Abstract
This paper employs a psychodynamic perspective to examine the development and maintenance of a leader’s identity, building on the premise that such identity work involves both conscious and unconscious processes. We focus on the latter by suggesting that those in coveted leadership roles may engage in projective identification to shape and sustain an identity befitting those roles. Projective identification is the unconscious projection of unwanted aspects of one’s self into others, leading to the bolstering of a conscious self-view concordant with one’s role requirements. Recipients of a leader’s projections may manage these by projecting them back into the leader or into third parties, which may lead to ongoing conflict and the creation of a toxic culture. We use examples from the Gucci family business to illustrate this process.
The ways individuals craft, uphold, and revise their identities—captured within the definition of “identity work” (Snow & Anderson, 1987; Sveningsson & Alvesson, 2003)—have gathered much attention from organizational scholars. Researchers have elucidated how individuals shape their self-conceptions, within social interactions, in order to transition into or sustain a desired role (Ibarra, 1999; Kreiner, Hollensbe, & Sheep, 2006; Pratt, Rockmann, & Kaufmann, 2006; Thornborrow & Brown, 2009) or to avoid the taint associated with a stigmatized one (Ashforth & Kreiner, 1999; Snow & Anderson, 1987). These studies have deemed identity work successful if individuals manage to craft identities that sustain their self-esteem and grant them social validation in their roles, and have provided the foundations for an emergent stream of organization studies concerned with the identity dynamics underpinning the emergence and exercise of leadership (Day & Harrison, 2007; DeRue & Ashford, 2010; Ibarra, Snook, & Guillen Ramo, 2010; Lord & Hall, 2005).
This stream of scholarship rejects the assumption that leadership is synonymous with occupying positions of formal authority or enacting requisite styles, and endeavors to account for the interaction of intra-psychic and social dynamics in the making, and demise, of leaders (DeRue & Ashford, 2010). It suggests that leaders are most effective “when their message is deeply personal and yet touches shared concerns” (Petriglieri, 2011, p. 6); that is, when they are able to legitimately claim leader identities that are congruent with their life story (Shamir & Eilam, 2005) and symbolize what is good and unique about their groups and organizations (Van Knippenberg & Hogg, 2003). We contribute to this field of inquiry by exploring two questions that have received less attention: How do individuals deal with unwanted aspects of themselves in the process of crafting identities that befit coveted (albeit not necessarily formal) leadership roles? And are there unintended consequences—that is, a hidden price to pay—for being able to tailor one’s identity to such roles?
We address these questions from a systems psychodynamic perspective (French & Vince, 1999; Gould, Stapley, & Stein, 2001; Hirschhorn, 1988; Long, 2008). First employed by social scientists at the Tavistock Institute of Human Relations in London, who combined open systems and psychoanalytic theories to advance understanding of organizational and social phenomena (Menzies, 1960; Miller & Rice, 1967), this perspective rests on the assumption that conflicting elements coexist within the self, and endeavors to explain how such conflicts are experienced and managed intra-psychically and in social interactions (Gabriel, 1999). Therefore, it is suited to answer extant calls for deeper inquiry into the multiplicity and dynamic nature of identity and identifications (Albert, Ashforth, & Dutton, 2000) as well as into the intra-psychic aspects of identity work (Kreiner et al., 2006). In addition, by focusing on the interaction of cognitive and emotional, intra-psychic and relational, conscious and unconscious factors, a systems psychodynamic perspective helps to cast a light on aspects of identity and identification that may otherwise remain invisible or unexplained (Petriglieri & Petriglieri, 2010).
The sequence of this paper is as follows. After locating our argument within the field of leadership studies and reviewing existing identity scholarship on leaders’ development, we outline the conceptual framework of projective identification in leaders’ identity work. Here we develop the paper’s central argument: that the mechanism of projective identification (Klein, 1946 [1975]) is likely to be employed by leaders to manage unwanted, often unconscious self-definitions in order to attain or uphold a desired identity. Projective identification, as used in this paper, refers to the unconscious projection of unwanted aspects of the leader’s self into others so that it appears that they, and not the leader, have these unwanted characteristics and the identities they imply. This mechanism shores up the boundary between conscious, desired features of the leader’s identity and its unwanted aspects. In doing so, projective identification reduces leaders’ inner conflicts and enhances their ability to credibly enact identities suited to the demands of their role—hence supporting “successful” identity work. This mechanism, however, also generates unexpected consequences, such as destructive interpersonal conflicts and organizational phenomena. We illustrate our conceptual argument with examples from the evolution of the Gucci family business. We conclude with a discussion of the paper’s contributions and implications for future research.
Identity Work in Leader Development
Leadership is the object of enduring popular fascination, and yet scholars have had a difficult time with it (Nohria & Khurana, 2010). There is no widely agreed-upon definition of leadership and a plurality of views persist on how it should be researched, practiced, or developed (Avolio, 2007)—a state of affairs reflective of our evolving understanding of the concept (Day & Harrison, 2007) and the many disciplinary lenses through which it is studied. Reviewing the field of leadership studies, and/or offering yet another definition of leadership, would be beyond the scope of this paper (for recent reviews, see Alvesson & Spicer, 2011; Glynn & Raffaelli, 2010; Hogg, 2007). Three recent trends, however, are notable. The first is the resurgence of a perspective less preoccupied with leaders’ impact on organizational performance and more with their function as sources and symbols of the values and meaning making of organizational members (Podolny, Khurana, & Hill-Popper, 2005; Smircich & Morgan, 1982). The second is a move beyond the study of traits, behaviors, and contingencies that allow leaders to exert their influence over followers (Reicher, Haslam, & Hopkins, 2005). The third is a transcendence of traditional views of leadership as the preserve of individuals in positions of formal authority (DeRue & Ashford, 2010).
This paper rests upon and contributes to these trends, joining the efforts of scholars who are marking a trail “away from a static and hierarchical conception of leadership and toward a more dynamic, social and relational conception of the leadership development process” (DeRue & Ashford, 2010, p. 629). Rather than ascribing it to individuals by virtue of their characteristics or roles, this conception views leadership as a dynamic relationship between leaders and followers characterized by reciprocal influence (Hogg, 2007). It accounts for observations that leadership can be exercised by individuals or be distributed within groups (Carson, Tesluk, & Marrone, 2007), that those in formal leadership roles are not always followed while people without such roles can and do lead (Hackman & Wageman, 2007), and that the same people can be leaders or followers in different settings and/or at different times (DeRue & Ashford, 2010). The question then is not only what leaders do, but also to whom leadership accrues, and how.
A rich vein of contemporary scholarship examines the emergence and effectiveness of leaders through the lens of social identity theory (for a review, see van Knippenberg & Hogg, 2003), suggesting that “the secret of successful leadership lies in the capacity of the leader to induce followers to perceive him or her as the embodiment of a positive social identity that they have in common and that distinguishes them from others” (Ellemers, de Gilder, & Haslam, 2004, p. 469). This work echoes psychoanalytic conceptualizations of the bond between leaders and followers. Freud suggested that groups confer leadership on those who best embody and articulate group members’ “ego ideal,” that is, an idealized version of themselves (Freud, 1921). Building on this insight, systems psychodynamic scholars have argued that leaders are best located at the boundary between their own and other groups or organizations—where they can represent the group to both insiders and outsiders (Miller & Rice, 1967). These perspectives converge on the idea that leaders’ identities and activities have a symbolic function. They help followers make meaning of their circumstances, of their intentions, and of who they are (Smircich & Morgan, 1982). The characterization of leaders as entrepreneurs of identity (Reicher & Hopkins, 2003) captures the reciprocal influence through which groups bestow leadership on members who best represent the group’s identity and how, in turn, these leaders mold the group’s identity (see also Reicher et al., 2005). This implies that leaders must manage their identities to gain the right to influence others.
Building on this insight, a literature on the role of identity in the process of leader development is emerging (Day, 2001; Day & Harrison, 2007; DeRue, Ashford, & Cotton, 2009; Ely, Ibarra, & Kolb, 2011; Ibarra et al., 2010; Lord & Hall, 2005; Petriglieri, Wood, & Petriglieri, 2011). This work suggests that two central features of leaders’ development are the internalization of a leader identity within the individual’s self-concept and the validation of that identity in social interactions. The intrapersonal portion of the process involves achieving congruence between the individual’s view of himself or herself and his or her view of what leadership is (DeRue et al., 2009). The interpersonal portion of the process involves potential followers granting the individual’s claim to leadership on the basis of its congruence with their view of what leaders should be like (DeRue & Ashford, 2010). This conceptualization accounts for the emergence of formal or informal leadership at all levels of an organization.
A corollary of this argument is that, the more visible and demanding the leadership role, the more pressure there will be from the self, followers, and the public for the person holding the role to embody views of what the leader should be like in representing the group or organization. One can see this pressure, for example, in the scrutiny surrounding leaders’ biographies. We prefer the life story of a central bank chairperson to display poise and stability, whereas we expect the life of a revolutionary to be a tale of dissent with the establishment. Put another way, those aspiring to lead have little choice in positioning themselves vis-a-vis the identities valued by their potential followers—they must embrace them. Hence, leader development for such visible roles is likely to involve a “deep identification” (Rousseau, 1998)—that is, an integration between role requirements and the role holder’s personal identity—so that the person experiences an overlap “between self-at-work and one’s broader self-concept” (Rousseau, 1998, p. 218).
Internalizing and enacting a leader identity, then, involves identity work (Snow & Anderson, 1987; Svenningsson & Alvesson, 2003) aimed at resolving the intrapersonal and interpersonal incongruency between one’s personal identity and the leader identity to which one aspires. Building on McAdams’s (1999) definition of identity as “the internalized and evolving story that results from a person’s selective appropriation of past, present and future” (p. 486, italics added), scholars argue that identity work entails crafting, experimenting with, and revising identity narratives or stories about the self (Ashforth, Harrison, & Corley, 2008; Ibarra & Barbulescu, 2010; Snow & Anderson, 1987). They also argue that individuals can firmly inhabit roles only when they “resolve the conflicts and contradictions” between their potential narratives (Ibarra & Barbulescu, 2010, p. 32).
But what happens to those elements of an individual’s past, present, and future that are left out of the selective appropriation involved in crafting a version of the self suited for visible leadership roles? While some of these elements may perhaps be discarded, others may not be so easily eliminated, and we propose that they may be the matter of a less conscious aspect of leaders’ identity work. Identity scholars have suggested in passing that projective identification (Klein, 1946 [1975]) may be involved in managing the demands of roles that require deep identification (Kreiner et al., 2006). We build on this hint and propose a conceptual framework that articulates why, how, and with what consequences projective identification may be involved in leaders’ identity work.
Projective Identification: A Conceptual Framework
Our argument rests on the notion of the self as dynamic and comprising numerous identities (Markus & Wurf, 1987). Some of these identities are related to personal characteristics, others to roles held and group memberships (Gecas, 1982). All are shaped and refined in social interactions (Cooley, 1902; Mead, 1934) and include actual as well as possible selves based on one’s past history or images about who one might become, could have been, would like to be, or fears becoming (Markus & Nurius, 1986; Obodaru, 2012). Possible selves serve as points of orientation for identity work (Ibarra, 1999).
While leaders are likely to work hard to actualize and maintain selves that reflect what is unique about, and valued by, members of their organizations (referred to here as “wanted selves”), there also exists a reservoir of selves that they do not like or wish to become, as becoming that person would make them ill-suited to leading in their social context (which we refer to as “unwanted selves”). While “wanted selves” may often be selves broadly held in positive regard, in rather different and more extreme cases leaders may idealize and enact destructive selves (Rosenfeld, 1987). This somewhat perverse way of gaining and exerting power may occur, for example, in the case of gang leaders whose acceptance by members of the gang may hinge on displays of ruthlessness, lack of remorse, and social deviance. 1 Regardless of their specific contours, unwanted selves are powerful elements in the psychic economy and are more likely than desired selves to contain elaborations based on embarrassing past experiences (Ogilvie, 1987). Individuals assess their well-being more on the basis of how distant they are from unwanted selves than on how close they are to ideal ones (see also Carver, Lawrence, & Scheier, 1999).
The conclusion that individuals who feel too close to unwanted selves experience distress is common to identity research from different traditions. Building on a post-structuralist perspective on authenticity as a “workable fantasy of a unique and coherent self,” Costas and Fleming (2009, p. 358) described two modes of experience among a consultancy’s employees. Some were able to draw a boundary between their “authentic” selves, enacted outside of work, and the unwanted selves prescribed by the organization. Disidentification from the latter afforded them a modicum of dignity, even if nursing hidden, truer selves may well have helped them integrate within the fabric of the firm (Fleming & Spicer, 2003). Other consultants suffered a more problematic fate. Their experience, a contemporary form of self-alienation, revolved around the unhappy awareness that the boundary between “who they really were” and the unwanted corporate self had failed. Commenting on these findings, Ybema et al. (2009) suggested that what makes an identity more or less real, more or less actual, is the “continuing capacity to enact [it]” (p. 306). This capacity, we argue, is sustained by conscious and unconscious work.
In crafting a personal identity suitable to the demands of leadership roles, individuals may be confronted with contradictions and aspects of their history, affiliations, behaviors, and imaginary self-conceptions that they find unpalatable because they threaten, at least unconsciously, the congruence of the leader’s self. Social identity theorists suggest that only one version of the leader’s self will be active at any given time (van Knippenberg, van Knippenberg, de Cremer, & Hogg, 2004), a view that has recently been challenged (Swann, Gomez, Seyle, & Morales, 2009). We propose that, while at a conscious level only one version of the self may be active at a time, other versions may simultaneously be activated at an unconscious level. Therefore, the more the leader’s conscious identity work involves striving to craft and maintain a wanted version of the self, the more unwanted selves are likely to be worked on unconsciously.
One way of dealing with unwanted selves that are incongruent with leader identity requirements is through the use of what Melanie Klein (1946 [1975]) referred to as “projective identification,” a concept that builds on Freud’s (1984) idea of projection and has gained wide currency and generated much debate in psychoanalytic circles (for different perspectives see Goldstein, 1991; Ogden, 1979; Sandler, 1987). When resorting to projective identification, individuals unconsciously split off certain aspects of themselves and project them into others. These others are then experienced as having the characteristics that have been projected into them, and the individual who is doing the projecting unconsciously identifies with them (Klein, 1946 [1975]). Klein argued that projective identification involves “splitting” the self into “good” and “bad” so that either unpalatable aspects of the self or, conversely, desired aspects of the self may be projected—leading respectively to negative or positive identification with the recipients of the projections. Much of the psychoanalytic literature focuses on the problematic nature of the former predicament, as we do here.
Projective identification is first used in infancy as a form of emotional regulation and communication, a way for children to manage affect and, at the same time, let their caretakers know how they feel. During child development, projective identification is gradually modified in response to parental “containment” so that later in life it may be generally employed as a means of non-verbal communication of internal states (Bion, 1967), rather than in the more problematic form discussed here. The role of projective identification in sustaining followers’ idealization or denigration of leaders has long been recognized. Either through identification with, or cynical detachment from, a leader imbued with extraordinary powers, followers can shield themselves against anxiety and ambivalence (Gabriel, 1999; Schwartz, 1990). Here we conceptualize the role of projective identification in relation to another side of the unconscious dynamic that binds leaders and followers, namely, the way leaders sustain identities that match the collective expectations embedded in their roles.
Projective identification is never a conscious strategy, but rather an unconscious operation as instantaneous and compelling as it is out of awareness and control. While it can never be fully controlled or captured in a conclusive process model, it can be examined and struggled with as we attempt to do here. We suggest that several factors may unconsciously ignite leaders’ engagement in projective identifications. One is the need to protect themselves from consciously experiencing unbearable feelings, in which case projective identification functions as a defense mechanism (Feldman, 1992). With their dislike of their unwanted selves now directed toward others (Klein, 1946 [1975]), leaders are unconsciously relieved of the affect associated with unwanted selves and less conflicted in the expression of wanted ones. Projective identification may also be motivated by the desire to control and dominate another (Rosenfeld, 1987; Joseph, 1989). A leader who projects unwanted qualities into a follower exercises control by evoking those qualities in that person and/or by imagining him or her to have those qualities. Projective identification may also be motivated by envy. By making recipients appear to possess despised characteristics (Rosenfeld, 1987), leaders are liberated from envious feelings toward them. In this sense, projective identification may be both a defense against envy and an enactment of it (Rosenfeld, 1987). Finally, leaders may employ projective identification to extrude unwanted selves that are inconsistent with their followers’ expectations. In doing so, they are freer to introject those expectations and become the leaders their followers, more or less consciously, want them to be.
While projective identification allows leaders to internalize and enact identities that befit their roles, it also creates ongoing difficulties. Leaders are unlikely to work effectively with those who are felt to embody their unwanted selves. Given the role of trust (Burke, Sims, Lazzara, & Salas, 2007) in establishing productive relationships between leaders and followers, projective identification may diminish the extent to which leaders feel they can depend on others. This is exacerbated by the likelihood that leaders who project into others will experience paranoid (Rosenfeld, 1987) or persecutory anxieties (Bott Spillius & Feldman, 1989), which result in lingering fears of retaliation by the recipients of their projections.
While it is possible to engage in projective identification in relation to a distant recipient who is not affected by it, a leader’s projections usually affect nearby recipients deeply because they are manipulated to introject and enact the leader’s unwanted self—and are drawn into ongoing conflict. This is known as the “evocatory” aspect of projective identification (Bott Spillius, 1988). Being the recipient of painful, palpable projections, the other person may feel impelled to unburden himself or herself by engaging in the unconscious return of those projections into the leader or other people. Such returning of projective identification, or unconscious enactments (Bott Spillius & Feldman, 1989), may manifest themselves as vengeful retaliation against the leader—a kind of unconscious retributive justice. The result is that both sides spend much energy attempting to lodge the projections into each other, while on the surface their relationship appears stuck and ossified.
In spite of the discomfort and conflict, leaders engaging in projective identification are likely to feel compelled to remain in proximity with those into whom they have projected their unwanted selves. Such proximity provides them with ongoing opportunities to compare themselves favorably with the recipients of the projections, to deny their unwanted selves, and to attack and attempt to destroy the unwanted selves lodged in others. Having nothing to do with these others would not provide an adequate solution to the intra-psychic conundrum because the leaders would thereby relinquish the opportunity to deny, control, and attack the unwanted selves in others rather than within themselves. Projective identification thus transforms inner conflicts into interpersonal ones. Leaders may thereby develop a “compulsive tie” (Klein, 1946 [1975]) to these others and become interminably entangled in trying to extrude or destroy what cannot be extruded or destroyed: the unwanted parts of the self (Bott Spillius & Feldman, 1989).
Such processes may lead to the development of toxic environments in which problems that are created in one area are systemically transferred to others. Recipients of projective identification who feel impelled to return these projections to the leader or to a third party are likely to become embroiled in ongoing, damaging struggles that can become toxic (Maitlis & Ozcelik 2004; Stein, 2007). Since leaders function as sources of meaning making (Podolny et al., 2005; Smircich & Morgan, 1982), the unconscious use of others as recipients of unwanted aspects of the self may become a collective modus operandi that damages the organization and may even cause its destruction. In the following section we illustrate this conceptual framework by drawing on the history of the Gucci family business.
Projective Identification: A Case Example
Founded by Guccio Gucci at the turn of the 20th century, Gucci began as a modest family business and grew to become a global fashion powerhouse by the 1970s. During the following two decades, however, the family descended into an acrimonious struggle that resulted in tragic consequences. Multiple conflicts between members of the second generation (the brothers Aldo and Rodolfo Gucci) and the third generation (Paolo, Giorgio, and Maurizio Gucci) led to a series of costly court cases and a breakdown in the capacity to work effectively. To summarize, Aldo, who presided over Gucci’s international expansion, was ousted from office by Paolo (his son) and Maurizio (his nephew) and sent to prison at age 82 on the basis of evidence of tax evasion provided by his son. Paolo himself remained in bitter conflict with the rest of the family—being sacked and reinstated on several occasions—and died in tragic circumstances in 1995. Maurizio, who became the last Gucci to serve as the firm’s CEO, was pursued in the courts by the rest of the family across the United States and Italy. Shortly after being forced to sell the business, he was murdered. His ex-wife was tried and found guilty of commissioning the murder. By then, the family was in chaotic circumstances and held no share in the firm (Forden, 2000).
We chose the case of Gucci as an illustration because the individuals’ identity work to fit coveted leader roles—and the associated dynamics and unintended consequences—is highlighted in a particularly vivid way. In a family business carrying the family name, the boundary between personal and work identities is likely to be thinner than in most cases. Arguably, fashion industry leaders’ visibility, and the pressure on them to introject public expectations and embody the organizations’ identity, may be higher than in other industries. Finally, the case is set in a context of great and longstanding success. This is the exception rather than the rule for fashion houses that carry the founder’s name, as in these creativity and experimentation often take precedence over sound administrative practice (Corona & Godart, 2010).
Gucci is one of a handful of firms that defined the fashion industry as we know it today. Because of the dramatic nature of the events and the iconic status of its brand, there are a wealth of public accounts documenting the history of the Gucci family and its business. Those we consulted included books based on primary data (Forden, 2000; McKnight, 1989; Pergolini & Tortorella, 1997), including one written by a member of the family, Paolo’s wife, Jenny (Gucci, 2008). Forden (2000), for example, interviewed approximately 100 family members and associates. We supplemented those books with newspaper articles in both English and Italian. We draw examples from these secondary sources not to “build” or “confirm” the conceptual framework presented in this paper. Rather more modestly, we aim to illustrate the framework’s potential value to understand (at least one layer of) the complex predicaments of one organization and its leaders. To select the vignettes we offer below, both of us acquainted ourselves with the texts above, and developed an initial set of instances that seemed to illustrate projective identification linked to the development and maintenance of leaders’ identity. We debated these in two days of conversation followed by another round of separate reviewing. During three further meetings we narrowed our choice of illustrations, and debated whether and how an interpretation based on the framework presented above offered a richer understanding of accounts than that afforded by alternative explanations.
There are risks and limitations in what we set out to do. A conceptual paper with an extended case illustration fits neither of two established genres in academic writing—the “theory paper,” where the theory’s application is left to the reader, or the “empirical paper,” where data is the ground on which the theory is developed or tested. If the format of this “theory paper” does challenge a norm, it does so to offer a template that is, in our view, well suited to portraying the value and limitations of a system psychodynamic perspective on organizational phenomena. In keeping with this perspective, our main heuristic instrument is interpretation, which is always provisional and intended to provoke and open up understanding—rather than to capture essential truths and draw definitive conclusions. The value of a psychodynamic framework lies in enabling the articulation of a layer of meaning (among many) that gets, so to speak, a conversation going.
Interpreting texts—from primary or secondary sources—inevitably implicates and exposes the subjectivity of the interpreter, despite the availability of multiple accounts of the same events. The accounts we examined may be interpreted differently depending on the perspective and purpose of the examination. Here we suggest that one way to understand these events, and the dynamics of what scholars have identified as a successful yet problematic family business (Stein, 2005; Kets de Vries & Simmons, 2010), is through the lens of projective identification aimed at crafting and sustaining leaders’ identities.
Creating the leader’s identity
We turn first to Aldo, a son of the founder of Gucci. Aldo was at the helm of the firm for several decades, and many accounts exist of his creative flare, bold strategic decisions, and temperamental leadership style. As we discussed earlier, however, the focus of this paper is not leadership as position, strategy or style. We are concerned with the identity work that allows leaders to sustain their role at the boundary—in this case—between the family, the firm, and the public. Aldo appears to have crafted two related identities for himself that, we argue, allowed him to emerge as a leader and be regarded as “the driving force” in the transformation of his family’s leather workshop into a global brand (Forden, 2000, p. 30).
One identity concerned the idea that Aldo was a highly aristocratic person, part of the “Tuscan Royal Family” with himself as monarch. He had, in fact, come to be known as L’Imperatore (the Emperor) in certain circles (Forden, 2000, p. 66), with many—including reputable newspapers—speaking of the family as having “blue blood” (Forden, 2000, p. 120). A second, related identity cast Aldo as a moral authority with unwavering integrity. As he exclaimed during an interview, “I want to be like the Holy Father” (Forden, 2000, p. 94). These identities were highly symbolic of the exclusivity, purity, and impeccable style with which the Gucci brand came to be associated.
Neither identity, however, had much grounding in Aldo’s personal history or business practice. The family had no genealogical connection with royalty. Indeed, Aldo’s father and the founder of the business, Guccio Gucci, had been born into poor circumstances. When his own father’s business went bankrupt, Guccio made his way to England, where he worked at London’s Savoy Hotel before returning to Florence. There he married, had a family, and lived in modest circumstances (McKnight, 1989, pp. 17–18). It was only in 1922—when Aldo was 17—that Guccio started his leather workshop and store. Initially, he struggled to build this up, coming within a hair’s breadth of closing it down shortly after it opened (Pergolini & Tortorella, 1997, pp. 56–57). Although these hardships befell Guccio, they also impacted Aldo, who, throughout his youth, was exposed to the uncertainties of his father’s work and to the anonymity associated with it.
Aldo’s identity as a moral authority stood in sharp contrast to his business conduct. Under his leadership, millions of dollars in taxable revenues were “siphoned to offshore companies under a system of false invoicing” (Forden, 2000, p. 86). The problem came to a head when an executive trained in the law discovered “massive fraud” at every level of the company (Forden, 2000, p. 106). When he tried to persuade Aldo of the gravity of the situation, the latter responded that, because he had built up the company, he “deserved to get something back” (Forden, 2000, p. 107).
Projective identification
The “aristocratic” and “moral” leader identities served Aldo well in his role as a symbol of all that was special and unique about Gucci, and they were socially reinforced by his induction into a circle that did include royalty and celebrities. Sustaining these identities, however, required keeping his discrepant history and practices—and the identities they implied—at bay. We suggest that Aldo could persuade himself that he fit the desired “aristocratic” and “moral” identities, and could credibly enact them, because he had disowned those much less appealing, hence unwanted, identities and projected them into others.
In light of our theorizing, it may be argued that Aldo dealt with potential feelings of not being special—highly discordant with an “aristocratic” self—by projecting them into other people. In public, for example, he often referred to his first son Giorgio as “the black sheep of the family” (Gucci, 2008, p. 87) and repeatedly told his wife that she was “a nobody, a nothing” (Gucci, 2008, p. 82). Aldo may also have been able to sustain his “moral” self by projecting the most controversial aspects of his practices into others. For example, while the US Internal Revenue Service was investigating his financial affairs (Forden, 2000, pp. 105–106), Aldo accused his nephew Maurizio of receiving his inheritance by fraud (Forden, 2000, p. 117). In addition, Aldo sent to the Italian chief prosecutor, fiscal police, tax inspection office, and Ministries of Justice and Finance, documents purporting to reveal how Maurizio had financed the purchase of his yacht with illegal funds (Forden, 2000, p. 126). In short, we suggest that constantly finding and eliciting the “nobody” and “criminal” among close others was instrumental to sustaining Aldo’s “aristocratic” and “moral” identities. This may have bolstered his leadership, but at a price.
Returning projective identifications
As we described earlier, the process of projective identification is likely to be unconscious and the consequences significant—especially if the recipients do not recognize that they are being used as a character in the leader’s drama, or the leader is close to them, or the projections resonate with aspects of their own identity. In some cases, recipients may introject the projections and collude in enacting the leader’s perception of them. In others, particularly when projective identification triggers anxieties about their own identities, recipients may feel compelled to get rid of the projected elements and “return” them to their source.
While many at Gucci appeared to accept Aldo’s projections, the “return” dynamic may have fueled the conflict between Aldo and his son Paolo. Like his father, Paolo had a problematic ethical record. For example, following his divorce, he was arrested for failure to pay alimony (Gucci, 2008, pp. 225–226) and subsequently returned to jail for contempt of court (Gucci, 2008, p. 251). Further, amid an extended conflict with the family, Paolo was sacked but kept receiving a salary illegally (Gucci, 2008, p. 136). When he died, it emerged that Paolo was worth around £30 million, with “most of his money … in secret Swiss bank accounts and his properties … in the name of obscure offshore companies” (Gucci, 2008, p. 272). Because of his own problematic record, we suggest, Paolo was unable to tolerate Aldo’s projections and responded by returning them—putting much effort into exposing the purported corruptness of his father. Paolo spent a decade trying to send Aldo to prison. This involved ten court cases, the defense of which cost Gucci millions of dollars in legal fees (Forden, 2000, p. 78). Later, Paolo and his cousin Maurizio hatched a plot that ousted Aldo from office (Pergolini & Tortorella, 1997, pp. 99–102). Finally, Paolo succeeded in getting his father sent to jail, where he served four months of a one-year term (McKnight, 1989, p. 196). Aldo had been forceful in painting his son as starkly different from himself, calling him a “son of a bitch” and “crazy” (Forden, 2000, p. 83), firing him repeatedly, and claiming that his ventures were illegal. Paolo could thus be understood to have retaliated by returning the undesired criminal identity into his father while claiming to be above illegal activity and highly “moral” himself; hence, claiming to be a better fit for leading the firm.
Projective identification and toxicity
The recipients of projective identification may not just introject or “return” the leader’s projections but may also become inclined to project into others. This can fuel a destructive cycle that impacts an entire organization. Through emulating leaders, projective identification may become the prevalent means of making sense of self and others at all levels of the organization. This can result in a toxic culture in which anyone’s identity is bolstered through the manipulation of someone else. All contact is then experienced as poisonous, with trust and collaboration becoming all but impossible.
At Gucci, we suggest, projective identification may have been intertwined with the long-running, multiple fights that involved not only Aldo and Paolo but also Rodolfo (Aldo’s brother), his son Maurizio, and several other family members and colleagues. Rodolfo also carved out an “aristocratic” identity, which may have been bolstered by projecting his own working-class origins into others. For example, he objected to his son’s choice of Patrizia as his wife, calling her “a social climber” (Forden, 2000, p. 42) who was not of their class. This toxic culture also affected the relationship between Paolo and Maurizio, the latter of whom told investment bankers that Paolo was “a complete liability” and his other cousin, Giorgio, was “totally hopeless”—describing them as “Pizza brothers” (Forden, 2000, p.141). For his part, Paolo did his best to expose Maurizio’s illegal activities. He passed “papers” on to the Italian authorities that enabled them to raid Maurizio’s home and office (Gucc,i 2008, p. 247). Maurizio retaliated by sending the police to break up the launch party for Paolo’s “P.G.” brand (Pergolini & Tortorella, 1997, p. 120). The acrimony of these encounters led to a downward spiral that precluded the possibility of effective work.
Such toxicity was by no means confined to family members. Gian-Vittorio Pilone, Maurizio’s chief adviser (McKnight, 1989, p. 199), was centrally implicated in Maurizio’s conflict with his wife Patrizia and his cousin Paolo (McKnight, 1989, pp. 114, 189). Another non-family member, Domenico De Sole, was appointed by Maurizio as head of Gucci US (Forden, 2000, p. 111) and ended up vehemently antagonizing Aldo, whom he claimed was guilty of “massive fraud … [and would be] … going to jail” (Forden, 2000, p. 106). Later, when Maurizio removed him from Gucci US (Forden, 2000, p. 111), De Sole took Maurizio to court (Forden, 2000, p. 217). On occasion, even Gucci employees treated customers contemptuously. New York Magazine, for example, ran an article about the Gucci shop assistants’ “drop-dead put-down” under the headline “The Rudest Store in New York” (Forden, 2000, p. 66). As Forden (p. 66) put it, “My-Gucci-story-is-more-outrageous-than-yours” became a familiar discussion point in elite New York circles.
These examples, which we use to illustrate our conceptual framework of projective identification in leaders’ identity work, could also be interpreted through other lenses. One alternative explanation is that, rather than exchanging projections of unwanted aspects of themselves, Gucci members simply voiced accurate views of each other’s character. Besides denying the social nature of identity, this reading does not account for the attempt to deny those identities in oneself, which was widespread in this case. This may be explained by the theory of “social comparison” (Festinger, 1954), which suggests that individuals in part bolster their self-definitions by distancing themselves from others who are portrayed as different. Social comparison, however, is a cognitive process and does not account for the emotional intensity of the attacks or the entanglement with denigrated others.
Among psychodynamic theories, an alternative explanation for the dynamics described above is that they were manifestations of sibling rivalries and Oedipal conflicts in a dysfunctional family (Kets de Vries, Carlock, & Florent-Treacy, 2007). This might be possible. We chose illustration from a family business, as noted above, because the pressure to align personal and leader identities, for family members in the firm, is particularly visible. The flip side of this choice, however, is the difficulty of ascertaining where family dynamics end and organizational ones begin. These dynamics, however, were by no means limited to parents and children, siblings, or family members alone. In addition, even if those involving the family were indeed Oedipal conflicts, what was fought for was exclusive claim of the identities that legitimized leadership in the eyes of Gucci’s employees and public, not simply among Gucci relatives. At a minimum, these constituencies’ expectations of Gucci leaders amplified, and dictated the prize of, family conflicts.
Finally, a simpler explanation might be that the dynamics described here were conscious attempts to create a carefully constructed image to serve personal economic interests. This also might be possible. However, this argument does not account for the compulsive ties within these relationships. Had Aldo simply aimed consciously to deceive others about his humble origins, he would not have needed continuously to diminish others. Had he just consciously tried to cover up his illegal financial activities, he might have been better served by turning a blind eye on supposed wrongdoing in other parts of his firm. Had Paolo simply wanted to take control of the company and its resources, he would have stopped trying to put Aldo in jail after his father was no longer in charge, especially given the significant financial costs involved in prosecuting Aldo. While relationships in the family and firm became increasingly acrimonious and damaged the Gucci organization in various ways, it seemed difficult for anyone to leave or break the cycle. Even while family members accused each other over the approaching demise of the firm, they did not leave or sell their shares when they were still worth a fortune. We suggest that one reason for their reluctance to do so was that their identity was inexorably linked with, and dependent on, both the rewards of their leader identities and the projection of unwanted selves into others, whom they then could not let go.
Discussion
The conceptual framework presented in this paper casts a light on unconscious aspects and unintended consequences of what may, from a social cognition perspective, be deemed “successful” identity work, that is, work that results in experiences of authenticity and social validation. Our argument contributes to the literature of four areas: identity work, organizational identification, the psychodynamics of leadership and organizations, and leader development. It also suggests potential avenues for future research.
Contributions to theory
Identity work
We contend that a system psychodynamic perspective on leaders’ identity work can complement scholarship based on symbolic interactionism and social cognition (DeRue & Ashford, 2010). It has long been suggested that identity work is ongoing but only consciously undertaken when one encounters novel or surprising situations (McAdams, 1999; Van Maanen, 1998). We argue that projective identification bolsters unconscious identity work at times when leaders are working to consolidate and maintain their desired identities. This complements existing work focused on how individuals craft and pursue desired identities within cultural scripts and discourses (Ibarra, 1999; Ibarra & Barbulescu, 2010; Pratt et al., 2006; Svenningsson & Alvesson, 2003; Thornborrow & Brown, 2009) and in spite of stigmatized roles (Ashforth & Kreiner, 1999; Snow & Anderson, 1987). By focusing on the fate of unwanted selves in a leader’s identity work, we contribute to work that, to date, has paid more attention to the crafting and maintenance of desired identities than to how individuals in coveted roles “work” on unpalatable elements of their self-concept.
Identification
Research on identification has recently moved beyond focusing on the individual’s relationship with a collective (organization or social category) to considering the influence of interpersonal phenomena on identity and identification. Organizational scholars have generally sidestepped exploring personalized relationships at work (Sluss & Ashforth, 2007) and called for more research on the dynamics underpinning harmful work relationships (Gersick, Bartunek, & Dutton, 2000). We propose a link between negative interpersonal identification, sustained by projective identification, and positive organizational identifications. Specifically, we argue that to craft an identity that befits a coveted leadership role, individuals are likely to unconsciously develop problematic relationships with people who embody their unwanted selves. Hence, the more identified a leader becomes with an organization, the more likely he or she will be to engage in projective identification to reduce the gap between his or her personal and organizational identities.
This expands views of over-identification as a pathology of organizational identification. Dukerich, Kramer, & McLean Parks, (1998) highlighted the negative consequences of over-identification for the individual, such as diminished willingness to question organizational practices and take responsibility and/or increased vulnerability to identity threat. Over-identification is riskier for individuals in “highly visible, high status, and intrinsically motivating roles, which offer highly seductive identities for their incumbents” (Ashforth et al., 2008, p. 338). Our argument suggests another problematic effect of organizational over-identification, namely, the unconscious manipulation of others and the resulting development of interpersonal conflicts and toxic cultures. In other words, just as over-identification “may be a substitute for something that is missing in one’s life” (Dukerich et al., 1998, p. 254), it may also generate pressure to project aspects of one’s life into others.
Psychodynamics of leadership
Psychodynamic scholars have articulated the relationship between the character of leaders and organizational cultures (Kets de Vries & Miller, 1984) and unveiled a host of unconscious and irrational dynamics underpinning organizational phenomena (Long, 2008; Menzies, 1960; Stein, 2005, 2007). But while identity development has been a central concern for clinical psychodynamic authors (Erikson, 1959 [1980]), little has been written on identity dynamics in organizations from a psychodynamic perspective, with a few exceptions (Brown, 1997; Brown & Starkey, 2000; Dubouloy, 2004; Petriglieri & Petriglieri, 2010). The clinical literature on family therapy has long described projective identification as the mechanism underpinning the unconscious manipulation of others to sustain a desired identity (Waddell, 1981). We suggest that a similar dynamic underpins the unconscious identity work of leaders, who manage to maintain a positive identity not despite, but because of, denigrated counterparts. We also articulate the costs of this operation. This complements existing work on projective identification in followers’ relatedness to leaders (Gabriel, 1999; Halton, 1994; Schwartz, 1990).
Leader development
We contribute to the literature on identity work in the process of leader development (Day & Harrison, 2007; DeRue & Ashford, 2010; Ely et al., 2011; Ibarra et al., 2010; Lord & Hall, 2005; Petriglieri et al., 2011) by arguing that the internalization and enactment of a leader identity can generate inner conflicts between the features experienced as requisite of the leader identity and aspects of one’s history and behavior. Researchers have suggested that some leaders resolve this discrepancy by developing “before and after” narratives that highlight a turning point of personal transformation, to distance the past and embrace the future (Ibarra & Lineback, 2005; McAdams, 1999), “crucible” narratives that integrate the discrepant feature of the self and portray it as a moment of revelation (Bennis & Thomas, 2002), or narratives that feature a moment of personal suffering as pivotal in the development of transformational leaders’ vision and purpose (Parameshwar, 2006).
We postulate that when leaders are unable to sustain such narratives, they are likely to split off negatively charged self-conceptions—especially those experienced as impinging on the demands of the leader role—and engage in projective identification to keep such self-conceptions at bay. This suggests that it is not only psychological resources or infantile experiences with parental containment that determine whether an individual will be able to integrate problematic aspects of the self into his or her leader identity. Whether or not leaders resort to projective identification also depends on the degree of containment provided by followers and other stakeholders in the present. The more focus an industry, organization, or group puts on a leader to be a symbol of the organization, the more pressure there will be on the leader to develop a fitting identity and to project discrepant aspects of the self into others. Conversely, the more leaders are surrounded by others who help them develop integrative identity narratives and who are able to authorize a leader whose identity is complex and multifaceted, the less need leaders will experience to engage in projective identification.
Future research
Future research may profitably build on the arguments developed in this paper by focusing on a number of areas. The self has received much attention in the last two decades as a conceptual domain in which dialogue, if not integration, can occur between the often diverging perspectives of psychodynamics and social cognition (Curtis, 1991; Westen, 1992). The operation of classic projection first described by Freud, for example, has been demonstrated in a series of laboratory experiments (Newman, Duff, & Baumeister, 1997). Similarly, its function as a defense has been supported (Schimel, Greenberg, & Martens, 2003). A vast body of research has confirmed the conceptual cornerstone of psychodynamics—that much mental functioning occurs below the surface of awareness (Barsade, Ramarajan, & Westen, 2009).
It is unlikely that projective identification will ever be replicated in a psychology laboratory. This is not only because of its unfolding in the long term, but also because “laboratory studies” and “projective identification” belong to different epistemologies with distinct assumptions, discourses, and practices (Long, 2001). Researchers aiming to investigate the phenomenon empirically will need to employ naturalistic, qualitative methods—combining, for example, the collection of individuals’ autobiographical material with repeated observations of their relationships in work roles. Using such methods, researchers could endeavor to unveil the operation of projective identification in leaders’ identity work by comparing features of the self that leaders are embarrassed of, with features they denigrate in people with whom they have conflicts at work.
Other useful settings in which to investigate projective identification in the emergence and practice of leaders are group relations conferences (Miller, 1989; Rice, 1965 [1999]) and experiential leadership development courses designed to serve as “identity workspaces” (Petriglieri, 2011). These temporary institutions provide a magnifying glass on a host of unconscious personal and social dynamics and legitimate their exploration and interpretation. They are, therefore, eminently suited to researching the phenomena described here both as they relate to the formal leadership of these events (i.e., their staff) and to the informal leadership that emerges within their bounds.
Research in this area might focus on articulating what personal resources and social contingencies influence the likelihood that leaders resort to projective identification to sustain their identity. It would also be useful to investigate the relationship between projective identification and well-established motivators of identity work, such as role transitions (Ibarra, 1999) or identity threat (J. L. Petriglieri, 2011). Finally, scholars might enrich and develop the conceptual framework presented in this paper into a process model—including antecedents, moderators, and outcomes of leaders’ identity work—that predicts when leaders are likely to develop and claim integrative life narratives and when they may resort to projective identification to disown unwanted portions of their life story.
Conclusion
The arguments presented in this article underscore the importance of self-awareness and self-management in leader development. We contend that leaders who do not cultivate spaces for reflection amid the turmoil of organizational life may fall victim to pressures to over-adapt, thus potentially resorting, unconsciously, to projective identification to deal with unintegrated and unwanted parts of themselves. This, in turn, will limit their effectiveness and contribute to conflicts in their workplace. In addition, we suggest that self-reflection alone is not enough. When leaders operate under great visibility and pressure, they will likely need support from responsible followers and outside professionals to minimize the chances that they will unconsciously resort to projective identification and thus experience its consequences. Interventions aimed at reducing the destructive phenomena described in this paper will need to focus at both the individual and the group/organizational levels. Day and Harrison (2007, p. 363) argue that “what is missing in most [leadership] development initiatives is the interpersonal context.” In line with that observation, we suggest that only by addressing both individual leaders’ reflective capacity and followers’ expectations—through individual and group-based learning opportunities—can we begin to diminish the occurrence and problematic consequences of projective identification.
Footnotes
Acknowledgements
We are grateful to Jonathan Gosling, William Halton, Otilia Obodaru, and Jennifer Petriglieri for their thoughtful feedback on earlier drafts of this paper. We would also like to acknowledge the helpful comments of participants in the stream on “Psychoanalysis in search of meaning: Love, hate and desire for knowledge in organizations” at the 2009 EGOS Colloquium in Barcelona, in the INSEAD Leadership Brown Bag Seminar series, and at the 2010 Academy of Management meeting in Montreal. Finally, we are grateful for the guidance provided by Special Issue Editor Marianna Fotaki and three anonymous reviewers in developing the manuscript.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.