Encountering the Arugula Leaf: The Failure of the Imaginary and its Implications for Research on Identity in Organizations

The article reviews research on identity in organizations. It suggests that current research reiterates imaginary constructions of identity by which identity can be defined as coherent or fragmented. Based on a psychoanalytic understanding of subjectivity, it explores how articulating identity as lack may unsettle such imaginary constructions. The article develops the significant implications this has for how identity is conceptualized and researched and, importantly, how the failure of imaginary identity constructions relates to resistance and control in organizations. The article provides new directions for the study of identity in organizations particularly with respect to widening the discursive spaces in which creative identity struggles occur.

The Failure of the Imaginary and its Implications for Research on Identity Michaela Driver I develop in the article invites us to explore this failure and the creative identity struggles that accompany it when lack is articulated rather than covered up. This in turn provides new directions for identity research as an exploration of failures of the imaginary and importantly as a discursive space in which the absent subject can be experienced. Therefore, an important contribution to be made here is to develop a perspective through which identity research can relate differently to lack and conceptualize control and resistance in organizations in relation to how lack is articulated in identity discourse and how it may absorb, suck up or (re) appropriate larger, potentially controlling discourses.
The article proceeds as follows. First, I examine how the challenge of identity work is currently addressed in organizational research particularly in view of recent empirical investigations and the set-piece approach to coherence, fragmentation and refl exivity. Second, I review key concepts of Lacanian psychoanalysis (Lacan, 1977a(Lacan, , 1977b(Lacan, , 1988a(Lacan, , 1988b. Third, I discuss what these may tell us about how identity is theorized and explored and conclude by discussing implications for the theory and practice of identity research.

Empirical Research and the Set-Piece
I now examine how the challenge of identity complexity is dealt with in current research, particularly empirical research, on identity work in organizations. In this review I will focus on studies that investigate how the self is constructed in everyday language (Potter and Wetherell, 1987) and how subjects position themselves in interactions with others by drawing on different discourses (Davies and Harre, 1990). That is, I focus on research that conceives of identity in relation to discourse as 'a more or less integrated, prefabricated line of using language and reasoning in which the phenomenon [here identity] is constructed rather than revealed or mirrored' (Sveningsson and Alvesson, 2003). My review will try to map out fi rst how the themes of coherence and fragmentation are played out in current research in relation to the research fi ndings but also authorial refl exivity.
I begin my review with a study of managerial identity in which ethnographic materials furnished the contextual framework for an in-depth case study of an administrative manager, referred to as H, at a multinational fi rm in the high tech industry (Sveningsson and Alevesson, 2003). The theme of identity coherence emerges in this study as the authors fi nd that H draws on a relatively stable and coherent identity, which provides security from more ambiguous, contradictory and fragile constructions. Particularly, H's identity as a farmer provides this security but also the ability to resist various organizational discourses. The theme of fragmentation emerges as H's stable self-identity creates tensions and confl icts with dominant discourses of globalization, creativity, networking and management control found in her organization. As she defi nes herself in opposition to some but not all of these discourses, what seems to be a stable identity is continuously in tension with them, which also creates insecurity for her.
The authors conclude that the study provides evidence for both coherence and fragmentation by drawing on the concept of self-identity, which they defi ne as a more stable, coherent core that may be contrasted with more superfi cial and therefore more fragmented identity narratives. Additionally, they conclude that identity work is a force for integration and an increased susceptibility to conformity and managerial control as well as fragmentation and the resistance to organizational discourses and therefore managerial control. The authors suggest that these fi ndings point to the complexity of identity work and refl ect on the need to get thicker descriptions of identity processes via in-depth studies perhaps focusing on a single individual while taking an open-minded approach that avoids reducing identity work to the labelling of social categories, such as organizational identity or gender.
By refl ecting on the: 'need to listen carefully to the stories of those we claim to understand' (Sveningsson andAlevesson, 2003: 1177), the authors propose that the complexity of identity work may be unravelled more constructively through a thicker and more focused approach. Yet, their own study does not seem to go much beyond previous research establishing that identity is indeed coherent as well as fragmented and a force both for control as well as emancipation (Alvesson and Willmott, 2002). In fact, one might even make the case that their study represents a certain regression as the concept of self-identity developed in the study seems to indicate recourse to conceptions of identity as a pure or inner core that were previously refuted (Karreman and Alvesson, 2001).
But the point here is not to deconstruct individual studies of identity. Rather the point is to show how current conceptions of identity complexity, as, for example coherence versus fragmentation, are not only played out in current empirical research but also constrain it to some extent. In particularly, empirical studies rarely seem to escape from documenting recurring identity themes of coherence and fragmentation and, as a result, generate somewhat limited implications. A common implication for control and resistance is that the latter is always possible but we do not know what specifi c role identity plays in either, only that some identity dynamics seem more conducive to control while others seem more conducive to resistance and this varies from study to study.
For instance, in a discourse-oriented study of identity in an IT consulting fi rm, the authors fi nd the identity discourse of the consultants to be remarkably coherent as they identify quickly and seemingly without resistance with hegemonic organizational discourses (Karreman and Alvesson, 2004). As the consultants internalize these discourses defi ning what it means to be a competent and high-performing member of the fi rm, the study shows how bureaucratic and normative controls work in tandem. This in turn makes identity work unlikely to lead to resistance but highly likely to lead to dysfunctional behaviours such as consulting work that is neither creative nor of high quality (Karreman and Alvesson, 2004). In a later The Failure of the Imaginary and its Implications for Research on Identity Michaela Driver review of this and related studies, these fi ndings are somewhat modifi ed by suggesting that while processes of subjugation to dominant and appealing organizational identities, such as being elite, provide for coherence and stability of identity narratives, a closer analysis of discourses that people draw on in a given situation may reveal contradictions and fragmentation (Alvesson and Robertson, 2006). Therefore, while there is evidence that identity coherence may be linked to increased organizational control, there is also evidence that it is linked to decreased control (Sveningsson and Alevesson, 2003).
A study of the public services sector exploring new public management discourses reiterates this approach. In it identity is found to be a contradictory process of stabilizing and destabilizing discourses 'as individuals negotiate the complexity of "being"' (Thomas and Davies, 2005: 688). The study fi nds that identity work is never as coherent as individuals and/or organizations may wish it to be. Individuals draw on and interpret discourses differently from situation to situation thereby constructing a dynamic identity that reiterates but also subverts dominant subject positions. For example, Kate, a personnel manager for a police service, draws on the discourse of being a mother to resist the discourse of masculine competitiveness in her organization but in so doing also reproduces the gendered conception of the other this discourse relies on (Thomas and Davies, 2005: 693). The authors conclude that resistance is the exploitation of the spaces created by the ambiguities and polyvalence of discourse as the dynamics of identity work unsettle and weaken larger, perhaps more dominant and coherent, discourses (Thomas and Davies, 2005: 701). They also refl ect on their interactions with research subjects describing their interviews as co-constructed social events and their subsequent interpretations as constructions of constructions. It remains unclear how such constructions may have facilitated or subverted larger more dominant discourses and as such how perhaps the researchers' construction of identity may be implicated in the identity dynamics uncovered.
Such refl ection was undertaken somewhat more explicitly in a discursive study of the professional identity of women engineers (Jorgenson, 2002). Particularly, the author aims at uncovering the challenges faced by women in a male-dominated profession but quickly fi nds that the coherence she may have attributed to her respondents is highly contested in practice. The author refl ects extensively on the research process as a communicative encounter in which the researcher discursively positions research participants who in turn position themselves by either conforming to or resisting these positions. The author concludes that identity is constructed in discourse as multiple, negotiated positions and that research participants need to be given more voice 'as authors of their experiences' (Jorgenson, 2002: 365). What the author may be alluding to but does not refl ect on explicitly is that identity coherence as well as fragmentation may be as much a product of how research or the research encounter is produced as of the underlying identity dynamics uncovered in the process.

Organization 16(4) Articles
This aspect was explored further in a discourse study of gay identity in a government department (Ward and Winstanley, 2003). The study examines identity discourse in relation to silence or the absence of negativity. That is rather than being constructed coherently around a discourse of 'the Department [being] a very good place for sexual minorities to work' (Ward andWinstanley, 2003: 1262), the authors fi nd identity being fragmented around various forms of silences and what was not being said. For example, the discourse of revealing one's sexual identity at work may on the one hand lead to greater control for the subject as he or she is no longer silent about his/her sexual identity but it may also lead to a loss of control as the revelation is now part of and perhaps controlled by a more dominant heterosexual discourse.
The authors refl ect extensively on how identity may be co-constructed in discourse between researchers and researched and how respondents may become mere vehicles for the researchers' own identity work turning authentic insight into 'acts of ventriloquism' (Ward andWinstanley, 2003: 1266). They conclude that researchers may harm research participants in this process and examine how particularly in this study they had to take responsibility for naming identity or naming it wrongly. This points to the co-construction of identity discourse and may implicate researchers/authors in processes of control and resistance. That is, engagements with research participants may not only be construed as acts of control or resistance, as in the study of women engineers cited earlier (Jorgenson, 2002), but, importantly, may contribute to making identity discourse available for control or resistance.
This theme seemed to emerge also in a cross-cultural discourse study of management consultants in the UK and Finland (Merilainen et al., 2004). It finds fragmentation in how consultants construct their identities relative to discourses of work/life balance. Consultants in the UK seemed to construct issues around the integration of work and life as a form of resistance to dominant organizational discourses in which work takes precedence, while consultants in Finland constructed the same issues by way of conforming to a larger organizational and societal discourse around ideals for the integration of work and life. As such the theme of fragmentation emerged in the analysis of professional identities of consultants while coherence was found relative to cultural identities with consultants from the same country drawing on similar discourses. The authors refl ect extensively on how research subjects may have been constructed in the research process. Additionally, they consider their own identity work as members of a particular culture may have co-produced the interviews as a lived experience. Here they refl ect at least implicitly on how both coherence and fragmentation are constructed in researcher/researched interactions and how the dominance and normalization of larger discourses such as cultural ideals of work/life balance may be reiterated in and through these interactions.

Michaela Driver
This kind of refl ection is taken up in a study of entrepreneurial identity (Down and Reveley, 2004). It examines how entrepreneurs attempt to construct coherent self-narratives in and through encounters with older managers in which they defi ne their younger, entrepreneurial selves in opposition to the older, less entrepreneurial identities they attribute to the older others. The authors conclude that localized materials and interactions rather than conformity to larger, hegemonic discourses inform identity work. They also refl ect on the subjective nature of identity research and the importance of letting the researched tell their story. They refl ect extensively on how the identity of one of the authors as entrepreneur and friend of the researched as well as the personal nature of the research may have affected the researchers' identity work. For example, they consider how this work has enabled one of the researchers to overcome the anxiety and identity threats of fi eldwork by establishing 'a sense of 'sameness'' (Down and Reveley, 2004: 236) with the research subjects. They do not explore how this sense of sameness may have informed the identity work of the researched and reiterated the larger, hegemonic discourses they seem to have found in their study. For example, they do not consider how their interactions with the respondents may have reiterated generational norms as researchers and researched co-constructed the discourses of 'young guns' and 'old farts' (Down and Reveley, 2004: 233).
The point again is not to deconstruct individual studies but rather to highlight how conceptions of identity are played out in current research. When we review the studies described above and others described in Table 1 what I think emerges is that current research seems to be confi ned to what I have referred to as the set-piece of identity research and the following insights.
First, identity may cohere by conforming to dominant discourses or by constructing an oppositional subject position (Bergstroem and Knights, 2006;Down and Reveley, 2004). Second, identity may fragment in the very act of unsettling dominant discourses (Sveningsson and Alevesson, 2003) and the varying positions that may be constructed by drawing on different aspects of the dominant discourses and doing so differently in different situations (Thomas and Davies, 2005). Third, researchers seem to play a part in all of this as they co-construct the identities of research subjects (Jorgenson, 2002) and may in turn subject research participants to their own dominant identity discourses (Ward and Winstanley, 2003), share participants' identity discourses (Down and Reveley, 2004) or subvert and/or normalize the discourses that can dominate identity work in particular contexts (Alvesson, 1998;Merilainen et al., 2004;Thomas and Davies, 2005). I would now like to examine why this set-piece seems to exist and what underlying dynamics and new avenues for research may be missed by reiterating it. To this end, I turn to the theorizing of French psychoanalyst Jacques Lacan. Individual identities and career discourse in the retail sector (Coupland, 2004) Coherence to the extent that individuals use dominant career discourse to construct identity; fragmentation in how different subject positions are constructed from this discourse Different subject positions alternately normalize and subvert dominant career discourses How the subject positions that emerged in the interviews were co-constructed as the researcher positions subjects in discourse and prompts them to give an account of themselves Professional identity of project managers (Hodgson, 2005) Coherence in how dominant discourse is used for identity work, particularly front stage; fragmentation how this is performed backstage Project managers perform identity work that conforms to but also subverts dominant discourses of professionalism How subject positions may be performed for researchers (backstage as opposed to front stage)

Organization 16(4) Articles
Identity of consultants who sell advise on fl exible work arrangements (Whittle, 2005) Coherence in how consultants reproduce dominant organizational discourses; fragmentation as they also dis-identify with these discourses Dis-identifi cation reduces pressure for identity conformity while allowing individuals to perform in ways that conform How interpretations were co-constructed from the author's understandings of the participants' meanings as well as academic concepts Occupational identity of airline pilots (Ashcraft, 2005)

Subjectivity and the Failure of the Imaginary
Lacanian psychoanalysis (Lacan, 1977a(Lacan, , 1977b(Lacan, , 1988a(Lacan, , 1988b(Lacan, , 2001 with its focus on how identity is constructed in discourse has been shown to have increasing relevance for organizational studies today (Arnaud, 2002(Arnaud, , 2003Arnaud and Vanheule, 2007;Driver, 2005Driver, , 2008Driver, , 2009aDriver, , 2009bJones and Spicer, 2005;Roberts, 2005;Vanheule et al., 2003). It suggests that the conscious language in which a person articulates the self is typically an imaginary construction from which the authentic subject, or the subject of the unconscious, is missing. It is missing because we always construct the self in language. But language is a symbolic order that can never provide us the unmediated experience of the world we have had prior to attaining language and (self) consciousness.
The self we construct in this symbolic order is alienated as an object constructed in the language of an other, internalized others or even the conventions of language itself handed down through generations of others. Therefore, the answers we continuously seek about who we are and what we want never actually answer our questions or, importantly, fulfi l our desires. Lacanian psychoanalysis does not offer a cure for this ill. Rather it offers the insight that the failure of our conscious imaginary self-constructions is a highly valuable experience that we should refl ect on and explore further. In so doing we will not be able to rid ourselves of the imaginary or somehow learn who we truly are so we can fi nally get what we truly want. Instead, we may learn that the failure to do those things is not a personal shortcoming but rather a structural condition. It is an impossibility we can never overcome but live with.
Particularly, we can learn to render transparent the lack that fundamentally marks who we are as subjects (Ragland, 1996). We can refl ect on lack as an articulation of our unconscious that speaks through us even though we cannot understand it. We can let ourselves be reminded again and again that there is something we long for, someone we want to be, but when we try to articulate this, we come up empty. Finally, we can come to notice that we continue to circle around this lack in everything we say and do, but that this lack also provides us with immense and creative potential (Lacan, 1988b). This potential rests in interactions and discourse in which we amplify, render transparent and further refl ect on our struggles with identity and desire and the failure of the imaginary. Because it is here that we come closest to experiencing the self as subject uniquely marked by its lack (Soler, 1996).
To make it more obvious what all this means particularly for identity research, I would like to share an analogy I recently presented at a conference where I asked participants to imagine the perfect dinner date. You are there. He/she is there. The restaurant is perfect. The menu is perfect. You order your appetizers and they are perfect. You anticipate an evening in perfect bliss with the one person you really want to be with. Then you take your fi rst bite of your perfect appetizer, a perfectly tasteful and beautifully presented arugula salad (leafy greens often served with or instead of green Organization 16(4) Articles salad). And you realize that one arugula leaf has become lodged between your teeth. You suspect, but are not sure, that he/she can see it, too. From then on your experience of your perfect dinner date revolves around this arugula leaf. While you are perfectly aware that it is silly to let such a small thing distract you, you are nonetheless distracted. You talk in ways that conceal your arugula leaf and listen only with half an ear while planning for the moment you can escape to the restroom to dislodge it knowing you may not be able to because your dental fl oss is at home.
Imaginary constructions of identity are like this perfect date. We aim for an unmediated and of course unmarred experience of the self and the world we live in but there is always an arugula leaf that mediates and mars this experience. This arugula leaf is language and the otherness and alienation that it creates for us when we articulate ourselves. But the arugula leaf also offers an important opportunity. It indicates an encounter with a failed fantasy or the failure of imaginary self-constructions. If we do not engage with this failure or spend our time covering it up, this opportunity is lost. Unlike actual arugula leafs the failure of the imaginary cannot be removed. However, we can stop thinking about leaf removal and focus instead on how we can enjoy this evening with an arugula leaf between our teeth. We can enjoy how alive we feel when we are confronted with the arugula leaf and how we experience every detail at that moment of struggle. Importantly, we can enjoy this as an opportunity to experience each other in whole new ways. What if he/she also has an arugula leaf between his/her teeth (which he/she is bound to of course)? What if we actually admitted this and thereby discussed our ambivalent feelings, missed expectations, etc., all by way of uncovering new questions?
In short, when we engage with the failure of the imaginary and further refl ect on it, we can not only experience our creative potential but also relate to others in new ways (Vanheule et al., 2003). Psychoanalytic praxis suggests, in order for the latter to happen, it is important to shift our focus in identity discourse from what is said to what is not said (Lacan, 1988a). This means focusing on the many failures that commonly mark our discourse such as omissions, tangents, slips of the tongue, inconsistencies, distractions and other linguistic and rhetorical constructions that point to the failure of the imaginary. If we avoid covering up such failures by completing the incomplete, correcting the erroneous or making 'reasonable' interpretations and inferences, we make space for creative struggles with language and identity by simply allowing the failures to be noticed and explored further.
What does all this mean for research on identity in organizations? First, it does not mean that research is an analytic setting in which psychoanalysis should be practiced by putting research subjects on the couch. Rather, as in other studies drawing on Lacanian thinking (e.g. Driver, 2005Driver, , 2008Driver, , 2009aDriver, , 2009bJones and Spicer, 2005;Roberts, 2005), I suggest that such thinking can provide new insights about how we conceptualize the construction of identity in discourse and importantly how we may theorize The Failure of the Imaginary and its Implications for Research on Identity Michaela Driver and encounter subjectivity with the kind of fl uidity offered by a Lacanian perspective (Elliott and Frosh, 1995).

Identity Research as an Encounter with the Arugula Leaf
Specifi cally, the contribution of Lacanian theorizing of subjectivity to identity research is threefold. First, it provides a way to uncover and render noticeable imaginary constructions of the self. Second, it provides a way to explore their failure. Third, it offers the potential to unsettle the imaginary and explore or even widen the space for creative identity struggles. In so doing it may provide new possibilities for the theory and practice of identity work in organizations and move us beyond the set-piece of coherence and fragmentation. Here is how.
Lacanian theorizing suggests that any construction of identity as a defi nable object, or by way of answering questions about who we are, is an imaginary construction. Particularly, discourse in which we present a coherent and stable self is such a construction. Therefore, whenever we conceptualize or fi nd, in empirical research for example, that someone has a coherent and stable identity, we are articulating an imaginary construction of subjectivity. In this sense the theme of identity coherence in the literature can be viewed as an identity articulation that is fi rmly rooted in the imaginary. But this may or may not mean that the theme of identity fragmentation offers a way out. Specifi cally, it does not offer a way out (of the imaginary) if we conceptualize fragmentation as yet another (imaginary) answer to what identity is. That is, to the extent that identity is conceptualized as fragments that can be put together to form a whole, fragmentation becomes another reiteration of coherence.
For example, coherence may be found when someone has a professional identity completely defi ned by their organization's professional norms. But coherence may also be found when someone articulates their professional identity as being a mix of different subject positions, by for example, drawing on some but not all the norms offered or imposed by the fi rm. So rather than dealing with one coherent, monolithic identity, we fi nd instead pieces of the puzzle. But in the end, we come back to putting them together to see the whole picture. In this sense what has been reiterated in research is that there is very much still coherence but we are showing how this coherence is like a big jigsaw puzzle. So we no longer show a smooth picture but a puzzle in which we can see the seams of the pieces. However, the picture is still there.
As long as this picture is not only there but offered in research as an answer to what identity is, fragmentation is not much more constructive than coherence in moving us beyond imaginary constructions of subjectivity. The alternative offered by Lacanian theorizing is that fragmentation may also be conceived of as the failure of the imaginary. This in turn makes it possible to examine the fragments not as parts of a coherent whole we can uncover but as disruptions articulating an underlying lack that we can Organization 16(4) Articles make present. In this sense, it is not about putting the pieces of the puzzle together but about exploring the seams between the puzzle pieces to show how they do not fi t together.
Building on this perspective then, identity research is very much still about examining identity coherence and fragmentation but as an interplay in which fragments are explored as disruptions and a way to articulate identity as an absence rather than a presence. In particular, it is about uncovering how every subject struggles creatively with lack and how identity work is a unique articulation of the failure of a particular imaginary construction. Consequently, it is not about fi nding out that 'H' really sees herself as a gardener rather than a professional manager in a global company (Sveningsson and Alevesson, 2003). Rather it is about exploring how this imaginary construction of a self-identity (Sveningsson and Alevesson, 2003) fails again and again at the interstice of the many discourses through which subject positions are constructed. In H's case for example, it fails at the interstice of the four discourses dominating her organization, namely globalization, creativity, networking and management control (Sveningsson and Alevesson, 2003). Therefore, the lack that marks H's identity discourse may be triangulated by examining what she articulates as the lack in the four discourses and how, in her particular case, the gardener discourse is the discursive space (Kornberger and Brown, 2007) around which this lack is focused but also around which organizational discourse is (re-)appropriated to make space for the divided subject (Arnaud and Vanheule, 2007) whose lack absorbs any attempts at controlling it.
It has been suggested recently that the creation of identity lack on the part of organizations is a powerful means for inviting employees to bring in, as it were, their own identities and therefore to appropriate their identity discourse for organizational purposes (Maravelias, 2003). If we build on this idea but turn it around, we can say that identity lack on the part of organizational members is a powerful means for appropriating organizational discourse so that the individual may experience a space in which it can be divided (Arnaud and Vanheule, 2007) or lacking. The implication of such identity conception for control and resistance in organizations is that the articulation and experience of lack in identity discourse absorbs, sucks up perhaps, discourses aimed at identity control and thereby makes resistance possible as creative struggles in discourse and new relations to others and the self (Vanheule et al., 2003). This in turn may make more externalized conceptions of subjectivity possible in practice by creating a discursive space in which it becomes possible to undermine an organization's cultural management program by putting not one but one hundred stickers bearing the company logo on one's car (Fleming and Spicer, 2003) as a simple affi rmation that the imaginary always fails and that lack feeds on organizational resources.
To explore such dynamics further, identity studies would need to focus more on fragmentation as an articulation of lack. For example, rather than examining how masculinity provides discursive resources for coherent The Failure of the Imaginary and its Implications for Research on Identity Michaela Driver identity narratives in a fi rm seeking to impose femininity as a dominant discourse (Alvesson, 1998), we could investigate how the feminine organizational discourse is absorbed in struggles with masculinity and the lack it articulates repeatedly. In this sense, we might focus less on how dominant discourses may be unsettled by various local articulations of subjectivity (Thomas and Davies, 2005) but rather how such discourses are sucked into and feed the lack articulated in the 'dynamics of identity work' (Thomas and Davies, 2005: 701).
How this may be done has been illustrated perhaps already in a study I referred to earlier in which gay identity in a government department was explored by focusing on silence and negative space (Ward and Winstanley, 2003). The authors focused extensively on gay identity discourse as it was articulated around what was missing or what was not being said. They found that the dominant discourse of the department being a good workplace because it accommodated homosexual identity was continuously absorbed in particular articulations of gay identities and struggles over the failure of this discourse. Based on the perspective I am advancing here, this failure is not a failure of this particular organization trying to be good by accommodating sexual minorities (Ward and Winstanley, 2003). Rather it is the failure of any imaginary construction of goodness or sexual orientation to address the lack of the subject. This becomes clearer when we focus on the interplay of coherence and fragmentation in particular identity narratives and examine the failures of the imaginary articulated therein.
For example, rather than exploring how entrepreneurial identity is constructed by drawing on the discourse of generational differences (Down and Reveley, 2004), we might focus on how the latter discourse is used to articulate the lack in the former and vice versa. This may bring into focus how entrepreneurial identity and entrepreneurship remain lacking the more we seek to articulate them and, like a sublime object, instantly turn into something we did not want as soon as we do (Jones and Spicer, 2005). In this study, it could have been examined how the lack in entrepreneurial identity discourse was fi lled with generational discourse resulting in rather creative struggles to articulate (the failure of) entrepreneurial identity around 'guns' and 'farts' (Down and Reveley, 2004). Then it could have been examined how entrepreneurial organizations offer discursive spaces for such creative struggles and how many resources are appropriated in and through the lack in and of 'guns' and 'farts'.
This lack also extends to and is refl ected in the interactions between researchers and researched. As the authors of the above study suggest, there is 'sameness' between researchers and study participants (Down and Reveley, 2004: 236). This sameness is not just an imposition by researchers using research subjects to perform acts of 'ventriloquism' (Ward andWinstanley, 2003: 1266) but importantly the result of a shared fantasy that identity exists as a defi nable object and can be articulated in a coherent story. By way of undertaking refl exive research in general (Alvesson and Skoldberg, 2000) and refl exive identity research in particular, it is worth Organization 16(4) Articles exploring how the maintenance of this fantasy is necessitated by, as a reviewer for this manuscript put it, 'academic conventions' that make only coherent stories socially acceptable and therefore publishable. But I would argue that resistance to such conventions, if they constitute indeed a dominant and controlling larger discourse, is best explored as the failure of individual (researchers') identity discourse.
A promising avenue for exploring this exists in building on but also extending the refl ections currently undertaken on how researchers interact with research participants in identity studies. For example, in the refl ections undertaken in the identity study of female engineers reviewed earlier, the author underlines how research participants frequently broke the frame of the interview to re-negotiate how the author was positioning them in discourse (Jorgenson, 2002). Such moments can be explored further as failures of imaginary identity constructions especially those of the researcher seeking to validate particular answers to identity questions or at least the possibility that such answers exist. Here the lack of the researcher's identity discourse as scientist who can cover up lack by being the subject-supposed-to-know (Lacan, 2001) can be seen as appropriating the larger discourse of 'academic conventions'. The coherent story of identity work continues to be disrupted by the failure of the researcher to be the one who knows or can know.
But this failure is also an important opportunity as it shifts the focus from the imaginary competition of identity discourses of researchers and researched clamoring for more voice (Jorgenson, 2002) to uncovering and widening the space in which researchers and research participants can articulate lack and perhaps create something new in the process (Vanheule et al., 2003). What if, for example we conceived of identity research very much as a localized and situated enactment of identity work (Thomas and Davies, 2005) whose main goal is to document the unique and creative ways in which failures of the imaginary are articulated in organizations? What if we reconceived writing good stories (Kornberger and Brown, 2007) as unique articulations of the failure to write coherent ones? What if we listened carefully to our research participants but not to better tell their stories (Sveningsson and Alvesson, 2003), but instead to create new stories with them? What if we did do more in-depth single case studies (Sveningsson and Alvesson, 2003) but not to ensure that we can discern ventriloquism from authenticity (Ward and Winstanley, 2004), but rather to more fully explore unique articulations of lack as localized acts of disruption (Thomas and Davies, 2005) in which we play a part?
And could this be the beginning of identity research as localized and situated identity discourses that facilitate resistance and emancipation in organizations (Thomas and Davies, 2005)? My own imaginary identity construction of a researcher making such conceptions possible prompts me to answer yes. But I am sure that, at this very moment, as you are reading these lines this has become a failed fantasy. So there is nothing left but to ask more questions about identity and identity research. And that is of The Failure of the Imaginary and its Implications for Research on Identity Michaela Driver course the point! There are no answers because all answers fail. But failure is an opportunity to experience creative potential and relate to others in new ways (Vanheule et al., 2003).

Conclusion
It has been argued that identity is more contested and used more for managerial control than ever and that this is why organizational researchers should aim to make identity better understood (Alvesson and Willmott, 2002). The present article makes a contribution to this effort in three ways.
First, it suggests that conducting research with the aim of showing that identity is more contested in organizations today misses an important underlying dynamic of identity work. Specifi cally it misses that identity discourse, whether contested or otherwise, is often stuck in the imaginary and always marked by and articulated around lack. As I have shown, this confi nes it to reiterations of what I have referred to as the set-piece of identity research with somewhat limited implications for resistance and control.
The second contribution this article makes is to underline that managerial control is no more likely to be reduced by contested identity discourse than by uncontested identity discourse as long as imaginary identity constructions remain in place. Therefore, it suggests that more promising avenues for the theory and practice of resistance and control in organizations exist in pursuing contested identity discourses as localized failures of uncontested imaginary discourses and vice versa. In so doing, the lack that marks identity discourse can be rendered noticeable and, importantly, available for refl ection and dialog. This in turn offers new avenues for theorizing and practicing resistance particularly as we focus on how lack absorbs discursive and perhaps material resources (such as stickers with company logos?) and how interactions do not only articulate lack but also widen the space in which the divided subject can exist (Arnaud and Vanheule, 2007).
Related to this last point, the third contribution the article makes is to suggest that our efforts at better understanding identity are implicated in control and resistance but not necessarily in the ways typically refl ected on by researchers. Specifi cally, to the extent that efforts at better understanding identity are translated into fi nding answers to our own identity questions or into, at least, validating the possibility of fi nding them, such efforts keep us fi rmly enmeshed in the imaginary order. Therefore, by way of unsettling this order, it may be worth refl ecting on how we can undertake identity research that helps us, and others, notice and amplify the inevitable failure of this fantasy and connect the lack in the research participants' identity discourse to our own. For a recent illustration of how this may be done, please see Harding (2007). It may turn out that research interactions undertaken with this in mind constitute an important and signifi cant opportunity for resistance by re-appropriating dominant, potentially controlling, organizational discourses to articulate lack and creative struggle.