault's work is typical of the burgeoning organizational discourse literature that reflects a body of constructionist theories not specifically about leadership per se, but with great potential to illuminate it in ways that we have not yet seen. That potential motivates the writing of this book, which is less a literature review and more of an exploration of key discourse concepts and what they could mean for leadership. The voluminous research from leadership psychology serves as a useful point of contrast, springboard, and benchmark along the way. There is still much to learn about leadership, especially if we surrender to its protean tendencies. As Chapter 1 makes clear, discursive leadership and leadership psychology differ on both ontological and epistemological grounds. In a nutshell, leadership psychology has been on a quest to understand the essence of leadership, whether it be found in the individual leader, the situation, or some combination thereof (Grint, 2000). By contrast, discursive leadership rejects essences because leadership is an attribution and, very likely, a contested one at that. Discourse scholars like me depart from leadership psychologists' adherence to traditional science assumptions about realist conceptions of truth and representationalist views of knowledge. Influenced by the linguistic turn in philosophy, we ask instead that both perspectives be seen as alternative ways of knowing, talking about, and justifying leadership (Deetz, 1996; Rorty, 1982). By recognizing discursive leadership from this vantage, we have a means by which to embrace what leadership psychologists might see as the elusive, unwieldy, mutable, and maddening error variance in leadershipin short, its protean tendencies. I am certainly not claiming that discursive leadership has all of the answers to leadership's mysteries, but neither do I believe that discursive leadership is just one more approach to leadership. It represents instead a foundation for many new lines of research into leadership with potentially important implications for helping practicing leaders and others better understand how they coconstruct reality. It also represents an opportunity for new dialogue with leadership psychologists-a dialogue that I hope continues long after this book. There are several leadership psychologists who have been gracious enough to help me begin this dialogue in Chapter 8, the book's final chapter. They include Donna Chrobot-Mason, Steve Green, Jerry Hunt, Robert Liden, and Boas Shamir. Three discursive scholars, Kevin Barge, Franc;:ois Cooren, and Linda Putnam, also joined in. To all of them I am grateful for the effort that they put forth under a very tight deadline. I hope that the reader finds their comments as illuminating as I did. x DISCURSIVE LEADERSHIP Some of the reviewers for this book suggested different ways in which it might be read that I found quite useful. For example, if one prefers to start out with the details supplied through language and interaction ( what I call little 'd' discourse), the chapters should be read in chronological order. However, others may prefer to start with the generalities associated with a Foucauldian view of Discourse (big 'D' Discourse) as a system of thought and way of talking about a subject. In that case, I would recommend reading Chapters 4 and 5 before Chapters 2 and 3. As the book reviewers also noted, the potential readers for this book will have varying levels of familiarity with the different forms of discourse analysis. Thus, I have included a set of appendixes organized by type of discourse analysis. They are designed for quick and easy reference. Finally, except for interview discourse, the transcribed interaction in this text follows the conventions of conversation analysis (see Appendix Al). For those who do not appreciate the level of detail this provides, readers may simply skim over the detailed markings. Those who have read all or parts of my book along the way include Carey Adams, Kevin Barge, Mary Helen Brown, Mary Ann Danielson, Jennifer Butler Ellis, David Hoffman, Fred Jablin, Robert Liden, Patricia Parker, Paaige Turner, Patricia Witherspoon, and Ted Zorn. T hank you for the time and effort that you put into reviewing my work. I especially want to thank those scholars who gave me direct feedback. T hey include Kevin Barge, Suzanne Boys, Lisa Fisher, Angela Garcia, Donna Chrobot-Mason, Frarn;:ois Cooren, Rich Kiley, Linda Putnam, Edna Rogers, Marcia Schoeni, Mathew Sheep, James Taylor, and Heather Zoller. You have shaped my thinking in ways too numerous to mention. T hanks also to Jan Svennevig and Maria Isaksson of the Norwegian School of Management BI and organizers of the 2006 Association for Business Communication European Conference in Oslo, Norway; Pam Shockley, organizer of the 2005 Aspen Conference on Organizational Communication; and Angela Garcia, director of the Workplace Studies Group at the University of Cincinnati for providing forums for the presentation and discussion of my work. T hanks also to my department head, Teresa Sabourin, for her continued support and friendship; my graduate students, Justin Combs, Zhou Fan, Stephanie Hamlett, Elizabeth Prebles, Kim Richardson, and Brian Singson for their diligent work on my behalf; and to Sadie Oliver and Priscilla Ball for all of their help and office support. T hank you to Verne, Katie, Tom, and Kelsey, each of whom has a wonderful way of helping me to maintain perspective throughout this effort. I feel blessed every day for their love and support. To Todd Armstrong and Sarah , Quesenberry at Sage, thank you for your patience and expert guidance. To Teresa Hedinger and Libby Larson, I greatly appreciated the care that you showed toward my manuscript. Finally, I have been blessed with so many wonderful colleagues, including the late Fred Jablin whose work as a leadership and communication scholar remains forever with me as a standard of excellence. NOTE 1. Scholars from political communication (Hart, 1984, 1987; Trent, 1978; Trent & Friedenberg, 2004), political science (Burns, 1978), educational administration ( Gronn, 1982, 1983), and organizational development (Kets de Vries, 1990a, 1991, 2005) among others have also made important contributions to leadership study. However, the broadest comparison appears between leadership psychology and discursive leadership. Where relevant, work from these related fields is introduced into individual chapters. 1 Two Traditions S everal years ago, Henry Mintzberg (1970, 1973, 1975) was at the center of a debate in leadership studies when he turned the spotlight on what leaders actually do and the nature of managerial work itself. Although he was certainly not the first to do this type of research, 1 his work attracted a great deal of attention because he challenged the conventional wisdom that the manager's job was to plan, organize, coordinate, and control. First introduced by French industrialist Henry Fayol in 1916, these functions became almost passe as a result of Mintzberg's behavioral observations of five chief executives (Hunt, 1991). Mintzberg (1975) argued compellingly that Fayol's functions were just folklore because managerial work is in reality too d ynamic, fragmented, and unsystematic. Managers work frenetically in short bursts of time as they react to job demands and constant interruptions. From this research, Mintzberg created his well-known taxonomy of managerial roles.2 Applauded by many for his realism, Mintzberg certainly had his critics. Carroll and Gillen (1987) defended the classical functions and argued that Fayol's ideas would have been supported had Mintzberg asked for the reasons for the managers' observable behavior. Further, they argued that Mintzberg's observation approach was fundamentally flawed because "Managerial work is really mental work and the observable behaviors such as talking, reading, and writing serve as inputs and outputs to neuropsychological activities" (p. 43, emphasis added). Thus, observable behavior is not a reliable measure of what managers actually do. On the surface, this looked like a debate over whether to characterize what managers do as abstract functions or specific behaviors. Interestingly, Carroll and Gillen (1987) found a middle ground on this issue by urging researchers to consider the unsystematic ways in which management's classic functions may be achieved-for example, how planning, organizing, coordinating, or controlling occurs through unplanned, informal, and brief conversations. However, a deeper conflict was apparent in Carroll and Gillen's apparent need 2 DISCURSIVE LEADERSHIP to make "mental work" or "neuropsychological activities" the central and defining feature of managerial work. Observable behavior like talking, writing, and reading was then downgraded to simple inputs and outputs. Why couldn't mental work and social processes like talking both be of equal import? One reason may have been that Mintzberg advocated a radical approach to research that was everything mainstream leadership research was not at the time. Mintzberg (1982) urged his colleagues to get rid of their constructs before they collected data, throw away their questionnaires and 7-point scales, stop pretending the world is divided into dependent and independent variables, and do away with "artificial rigor, detached rigor, rigor not for insight, but for its own sake" (p. 254). Although he allowed that he may have been overstating his recommendations a bit, he felt strongly that leadership needed to be studied simply, directly, and imaginatively, and that traditional empiricist approaches were not getting the job done. Nevertheless, his were fighting words, words that can begin paradigm wars, although Mintzberg (1982) seemed only to be calling for a methodological overhaul. However, to take Mintzberg seriously, one had to acknowledge that behavior was worthy of study in its own right. His argument had implications both in terms of what leadership scholars studied and how they studied it. 3 Fast forward, if you will, to twenty-first century leadership study. At first blush, little seems to have changed-especially in the United States where a psychological lens and traditional empiricist methods still dominate (Alvesson & Sveningsson, 20036; Conger, 1998; Knights & Willmott, 1992). However, it would be a clear mistake to suggest that the legions of leadership scholars with psychology backgrounds are unconcerned with behavior. True, their first concern is with its cognitive or social-cognitive origins and the perceptions they generate; the weight given to the mental over the behavioral in the Carroll and Gillen quote is testimony to this. However, as leaders increasingly get depicted as 'managers of meaning' (Pondy, 1978; Shatter & Cunliffe, 2003; Smircich & Morgan, 1982), the style, content, and delivery of their message or 'visions' have been the subject of scrutiny in ways that Mintzberg, circa 1970s, might have welcomed (Emrich, Brower, Feldman, & Garland, 2001; Fiol, Harris, & House, 1999; Shamir, Arthur, & House, 1994). In addition, work by Komaki (1998) and Gioia and Sims (1986) has examined the impact of leader verbal behavior on employee performance, narrative has gained a foothold in leadership studies (Conger, 1991; Gardner & Avolio, 1998; Shamir & Eilam, 2005), and qualitative leadership research in general continues to be on the rise (Alvesson & Sveningsson, 20036; Bryman, 2004). However, another force was afoot to answer Mintzberg's call. Spurred on by the 1960s and 1970s critiques of traditional scientific canons such as realist conceptions of truth and representational theories of knowledge, the linguistic turn in philosophy affected scholars in such disciplines as communication, Chapter 1: Two Traditions 3 sociology, psychology, and European schools of management (Alvesson & Karreman, 2000a; Bochner, 1985; Deetz, 1992).4 Skirmishes waged and won in the back alleys of journal publication and other scholarly venues have been producing a body of scholarship relevant to leadership that is clearly outside mainstream leadership psychology. Heavily oriented toward discourse and communication, I have termed this scholarship discursive leadership. Grint's (1997, 2000) work on the paradoxes of leadership is a good example. Through in-depth case history analyses of several political and organizational leaders, he finds that leadership is more inventive than analytic. The mainstream leadership literature often suggests the opposite because its primary lens is individual and cognitive. It is less focused on the contested nature of leadership interaction, thus it may undervalue those creative aspects that explain why leadership is more art than science. Grint also finds that reason and rationality do not carry the day as much as persuasion does, and the study of leadership is too often rooted in irony rather than truth. He sees this in the collective identities upon which much leadership rests, which are not 'reflected' in empiricist data as much as they are 'forged' amidst challenge and conflict. Outcomes are far less predictable as a result, despite a literature body whose writers (particularly in the business press) often confidently proclaim the opposite. Thus, who may we say is better positioned to answer Mintzberg's call to study leadership simply, directly, and imaginatively-latter-day leadership psychologists or discursive leadership scholars like Grint? The answer may surprise: neither alone, and both in different ways. Given the variety of organizational discourse approaches available today and the cross-paradigmatic thinking some are generating, these approaches have the potentialmuch like Mintzberg's work-to challenge, inform, and complement the still-dominant psychological approaches upon which so much leadership research is based. With this view in mind, two deceptively simple questions guide this book. First, what do we see, think, and talk about with a discursive lens directed toward leadership? Second, what leadership knowledge is to be gained in the interplay between a discursive lens and one that is psychological? In posing these questions, I have no interest in debating whether discursive leadership or leadership psychology is the better overall lens. There is never only one conceptual or paradigmatic framework sufficient for answering all questions about leadership,5 and I would argue that it is wrong for any perspective to overestimate its influence at the expense of the other.6 In making the case for discursive leadership, the substantial contributions made by leadership psychology to our understanding of leadership should in no way be underestimated. My bias is a discursive one, yet my intent in finding fault with leadership psychology at times is never to forsake it. Complex social phenomena, like leadership, have many parts that act together and define one another 4 DISCURSIVE LEADERSHIP to form an entwined whole, although such interdependence may not be readily apparent. This orientation reflects Albert et al.'s (1986) notion of complementary holism, the goal of which is to provide more holistic social theory through intellectual frameworks "specifically contoured to understanding an interconnected reality" (p. 15).7 I do not know how much holism is possible between discursive leadership and leadership psychology; I do know that such a goal is impossible without more conversation between them. As Rorty (1979) suggested, conversation across diverse theories and frameworks is "the ultimate context within which knowledge is to be understood" (p. 389). Thus, this book's purpose-to put some contours around what is discursive leadershipis aided, in part, by the possibilities for its relationship with leadership psychology. Neither discursive leadership nor leadership psychology should be seen as derivative of the other; they are simply alternative, coconstructing lenses with both strengths and shortcomings.8 To begin this conversation then, it is necessary to map some of the contested terrain over definitions of leadership and discourse. Defining Leadership Any definition of leadership ultimately rests on one's ontological commitments. As such, most of the discursive approaches in this book, in varying degrees, meet the conditions of a broadly constructionist stance as outlined by Hacking (1999). Critical of the status quo, they argue for social construction precisely when leadership is taken for granted and appears inevitable. For example, consider the current interest in authentic leadership (Avolio & Gardner, 2005).9 That there even is such a phenomenon as authentic leadership appears inevitable once you have actor or analyst claims about specific leaders' facades (for example, those of many politicians) or others' genuine or true selves (for example, those of a Gandhi, a Warren Buffet, or an Oprah W infrey). Yet, as Chapter 5 will reveal, a discursive approach rooted in Foucault strikes down this inevitability because authenticity is equated with virtuosity by those influenced by positive psychology and, opposingly, the revelation of one's dark side by those subscribing to the traditional pathology model of psychology. These diametrically opposed conceptions of authentic leadership suggest social construction at work. Paraphrasing Hacking (1999), a constructionist stance on leadership holds that (I) Leadership need not have existed, or need not be at all as it is. Leadership, or leadership as it is at present, is not determined by the nature of things; it is not inevitable. Chapter 1: Two Traditions 5 However, often a constructionist stance will go further: (II) Leadership is quite bad as it is. (III) We would be much better off if leadership were done away with, or at least radically transformed. A thesis of type I that strikes down the inevitability of leadership is the common starting point for constructionist approaches. We have essentially made such a claim about authentic leadership by noting the various ways in which it may be defined. However, any given constructionist approach may or may not embrace the second and third theses. For example, Hardy and Clegg (1996) cast leadership as a mechanism of domination (type II thesis), a position held by many critical theorists who favor more democratic processes (type III thesis) (Deetz, 1995). A discursive approach that embraces critical theory (types II and III) is thus interpretive (type I), but an interpretive orientation does not presume a critical one (Deetz, 1996). The relationship is intransitive. In various ways, discursive approaches embrace the processes of social construction and its products vis-a-vis the operation of one or more texts. Sigman (1992) captured this process orientation by observing that, "the process of communication itself ... is consequential, and it is the 'nature' of that consequentiality that should ... be the appropriate focus" (p. 351). Thus, discursive approaches tend to focus on how leadership is achieved or 'brought off' in discourse-just as Shotter (1993) portrayed managers as practical authors, calling attention to their everyday language use, the performative role of language, and the centrality of language to processes of organizing. Drawing from ethnomethodology, Knights and Willmott ( 1992) cast leadership as a practical accomplishment where a social order may be experienced as routine and unproblematic, but is really a precarious, reflexive accomplishment. The implications of these and other constructionist views of leadership suggest that leaders must constantly enact their relationship to their followers (Biggart & Hamilton, 1987). All must repeatedly perform leadership in communication and through discourse. As we will later see, conceptualizing and studying leadership in this way are often two different things. Importantly, discursive approaches allow leadership to surface in myriad forms, whether it is street gang credibility, role-modeling heroism, or legitimate authority. Jettisoning the concept of leadership is not an option, as it has been for some 'weak leadership' approaches (Shamir, 1999) like that of self-management (Manz & Sims, 1987), or substitutes for leadership (Kerr & Jermier, 1978). As long as the concept of leadership is invoked by actors for attributions of personal potency (Calder, 1977), the concept is worthy of study. While Calder cautioned not to confuse lay constructions with scientific 6 DlSCCRSIVE LEADERSHIP constructs, each can be studied without necessarily undermining the truth claims of either (Edwards, 1997; Meindl, 1993). 10 However, if one is to accommodate the attributions and descriptions of both actors and analysts, searching for the definition of leadership is futile, as many scholars have already concluded (Alvesson & Sveningsson, 20036; Barker, 1997; Rost, 1991). The definition I prefer is a rather simple one by Robinson (2001): "Leadership is exercised when ideas expressed in talk or action are recognized by others as capable of progressing tasks or problems which are important to them" (p. 93). As the ensuing chapters make clear, this definition is useful for four reasons. First, leadership is a process of influence and meaning management among actors that advances a task or goal. Second, leadership is an attribution made by followers or observers. Third, the focus is on leadership process, not leader communication alone, in contrast to heroic leadership models (Yuki, 1999). Finally, leadership as influence and meaning management need not be performed by only one individual appointed to a given role; it may shift and distribute itself among several organizational members. Note that Robinson (2001) does not distinguish between 'leader' and 'manager' in her definition. It is a lead that I will follow unless the particular leadership literature under scrutiny makes the distinction relevant, such as in neo-charisma theories (Bryman, 1996). Note also that I am arguing for the utility of Robinson's definition of leadership, not its veracity. Defining Discourse Grant, Keenoy and Oswick (1998) observed that discourse too has been a highly contested term-over the inclusion of both written text and spoken dialogue (Gilbert & Mulkay, 1984; Sinclair & Coulthard, 1975); visual images such as art, architecture, and media images (Hodge & Kress, 1988; Kress & van Leeuwen, 1990); and reality construction processes shaped by discourse (Berger & Luckman, 1966; Foucault, 1972, 1980; Searle, 1995). Following Alvesson and Karreman's (20006) efforts to clarify the various meanings of discourse, I generally distinguish between two broad definitions. 11 The term discourse (also known as little 'd' discourse) refers to the study of talk and text in social practices. Viewed as a local achievement, discourse embodies cultural meanings; it is a medium for social interaction where the details of language in use and interaction process are central concerns for analysts (Potter & Wetherell, 1987). However, what does it mean to study talk and text? Talk-in-interaction represents sociality, the processes of messaging and conversing. It is the 'doing' of organizational discourse, whereas text is the 'done' or material representation of discourse in spoken or recorded forms (J. R. Ta ylor & Van Every, 2000). Even though written documents are the Chapter 1: Two Traditions 7 simplest way to conceive of organizational texts (for example, emails and annual reports), verbal routines inscribed in organizations such as performance appraisals or job interviews also exist as texts and are reconfigured through their continued use (Derrida, 1988).12 By contrast, the term Discourse (also known as big 'D' Discourse) refers to general and enduring systems for the formation and articulation of ideas in a historically situated time (Foucault, 1972, 1980). In this view, power and knowledge relations are established in culturally standardized Discourses formed by constellations of talk patterns, ideas, logics, and assumptions that constitute objects and subjects. These Discourses not only order and naturalize the world in particular ways, but they also inform social practices by constituting "particular forms of subjectivity in which human subjects are managed and given a certain form, viewed as self-evident and rational" (Alvesson & Karreman, 2000b, pp. 1127-1128). Bennis and Thomas's (2002) Geeks & Geezers: How Era, Values, and Defining Moments Shape Leaders is a business press example of Foucault's notion of Discourse. Bennis and Thomas compare the characteristics of the Great Depression and World War II era with the era of the Internet and end of the Cold War. Their goal was to discern the ways in which the forces of history and culture shaped two generations of U.S. leaders ('geezers' and 'geeks,' respectively) and their organizations. Foucault did much the same kind of analysis, albeit with somewhat more specificity and different topics. Discursive approaches such as sociolinguistics, ethnomethodology, conversation analysis, speech act schematics, interaction analyses, and semiotics in various ways focus on language in use and interaction process; they analyze little 'd' discourse. By contrast, critical and postmodern discourse analyses focus heavily on systems of thought; they analyze big 'D' Discourse. However, there is a third category that attempts both; it includes discursive psychology, rhetoric and literary analyses, ethnography of speaking, and Fairclough's ( 1995) critical discourse analysis. This book utilizes discourse analyses from each category, but makes no attempt to be exhaustive. For those unfamiliar with the discourse approaches used in this book, brief synopses of them appear in Appendixes Al -A7. 13 The differences between discourse and Discourse notwithstanding, discursive approaches vary in at least five other ways according to K. Tr acy (1995). They include (a) whether a transcript is required, and the type and level of detail a transcription should include; (b) the dominant kinds of texts used for analysis; ( c) the role of interviews and other kinds of contextual information; (d) disciplinary orientations and key theoretical questions that the discourse analysis is designed to answer; and ( e) the metatheoretical frame ( empiricist, interpretive, critical) within which the discourse analysis is viewed as a method (p. 200). Where relevant, Appendixes Al-A7 make these differences known. 8 DISCURSIVE LEADERSHIP The Case for Discursive Leadership In order to understand the contributions of discursive leadership, a fuller case must be made for its distinctiveness relative to leadership psychology. Unfortunately, any comparison risks unfairly representing all of the theories and approaches grouped under these two labels. Nevertheless, it does seem useful to try to find threads of unity within the diversity, while respecting the diversity as much as possible. This is because discursive leadership has its own ways of talking-its own language of leadership-that is different from leadership psychology. This will become evident in the following six comparisons between/among discourse and mental theater; decentered subjects/thin actors and essences; reflexive agency and untheorized/exaggerated agency; encompassing and dualistic conceptions of power and influence; textual, con-textual, and variable analytic; and communication as primary and subsidiary. DISCOURSE VERSUS MENTAL THEATER Reacting against Kantian philosophy and the ways it influenced contemporary psychology, the term 'mental theater' was used by Cronen (1995a) to refer to psychologists' need to "get beneath and behind experience to fret out the connections among cognitions, emotions, and behaviors" (p. 29). 14 As psychologists form and correlate the cognitive, affective, and conative variables that they believe capture experience, they must often reduce behavior to statements of intention or summary judgments of past behavior. Cronen argues that all sense of coordinated action (in its often messy, yet fine-grained detail) and any real sense of experience are thus lost in the projected play of mental operations. 15 Similarly, others suggest that when leadership is viewed as the result of variables 'inside' or 'outside' the person, the only interaction that is studied is a statistical one (Hosking & Morley, 1991; Meindl, 1993). There is a difference between studying actual interactional processes where relational patterns are always codefined, and studying reports of such processes as if a single relational reality exists (L. E. Rogers, Millar, & Bavelas, 1985), even though both may derive from theorizing leadership as socially constructed in some fashion. Theories from leadership psychology may miss this distinction when they theorize social processes, yet measure only one party's perceptions of same. These perceptions are retrospective summarizing judgments that gloss the details of interaction over time and may give the impression that a single relational reality can be assumed and measured. 16 Reinforcing this view, Gronn (2002) recalls the distinction between ontological, observational, and analytical units. Ontological units define the entity that one is studying. Observational units define who or what an analyst observes, while analytical units more specifically parcel out that which is to be Chapter 1: Two Traditions 9 deconstructed, measured, or explained. Leadership psychology focuses on the ascending series of individual-dyadgroup-organization as ontological units, which Gronn (2002) argues has historically been confounded with levels of analysis and overshadowed by the dominance of leader-centrism-an individualist concern relative to the other units. Leadership psychologists frequently observe individuals and analyze their perceptions and summary judgments, even if the ontological unit is a leader-member dyad, group, or whole organization (Yammarino, Dionne, Chun, & Dansereau, 2005).17 T he confluence of these ontological, observational, and analytical units thus biases leadership psychologists toward the study of the individual over the social or cultural. This is in no way to discount the study of the individual or demean the fascination with mental theater ( or in any way to ask psychologists not to be psychologists), only to deny the assumed isomorphic correspondence between cognitive operations and social process (Holmes & Rogers, 1995). Cronen (1995a) argues that without a clear and separate focus on social process, analysts have little recourse but to explain individuals' abilities solely in terms of hidden mechanisms and inner motors. Discursive approaches' ontological units include subjectivity, identities, relationships, cultures and linguistic communities, organizations as macroactors, linguistic repertoires, and Discourses as stand-alone systems of thought. On the surface, this looks roughly similar to the individual-dyadgroup organization series of leadership psychology. However, discursive approaches' ontological units are often combinations of more than one 'level' because analysts argue that clear boundaries are often undecipherable (Collinson, 2006). For example, subjectivity is more about a person's image or constructed self relative to the range of conflicting Discourses that vie for control (Deetz, 1992), while the organization as a macroactor focuses on how organizations come to have voice and agency (J. R. Taylor & Cooren, 1997). In terms of observational units, discursive approaches focus on language in use, interaction process, and/or discursive formations. Their analytical units are defined by their choice of text, of which there are all manners and varieties. As indicated below, texts can be written records, inscribed patterns, or memory traces. In the literature base for this book, texts are most often interview discourse as individuals' sensemaking accounts and meaning assignments are revealed in their language use, actual dialogue that captures language use in the backand- forth of interaction process, or discursive formations that may stand alone as systems of thought or appear as dialogically grounded linguistic practices. DECENTERED SUBJECTS/THIN ACTORS VERSUS ESSENCES Grint (2000) asserted that in trait, situational, and contingency theories of leadership, there is an 'essence' to the leader, the context, or both that suggests 10 DISCURSIVE LEADERSHIP one best way to lead. An essence suggests that things are what they are because that is their nature or true form, despite all appearances.18 For Grint (2000), trait approaches emphasize the essence of individual leaders-qualities that make them leaders regardless of the context or circumstances in which they may find themselves. Situational approaches like the Ohio State Leadership Studies emphasize the essence of particular contexts, the effective handling of which requires one leadership style over possible others. Finally, contingency approaches emphasize the essence of individual and context, where individuals gauge their alignment with the context and respond accordingly, for example, when a strong leader and a crisis coincide.19 The search for the essence of leadership derives from leadership psychologists' adherence to traditional science assumptions about realist conceptions of truth and conceiving of knowledge as representing reality (Rorty, 1982). To reject the notion of 'essence' is to embrace a socially constructed view of leadership because "what counts as a 'situation' and what counts as the 'appropriate' way of leading in that situation are interpretive and contestable issues, not issues that can be decided by objective criteria" (Grint, 2000, p. 3). Thus, Grint and other discursive scholars problematize the variability and inconsistency in actors' accounts and analyst findings, explicate the conditions of their production, and thus try to understand how conflicting truth claims about leadership come into being and may actually coexist. These analysts expect to find the research equivalent of the fog of war in the study of social interaction. They choose a constructionist path over essentializing theory because it supplies the necessary tools to grapple with communication's unending detail and variety.20 Included among those tools is the search for vocabularies and ways of talking that best address the purposes at hand (Bochner, 1985; Rorty, 1982). However, in his critique of two types of discourse analysis, conversation analysis and discursive psychology, Hammersley (2003a) pointedly objected to these analysts' unwillingness to "view actors as controlled, or even as guided in their behavior, by substantive, distinctive and stable mental characteristics such as 'attitudes; 'personalities; 'perspectives; or 'strategic orientations"' and their preference for treating actors as" employing cultural resources that are publicly available, and doing so in contextually variable ways" (p. 752, emphasis original). Hammersley thus reclaims the essentialist argument by arguing that a discursive orientation rejects anything that is unique or specific about actors in favor of what any member ( of a linguistic community) could do. Hammersley's critique raises key questions for discursive leadership scholars, namely, how should one think about behaviors that are distinctive to leadership actors across time and context? Is some essence worth hanging onto? To answer these questions, it is important to understand that leadership psychology traditionally relies on a Western conception of human beings as Chapter 1: Two Traditions 11 unitary, coherent, and autonomous individuals, whose 'selves' are separable from society (Holstein & Gubrium, 2000). Essentializing thus appears to be a natural way of making sense of leaders' complex inner lives as well as the contexts in which they operate. For most forms of discursive leadership, society and the individual are inseparable (Giddens, 1979). In postmodern thought, for example, the self is neither fixed nor essentialized for this very reason. Instead, subjectivity emerges as a historical product of sociocultural forces embedded within a specific context (Foucault, 1979, 1983). Such a focus often examines the discursive, gendered, multiple, and conflicting nature of subjectivities in this regard (Ashcraft & Mumby, 2004). However, neither conversation analysis nor discursive psychology goes so far as to portray actors as decentered subjects. Yet Hammersley (2003) still finds their model of the human actor to be rather "thin" compared to the actor with a strong inner motor. In his rejoinder to Hammersley, Potter (2003) argues that "a certain kind of thinness;' best characterized as lacking "a predefined model of the human actor;' is necessary in order to focus on social practices, the constitutive role of language, and the contributions of the cultural (p. 78-79). Interestingly, both Hammersley and Potter legitimate ·both actor orientations, although Potter flatly rejects Hammersley's suggestion that discursive approaches be viewed less in paradigmatic terms and more as methodologies. Indeed, few discourse scholars would stand for any minimization of their commitments to theory. However, Deetz (1996) offers a more accommodating solution, one that I prefer. That is, paradigmatic differences should not be seen as alternative routes to truth, but as specific (D)iscourses which, if freed from their claims of universality and/ or completion, could provide important moments in the larger dialogue about organizational life. The test ... is not whether they provide a better map, but whether they provide an interesting way to talk about what is happening in research programs. (p. 193) Deetz's poststructuralist solution finds further grounding in Rorty's (1979) notion of 'conversation; where the "focus shifts from the relation between human beings and the objects of their inquiry to the relation between alternative standards of justification" (p. 389-390).21 Discursive leadership and leadership psychology are thus usefully conceived as complementary Discourses or alternative ways of talking and knowing about leadership. ENCOMPASSING VERSUS DUALISTIC CONCEPTIONS OF POWER AND INFLUENCE The Western conception of an autonomous self-adopted by leadership psychology and the self that is inseparable from society embraced by discursive 12 DISCURSIVE LEADERSHIP scholars also implies different ways in which to view power and influence. Collinson (2006) observes that traditional conceptions of power in leadership psychology treat it as a negative and repressive property exercised in a top-down manner. Influence is thus treated independently, most often as embodying the very definition of leadership (Antonakis, Cianciolo & Sternberg, 2004; Rost, 1991; Yukl, 2002). To be more precise, leadership is understood as a "positive process of disproportionate social influence" ( Collinson, 2006, pp. 181-182). Indeed, by today's common standards, shaped heavily by leadership psychology, leadership fails when a leader must resort to his or her authority to gain compliance. Such a view also explains our admiration for charismatic and transformational leaders who excel at the influence game by winning the voluntary cooperation of followers, at times under extraordinary circumstances. In order to explain forced versus voluntary compliance, leadership psychology treats power and influence as dual notions. Many discursive approaches would not restrict their study of leadership to positive and disproportionate influence. Their views on the inseparability of self and society derive from a view of power that is much more encompassing, one that integrates various forms of power and influence and conceives of them in both positive and negative terms. Such a view draws heavily from Foucault (1990, 1995), who argues for the cultural and historical contingency of subjectivity along with its Discursive roots in power and knowledge systems. For Foucault, all power is local, relational, and embedded in specific technologies governed by Discourses with the power to discipline. As we will see in Chapters 4 and 5, such technologies are usually aided by systems of surveillance that turn individuals into knowable and calculable objects (Miller & Rose, 1990). With this kind of apparatus, we are able to see the individualizing effects of power, especially as individuals come to discipline themselves around that which a Discourse deems 'normal.' Power stays close on the heels of resistance here, traveling its same routes in order to overcome. Finally, when multiple Discourses are considered, the positive, productive, and creative aspects of power reveal themselves especially as individuals forge their identities ( Collinson, 2006). Thus, discursive scholars find that more encompassing views of power and influence are necessary to explain the inseparability of self and society. REFLEXIVE AGENCY VERSUS UNTHEORIZED/EXAGGERATED AGENCY Leadership is often viewed as a force for change (Bennis & Nanus, 1985; Hickman, 1990; Kotter, 1990), making it nearly synonymous with the terms 'agency' or 'action.' However, agency per se is an infrequent topic in leadership psychology, which has led critics to make two seemingly contradictory observations about this literature body. First, Hosking (1988) argues that leaders are Chapter 1: Two Traditions 13 too often untheorized as agents, which results in an odd disconnect between leadership research and the rest of the field. Bryman (1996) puts it more bluntly: "Leadership theory and research have been remarkably and surprisingly uncoupled from the more general field in which they are located" (p. 289). Hosking (1988) argues that, while the skills ofleadership are the skills of organizing, leadership psychologists have been too caught up in assuming the organization has an entitative statusneither questioning how the organization got to be an entity in the first place, nor how it maintains itself as an entity. When researchers ignore the processes of organizing, Hosking notes "a sharp divide between person and organization such that the agent, responsible for the latter, is left untheorized as an agent" (pp. 149-150). Consequently, leadership appears epiphenomena!. Second, Gronn (2000) makes the case for exaggerated agency by noting that the individualism and leader-centrism of leadership psychology results in a rather unsophisticated leader-follower dualism in which "leaders are superior to followers, followers depend on leaders, and leadership consists in doing something to, for, and on behalf of others" (p. 319). According to Gronn (2000), this "belief in the power of one" results in an exaggerated sense of agency because of an undertheorized view of task performance and accomplishment (p. 319). Indeed, Robinson (2001) too portrayed leadership psychology as floating ethereally above task accomplishment. If the division of labor were truly examined, Gronn (2000, 2002) reasons, leadership would surface as a more distributed phenomenon and the hero-anointing tendencies of, for example, neocharismatic leadership theories would be in check. Can Hosking and Gronn start from the same literature body and arrive at two different senses of agency? Yes, and both have a point. Gronn (2000) is correct, as others have noted the strong individualism and overstatement associated with the heroic capabilities of charismatic and transformational leaders (Beyer, 1999; Yukl, 1999). Yet Hosking (1988) is also correct because, across this genre of leadership theories and most others, agency is never explicitly theorized, the organization ontological status is assumed, and the disconnect between leader and organization perpetuates itself with inattention to the processes of organizing (Fairhurst & Putnam, 2004). L􀁦adership is still seen as a phenomenon embodied in persons, not as an organizing process grounded in task accomplishment (Fairhurst, 2006). Yet the move to study leadership as an organizing process cannot be done in the absence of discourse/Discourse and communication, a fact to which Gronn (2000, 2002) and Hosking (1988) only indirectly allude. As Bateson (1972) observes, what else do people have between them but the exchange of messages? Bateson argues that communication is the relationship because, following Schegloff (2001), it is the cellular biology or granularity from which perceptions are formed. As later discussion in this book makes clear, this is not 14 DISCURSIVE LEADERSHIP to bias the study of leadership in the direction of social process over cognitive operations. It is to suggest alternating the lenses so that one does not mistake the individual for the social in leadership study, as when survey or interview data substitute for a codefined leadership process. Importantly, the study of leadership discourse, not solely leader discourse, creates the kind of window in which to study the reflexive agency of its actors. 22 Such a view is based on the more general ethnomethodological argument of Garfinkel (1967) that action is organized from within-meaning that leadership actors are knowledgeable agents, who reflexively monitor the ongoing character of social life as they continuously orient to and position themselves vis-a-vis specific norms, rules, procedures, and values in interaction with others. What often seems paradoxical from the outside view of the researcher is logical and reasonable from the inside view of the actor, leading Garfinkel (1967), Giddens (1984), and others to object to the widespread derogation of the lay actor throughout much of the social sciences. With actors' language use, in particular, most discursive approaches view it as a window on human agency because "actions and the interpretations of their meanings are inseparable and occur simultaneously in the course of their production" (Boden, 1994, p. 47). Although discursive approaches certainly vary in how much knowledgeability they attribute to actors, most acknowledge that actors can be viewed as responsible agents who still do not fully comprehend or intend the nature of unfolding events (Giddens, 1979, 1984; Ranson, Hinnings, & Greenwood, 1980). Second, to attribute knowledgeability and reflexivity to actors is to put them in charge of their own affairs in a way that is marked by constraint as much as by freedom. As such, leadership actors must continuously manage the tensions between agency and constraint or structure (Giddens, 1984).As with actor knowledgeability, the issue of constraint is the subject of considerable debate. Charges of relativism have been ascribed to constructionist approaches generally and poststructuralist approaches specifically (Reed, 2000, 2001). Relativism suggests an exaggerated form of agency, an "anything goes" ability to construct reality despite the constraints of a material world (Gergen, 1991). By contrast, more realist constructionist approaches conceive of agency as constrained by material forces such as the brute facts of a physical world (for example, buildings, mountains, hurricanes, and so forth) or macro social contexts of institutions and power relations (Edwards, 1997; Hacking, 1999; Searle, 1995). Just how the material intervenes to constrain action continues to be the subject of considerable debate in constructionist thought in and around organizations. 23 TEXTUAL, CON-TEXTUAL VERSUS THE VARIABLE ANALYTIC As discussed above, leadership psychologists focus on the individual and a search for essences. A search for essences coalesces nicely with the variable Chapter 1: Two Traditions 15 analytic tradition, which holds that complex phenomena like leadership are best understood in terms of a fine-grained analysis of parts. As part of this research tradition, analysts value generalizable over local knowledge and are far more interested in answering cause-and-effect 'why' questions than the more descriptively oriented 'how' questions (as in, how is leadership brought off?). Consequently, leadership researchers in the variable analytic tradition try to capture the experience of leadership by forming and statistically analyzing a host of cognitive, affective, and conative variables and their causal connections. Context is not unimportant, but too much attention to its contingencies produces more local than generalizable knowledge. As a result, there is often less attention to leadership's historical and cultural/political conditions, while a heavy reliance on cross-sectional designs and quantitative methods further enable analysts to aggregate across contexts in the search for the generalizable (Bryman, Bresnen, Beardsworth, & Keil, 1988; Conger, 1998; Parry, 1998). It should surprise no one that survey researchers (many of whom are leadership psychologists) often view discourse analyses as fuzzy, unwieldy, and without a tangible payoff (Oswick, Keenoy, & Grant, 1997). Unconcerned with the search for essences or causal connections among variables, discourse analysts want to know how a text functions pragmatically, how leadership is brought off in some here-and-now moment of localized interaction. In complementary fashion, Discourse analysts ask, what kind of leadership are we talking about and how have the forces of history and culture shaped it? Both types of analysts reject prediction and control as key functions of theory, while never viewing description as mere description or prelude to the real work of theory building. Without the immediate concern of building generalizable theory, discourse scholars feel freer to embrace the context and its historical and cultural/political aspects. As Biggart and Hamilton (1987) write, "Leadership is a relationship among persons in a social setting at a given historic moment' (p. 438, emphasis added). Thus, local knowledge is key as text and context inevitably merge. Most discourse analysts take their cues from Bateson (1972) on this point, who argues that each action (which, once materialized, becomes text) is "part of the ecological subsystem called context and not ... the product or effect of what remains of the context after the piece which we want to explain has been cut from it" (p. 338, emphasis original). Thus, what is text one moment for the discourse analyst is con-text the very next.24 What may also be particularly disturbing to the survey researcher is the protean nature of 'text' versus that of the 'variable.' A variable usually refers to a welldefined class of behaviors that can take on different values. The concept of 'text' has great currency in the organizational discourse literature precisely because it assumes myriad forms such as written records, memory traces, materialized spoken discourse, verbal routines, and so on where size or amount of text matters little. Texts also possess qualities like inscription and restance, which 16 DISCURSIVE LEADERSHIP defines a text's "staying quality" (Derrida, 1988). Texts may even become a metaphor for the organization itself with their capacity to layer and interweave (Cooren, 2001; J.R .T aylor & Van Every, 2000).U nlike variables, texts may or may not have a unitary property whose order and coherence is the subject of analysis. Does a textual analysis preclude or supercede the need for variable analysis or vice versa? No, they often address different kinds of questions even with the same subject matter. For example, Schegloff (2001) suggests that in variable analytic studies connecting status/power and interruption behavior (Kollock, Blumstein, & Schwartz, 1985; Smith-Lovin & Brody, 1989), "what is lacking is the 'cellular biology' that 'closes the connection,' which explicates the mechanism linking the outcomes being studied, initiating interruptions and 'succeeding' with them, and the variables which assertedly engender these outcomes" (p. 315). Thus, Schegloff is not discounting the variable analytic work in this area. Rather, he is suggesting that it is how parties achieve the relevance of their status and power vis-a-vis linguistic forms like turn taking and category memberships (for example, based on gender) in a series of interactional moments that usefully provide this 'cellular biology' or granularity that he finds so missing in variable analytic studies. Thus, the variable analytic connection between power/status and interruption behavior serves as a useful starting point for a more finegrained textual analysis, such as conversation analysis. One of the desired goals of this book is that discursive scholars and leadership psychologists will find more complementary connections. Some of these connections will be made explicit in the chapters on sequence and temporal form (Chapter 2), membership categorization (Chapter 3), and narrative logics (Chapter 6), all chapters focusing on little 'd' discourse. Chapters focusing on big 'D' Discourse appear less amenable to variable analytic tie-ins. However, they certainly contribute to the discussions of leadership psychologists regarding how best to conceptualize leadership and explicate its practices.A s outside the mainstream, their role is an important one in challenging taken-for-granted assumptions and suggesting alternative ways in which leadership may be usefully conceived. COMMUNICATION AS PRIMARY VERSUS SUBSIDIARY More than the variable analytic tradition, it is the psychological orientation of mainstream leadership researchers that predisposes them to view the social and the communicative as subsidiary to individual (and broadly) cognitive operations. Communication merely plays out the cognitive and only partially, at best. This is part of the "received view" of communication, which Cronen ( 1995a) suggests "cannot be the site of the most important avenues of social inquiry because psychological, sociological, and cultural variables determine it ... we [referring to psychologists primarily] only care about communication because it can have consequences for other matters that are our real Chapter 1: Two Traditions 17 concerns" (p. 310). Stated otherwise, communication is of interest only to the extent that actors can impact each other's cognitive operations. Communication is of primary interest to discursive leadership scholars, although interest in human interaction varies as the distinction between 'discourse' and 'Discourse' makes clear. Interestingly, there is also substantial disagreement over the terms 'communication' and 'discourse' among discourse scholars. For example, in writing about the emphasis of conversation analysis on talk-as-action, Edwards (1997) views it as antithetical to what he called a "communication model;' in which communication is strictly a means of expressing speaker intentions and an act of transmission. For conversation analysis and discursive psychology, speaker intentions are at issue in the talk-ininteraction of participants: "intentions, goals, mental contents, and their intersubjective 'sharing' are analyzed as kinds of business that talk attends to, rather than being the analyst's stock assumption concerning what is actually going on" (Edwards, 1997, p. 107). As such, Edwards's "minds-in-communication" view is quite consistent with the received view of communication as depicted by Cronen (1995a). Yet most theorists in the discipline of communication neither endorse a strict transmission model of communication, nor equate the study of communication with speaker intentionality and its transmission aspects. 25 In contrast to Edwards (1997), some organizational communication theorists like J. R. Taylor (personal communication, May 2002) actually prefer the term 'communication' over 'discourse' because the latter term obscures the relationship between interactive speech and text, a relationship that he believes explains the way the organization emerges in communication (J. R. Taylor & Van Every, 2000). Preferences for 'communication' or 'discourse' aside, most discursive approaches eschew a strict emphasis on speaker intentionality and communication as a simple act of transmission, while embracing more meaning-centered models of communication. However, one can hardly resist essentializing leadership and then turn right around and claim that a meaning-centered model is the 'true' model of communication----even if 'meaning' itself has been a contested term when it comes to the interpretations, understandings, and readings of texts within different genres of discourse analysis (Putnam & Fairhurst, 2001). Like Grint (2000), Craig (1999) denies "that any concept has a true essence except as constituted within the communication process" (p. 127). He suggests that, in the case of 'communication; warrant can be found for both transmission and meaningcentered definitions of communication, depending upon the causes they serve.26 The Path Forward Now that I have sketched a broad outline of the differences between discursive leadership and leadership psychology, each of the chapters that follow will 18 DISCURSIVE LEADERSHIP address a key concept from the discourse/Discourse literature and apply it to leadership. The concept may or may not be tied to one specific discursive theory or approach; in fact, it could be several. As mentioned, Appendixes Al-A7 offer summaries for the uninitiated to discourse analyses and are designed for quick and easy reference. Echoing Edwards (1997), those who are new to this literature may find that some of the discourse concepts I have selected to build chapters around may appear a bit mundane at first. For example, when effective leadership can impact life-and-death struggles in high-reliability organizations like police units, conversational turn taking or category use may seem rather unremarkable in the grand scheme of things. But to borrow a distinction made by Staw (1985), it is when these concepts become problem driven through case analysis that they develop import and relevance for leadership. In that sense, they will seem far less literature driven than concepts from leadership psychology, whose debates are about gaps in the literature, inconsistencies, challenges to conventional wisdom, fresh perspectives, and so on. For this reason, my treatment of the discourse concepts is more heavily weighted toward enlightening examples and the use of theory not specifically designed to study leadership per se. 27 However, where relevant in the chapters, one or more theories or approaches from leadership psychology will enter the discussion of the concept; thus its literature base will be important to consider. My stance toward leadership psychology in the ensuing chapters is both appreciative and critical in this regard. With these caveats in mind, Chapter 2's discursive concept is sequence and temporal form in social interaction (little 'd' discourse). In leadership psychology, there is a tendency to study leadership apart from the tasks being performed, and in this chapter we will examine how a sequential orientation to leadership interaction can address this problem. There are a number of discourse approaches that focus on sequencing, but with different kinds of temporal units. Those units include (a) the actinteract-double interact, (b) turn taking and adjacency pairs, (c) narrative schemas and episodes, (d) scripts, and (e) script formulations. These units form the foundation of this chapter and collectively suggest that studying leaders' actions alone yields incomplete and ultimately distorting views of leadership interaction. This chapter also makes the case for distributed leadership and the sequential foundation of leadership command presence. From leadership psychology, Judi Komaki's work on performance monitoring also makes an interesting contribution to this chapter. Chapter 3's discursive concept is membership categorization (also little 'd' discourse). In contrast to leadership psychology theories that focus on the cognitive processes underlying categorization and its consequences for leadership, a discursive approach like conversation analysis examines the performative nature of categories. In other words, how are categories invoked, created, modified, or rejected in everyday leadership discourse and for what purpose? This chapter considers the consequences of category use, especially for organizational role and Chapter 1: Two Traditions 19 identity management in the leadership relationship, and how category work as a discursive activity can add some much-needed specificity to leadership as the management of meaning. While Chapters 2 and 3 focus on language in use and interaction process (little 'd' discourse), Chapters 4 and 5 explore those powerful historical and cultural forces that lie beyond language and interaction, yet serve as important resources for actors as they communicate (big 'D' Discourse). Chapter 4 addresses the role of history in the organizational sciences, including the study of leadership, and contrasts it with the work of Foucault, which is featured in this chapter. Through his conception of Discourse and disciplinary power, leadership actors are shown as subjects and objects of their relationships, organizations, and societies. It makes for quite an interesting contrast with leadership psychology's view of leaders as crucial agents. The resonances of Foucault's examination and confessional technologies are argued to operate in modern day performance management approaches such as 360-degree feedback and executive coaching. Moreover, the emerging executive coaching literature demonstrates how even the alpha males among leaders may be tamed and disciplined through the power of Discourse. Chapter S's focus is on the self and identity. The conception of the self in many leadership psychology theories is very different from the self in discursive leadership. In a discussion of these differences, we will continue to draw from the work of Foucault and his view of Discourse. However, the focus is also on the role played by multiple Discourses, including authentic leadership and gendered management Discourses, in the self-identity work of leadership actors. To carry out this task, we will adopt the translation of 'Discourse' into 'interpretative repertoire' by discursive psychology in order to understand the linguistic resources made available to leadership actors. Chapter 6's study of narrative explores the intersection of both discourse and Discourse. The chapter title, "Narrative Logics;' addresses itself to narrative as found in leadership interaction as well as the narrative resources afforded by various Discourses. Leader-member exchange (LMX) theory from leadership psychology is an ideal candidate in which to explore these narrative ties. Chapter 6 demonstrates that a narrative approach adds nuance and detailed meaning to the character and quality of LMX, particularly for the less well-understood, medium quality LMXs. It also reveals coconstructed, terse storytelling in LMX dialogue, and cultural contributions to LMX via the uniqueness paradox wherein culturally scripted narratives feel idiosyncratic. Chapter 7 addresses material mediations in leadership discourse. This chapter begins by examining the elusiveness of charismatic leadership and proposes that a discursive approach rooted in actor-network theory, known as the Montreal school, be used to study charisma. This approach recasts charismatic leadership in textual, scenic, technological, cultural, and embodied terms. Rudy Giuliani's leadership during 9/11 will be used to demonstrate that charisma is perhaps best seen not as residing in a single person, but as an 20 DISCURSIVE LEADERSHIP attributed product or outcome of a continuous networking strategy of human and nonhuman entities. Along the way, this chapter explores the criticism of discourse analyses vis-a-vis materialist concerns. Finally, Chapter 8 has two goals. First, the implications of discursive leadership research for leadership practice must be articulated, which suggests more conversation between discursive leadership scholars and practitioners. Second, I am looking to keep the conversation going between discursive leadership scholars and leadership psychologists. It seemed fitting to ask scholars from both groups to offer their take on the possible interplay between discursive leadership and leadership psychology. I am grateful to Boas Shamir, Stephen Green, Robert Liden, James (Jerry) Hunt, Donna Chrobot-Mason, Frarn;:ois Cooren, Kevin Barge, and Linda Putnam for their agreement to participate in this dialogue. As the ensuing chapters will make clear, and I hope readers will agree, discursive leadership is an interesting and powerful lens in which to view leadership and, speaking for myself and other discursive leadership scholars, our conversation with leadership psychologists has only just begun. NOTES For example, see work by Carlson (1951), Sayles (1964), and Stewart (1967). 2. Mintzberg's (1973) observations of managers produced a taxonomy in which there were three roles that dealt with managers' interpersonal behavior (leader, liaison, figurehead), three roles that dealt with information processing (monitor, disseminator, spokesman), and four roles dealing with decision making (entrepreneur, disturbance handler, resource allocator, negotiator). 3. Mintzberg's research was incomplete and is now outdated in many respects. For example, see Gronn (2000), Hales (1986), Reed (1984), Stewart (1983), and Willmott (1984). 4. For especially good introductions to the linguistic turn in the communication sciences, see Bochner (1985) and Deetz (1992), and in the organizational sciences, see Alvesson and Karreman (2000a). 5. This point echoes the debate over paradigm incommensurability in the organizational sciences (Burrell & Morgan, 1979; Conrad & Haynes, 2001; Corman & Poole, 2000; Deetz, 1996; Gioia & Pitre, 1990; Hassard, 1988, 1991; N. Jackson & Carter, 1991; Parker & McHugh, 1991; Scherer, 1998; Weaver & Gioia, 1994). 6. This would also include ignoring other spheres of joint influence such as the economic, biological, or physical aspects of leadership. 7. Albert et al.'s (1986) principle of 'complementary holism' derives from the work of physicist David Bohm and modern quantum physics. Among other things, it emphasizes "that reality is not a collection of separate entities but a vast and intricate 'unbroken whole"' (p. 12). 8. At least one writer has suggested that discursive approaches such as conversation analysis and discursive psychology be seen as methodologies (Hammersley, 2003a, 2003b), which is counter to the view expressed here. 9. See Leadership Quarterly's 2005 (16:3) special issue on authentic leadership. Chapter 1: Two Traditions 21 10. In Meindl's (1993) 'radical' social psychological approach to leadership, he concludes, "A key to understanding and conceptualizing leadership must be built on the foundation of a na"ive psychological perspective. How leadership is constructed by both na"ive organizational actors and by sophisticated researchers should constitute the study of leadership. For it is these very constructions on which the effects of leadership, defined in conventional terms, are like to depend"(p. 97). 11. Analysts often falsely assume a consensus instead of a range of positions regarding the meaning of 'discourse/Discourses.' See Alvesson and Karreman (2000b) for further discussion. 12. Texts have the capacity to layer and interweave processes of organizing, thus creating the notion of organizations writ large as texts layered within macro texts (Putnam & Fairhurst, 2001). 13. For a review of the organizational literature on these discourse approaches, see Putnam and Fairhurst (2001). 14. Thus, this excludes leadership theory and research where cognition and perceptual processing ;ire the primary focus, such as implicit leadership theories (Lord & Emrich, 2001). 15. An example of focusing on the play of mental operations can be found in the practice of modeling psychological variables and behavioral outcomes. For examples, see the June 2005 issue of Leadership Quarterly devoted to authentic leadership (Gardner et al., 2005; Ilies, Morgeson, & Nahrgang, 2005). 16. One currently prominent example of this is leader-member exchange theory, which often focuses solely on the member's perspective in assessing the quality of the exchange (Graen & Scandura, 1987; Graen & Uhl-Bien, 1995). However, a discursive approach may fall prey to this same criticism if it focuses exclusively on one-sided interview data, which may report more interactional detail, yet show only one person's definition of the situation or the relationship. 17. Yammarino et al. (2005) reviewed 348 journal articles and book chapters on leadership from the last 10 years in 17 areas of leadership study. They coded them for the degree of appropriate inclusion and use of levels of analysis in theory, measurement, data analysis, and inference drawing. They concluded that 91 % of the publications reviewed failed to adequately address levels of analysis issues. They characterized these findings as "troubling;' given that the 'levels' issue has been around for some 20 years (Dansereau, Alutto, & Yammarino, 1984). 18. Cronen (1995a) provides a particularly good example of essentialist thinking with factor analysis: "Some researchers argue that the factors-mathematically derived vectors through n-dimensional space-stand for the common essence shared by variables loading on the factor. When this interpretation is carried to the extent of treating the mathematically created factors as the most important reality, we have Platonism masquerading as empiricism" (p. 43). For Cronen, the factor or essence is not the reality, although it is frequently treated as such by empiricists. 19. Although not specifically addressed by Grint (2000), neo-charismatic models of leadership appear to essentialize leaders and leader-context combinations. 20. To what extent is essentializing as activity or thinking exercise just a natural part of attributional processes or the assignment of meaning in sensemaking processes? For example, Meindl, Ehrlich, and Dukerich (1985) observed that in causally indeterminate and ambiguous organizational conditions, leadership assumes a romanticized, exaggerated role in accounting for successful or failed outcomes. The attribution made 22 DISCURSIVE LEADERSHIP here is essentializing in that the people in this study drew 'hero-making' themes from the flow of experience in their sensemaking. However, here I would draw a distinction between essentializing as a sensemaking activity and essentialism, which is a philosophical position dating back to Aristotle and the phenomenology of Husserl ( 1962 ). Modern day versions of essentialism view social objects, in this case leadership, as given objects in the world-innately possessing a true nature or 'essence' whose meanings must be grasped/discovered rather than viewed as constructed. While any assignment of meaning by the actor or observer asked to reflect upon leadership essentializes it, it is one meaning assignment in a milieu of many others that may coincide, contradict, or lack any relation whatsoever. Essentialism in leadership study is thus counterbalanced by recognizing that which is contested about leadership, questioning the solidity or facticity of the social world, and viewing leadership as a situated, ongoing practical accomplishment (Garfinkel, 1967). 21. Rorty (1979) argued that conversation between alternative standards of justification can ultimately affect changes in those standards. Further, by abandoning the notion of knowledge as representation, our task as scientists is not the endless search for essences, but vocabularies and languages suitable for particular aims and goals (Rorty, 1982). 22. However, more empiricist discourse approaches like interaction analysis, if they theorize agency at all, embrace an externalized conception of agency, one that depends on a connection or relation in a network/system of relations thus ascribing constrained choice to actors but deemphasizing their active interpretive role in making those choices (Fairhurst, 2004). 23. There are several sources on this point (Astley, 1985; Chia, 2000; Conrad, 2004; Conrad & Haynes, 2001; Deetz, 1992; Foucault, 1972, 1980, 1995; Gergen, 2001; Gioia, 2003; Hacking, 1999; Parker, 1998; Potter, 1996; Reed, 2000, 2001, 2004; Shotter, 1993; Tsoukas, 2000). 24. Discourse analyses vary greatly in how much of the immediate context is incorporated into their analyses; interaction analysis and conversation analysis are particularly adept at accounting for the immediate communicative context (Fairhurst, 2004). 25. A number of communication scholars make this clear (Ashcraft & Mumby, 2004; Cooren, 2001; Craig, 1999; Cronen, 1995a; Deetz, 1992; Fairhurst, 2001; Pearce, 1995; Putnam, 1983; Putnam & Boys, 2006; Stohl & Cheney, 2001; Taylor & Van Every, 2000). 26. According to Craig (1999), transmissional views of communication possess great cultural currency and may bolster the authority of technical experts, while meaning-centered views promote the cause of freedom, tolerance, and democracy. From this vantage point, Craig does not view these definitions as mutually exclusive. 27. It is quite common in the little 'd' discourse literature for argument to proceed from example (S. Jackson, 1986; Jacobs, 1986, 1988, 1990; Pomerantz, 1990). Those adopting this practice are chiefly conversation analysts, speech act theorists, and discursive psychologists who, it must be remembered, are not making claims about behavioral regularities in the way that the more quantitative interaction analysts areor leadership psychologists for that matter. Instead, the emphasis is on claims about structural possibilities and coherent configurations generated by the particular system of discourse in question (Jacobs, 1986). 2 Sequence and Temporal Form W ithin the organizational sciences, the study of leadership is sometimes depicted as a discipline unto itself. For example, in his review of the literature, Bryman (1996) asserts that, "Leadership theory and research have been remarkably and surprisingly uncoupled from the more general field in which they are located" (p. 289). Both Gronn (2000) and Robinson (2001) concur and attribute it to an undertheorized view of task performance. However, Hosking ( 1988) argues that leadership researchers also err by conceptualizing the organization as an already formed entity, instead of one that is in a state of becoming. 1 As such, they fail to conceptualize leaders as agents whose actions, as Giddens (1984) would say, "make a difference" in the ongoing course of events. By contrast, when researchers embrace the processes of organizing, leadership becomes intrinsic to it because the skills of leadership are the skills of organizing (Hosking & Morley, 1988). As Hosking (1988) explains it is not enough to understand what leaders do. Rather it is essential to focus on leadership processes: processes in which influential 'acts of organizing' contribute to the structuring of interactions and relationships, activities and sentiments; processes in which definitions of social order are negotiated, found acceptable, implemented and renegotiated; processes in which interdependencies are organized in ways which, to a greater or lesser degree, promote the values and interests of the social order. In sum, leadership can be seen as a certain kind of organizing activity. (p. 147) Hosking's view restores agency to leadership theorizing because the organization is in a constant state of becoming only through the actions of its agents, thus making it difficult to cast leadership study in isolationist or epiphenomenal terms.2 Although Hosking was not particularly oriented to organizational discourse, her arguments take on new meaning with the study of sequence and temporal form in social interaction. To study sequence and temporal form is to 23 24 DISCURSIVE LEADERSHIP study interaction process, the dance between leader and led and its language of connectedness, temporalness, patternedness, and embeddedness (L. E. Rogers & Escudero, 2004b ). Hinde ( 1979) offers a deceptively simple example of this orientation by observing that it makes a relational difference whether a husband and wife consistently kiss after they quarrel or quarrel after they kiss, even though the total amount of quarreling and kissing may be the same. Too often, survey or interview data are used as a proxy for interaction process as if there is a single relational reality that can be presumed. 3 According to L. E. Rogers, Millar, and Bavelas (1985), this equates a description from the part (one member) with the description of the whole (the relationship itself), a slip that is especially seductive in dyad study where there are only two people in the relationship. From an interactional perspective, relational patterns are always codefined. This is because individuals in a leadership relationship do not relate and then communicate, they relate in communication (McDermott & Roth, 1978). A leader or a member's action is thus of little consequence unless it is a reaction or antecedent to the other's action.4 Interaction analysis is the study of interaction process. McDermott and Roth ( 1978) define interaction analysis as when "a person's behavior is best described in terms of the behavior of those immediately about that person, those with whom the person is doing interactional work in the construction of recognizable social scenes or events" (p. 321; emphasis original). McDermott and Roth were actually describing a number of approaches in the 1970s, with an emerging emphasis on sequence and temporal form, including ethnomethodology, conversation analysis, sociolinguistics, exchange theory, and network analysis. While the definition of interaction analysis has been narrowed significantly over the years (Fairhurst, 2004), the more general point remains. There is a loose confederation of discursive approaches, with seemingly few ties to any one theoretical tradition, that focus on the sequence and temporal form of behavior. These approaches currently include interaction analysis (in the form of coded interaction), conversation analysis, speech act schematics, and discursive psychology ( for brief overviews, see Appendixes Al -4). Their emphases on sequence and temporal form include acts, interacts, and double interacts; turn taking and adjacency pairs; narrative schemas and episodes; scripts; and script formulations. These temporal forms serve as the basis for this chapter's organization. Collectively, they will demonstrate why the individual leader or follower is no longer the unit of analysis. Individually, they will have much to say that goes to the heart of Hosking's ( 1988) enjoinder to study the organizing potential of leader-follower actions.5 For example, the study of acts, interacts, and double interacts in interaction analysis can inform the study of collective mind in high-reliability organizations as a global structure emerges from local interactions (Weick & Roberts, 1993). The study of turn taking and adjacency pairs in conversation analysis creates a stage for distributed leadership to emerge. Chapter 2: Sequence and Temporal Form 25 Because turn taking reveals the articulation points of action, it is leadership-asit- happens, however distributed or asymmetric the influence patterns (Boden, 1994). T he study of narrative schemas and episodes in speech act schematics argues strongly for the sequential foundation of leadership command presence and treatment of it as a strategic response, not a trait-like characterization such as we find in neo-charisma theories. Finally, the study of script and script formulations tells us about the organizational scripts of leadership actors, improvisation around their timing and pacing, and actors' reflexivity surrounding them. While topics such as collective mind, distributed leadership, command presence, scripts, and script formulations are present in varying degrees in the mainstream leadership literature, this chapter's emphasis on sequence and temporal form makes the case for a relational view of leadership, quite different from the individualism of leadership psychology. Consider first Weick's ( 1979) now familiar double interact. Act, Interact, Double Interact While a number of social scientists lay claim to the above title's terminology, Weick (1979) is perhaps best known for popularizing it within the organizational sciences. An 'act' is the behavior of one person, an 'interact' is the response of another to that behavior, and a 'double interact' is the response to the response. Consider one of his early examples: [S]uppose that a supervisor wants to get a worker to stop doing task A and start doing task B.- The worker's action is the doing of task A; the supervisor tries to influence the worker to do task B. Obviously, we must know how the worker responds to this directive before we can make any statement about the complete influence attempt. But to determine the worker's response, we need a specific description of the original activity as a basis for comparison. The worker's typical response pattern will probably be altered in some way by the supervisor's directive, and before we can understand the meaning of this alteration, we need to know the action that was already under way. (p. 89) Weick's (1979) model of organizing is built upon contingent response patterns like the double interact (in this case, how the worker responds to the directive), which he views as the building block of organizational growth and decline. For Weick, sets of double interacts aimed at reducing equivocality assemble into processes that become stable collective structures the more they are repeated (a point to which we shall return). Although Weick (1979) does not concern himself with organizational discourse very much,6 there are discourse analysts who hold Weick's appreciation for the value of interacts and double interacts. Interaction analysts form one such group (for a brief overview, see Appendix A2). To study acts, interacts, 26 DISCURSIVE LEADERSHIP and double interacts, they first code verbal behavior according to a predefined set of codes in order to analyze the patterns that form over the course of several minutes and multiple conversations. 7 More specifically, coding enables interaction analysts to quantitatively analyze the sequences (...