Corps de l’article

Introduction

Psychological harassment in the workplace is a problem that has been studied extensively for the last fifteen years. Numerous studies have focused on the two main elements that surround the definition of psychological harassment in the workplace (Quine, 1999; Faulx and Delvaux, 2005): the behaviour of the harasser (Leymann, 1990, 1996a; Zapf and Leymann, 1996; Brun and Kedl, 2006; Brodski, 1976; Faulx and Geuzaine, 2000) and the consequences for the victim (Brun and Kedl, 2006; Soares, 2004; Hirigoyen, 2001; Leymann and Gustaffson, 1996; Leymann, 1996a; Tehrani, 2003; Gold, 2003; Matthiesen and Einarsen, 2001, 2004; Geuzaine and Faulx, 2003). Most of these authors have also tried to identify risk factors related to harassment, such as the personality of the victim, group dynamics, the characteristics of the organization, etc.

On the other hand, little research has been dedicated to describing more dynamically how different processes interact with each other to generate psychological harassment. Methodologically speaking, it appears that approaching the problem of psychological harassment in the workplace through case studies would allow a better understanding of the subtleties of these processes (Matthiesen et al., 2003), and would complement quantitative approaches, which are much more numerous than the qualitative analyses of this problem (Hoel, Rayner and Cooper, 1999). This would also allow a better understanding of the nature of the relationship between the harasser and the victim and a better comprehension of the impact of the processes occurring around this relationship. Indeed, it is by paying attention to the stories as they are told that one can best understand the complexity of harassment phenomena (Brun and Kedl, 2006).

Through a detailed clinical case study, this article aims to achieve the following two goals: (1) to identify and analyze the personal, interpersonal, group and organizational processes occurring in situations of presumed psychological harassment in the workplace; (2) to identify and analyze the interactions between these different processes.

Theoretical Background: Risk Factors Identified in the Current Literature

Because our study takes into account four levels of understanding (personal, interpersonal, group and organizational), we will first focus on the suggestions that have been put forward to explain harassment at each of these levels. This will allow us then to correlate these explanations with the data available for each case.

Within the Organization

For many authors (Leymann, 1996a; Zapf, Knors and Kulla, 1996; Vartia, 1996, 2003; Ashforth, 1994; Liefooghe and MacKenzie, 2001; Hoel, Rayner and Cooper, 1999), task descriptions, work organization, management styles and the culture of the organization precondition the appearance of harassment.

Where tasks are concerned, the factors usually identified are quantitative overload and monotony (Leymann, 1996b), and low complexity and weak control over time (Zapf and Leymann, 1996). Where work organization is concerned, some authors mention the compulsory cooperation between employees (Zapf and Leymann, 1996), the lack of information, the lack of opportunity to discuss tasks and objectives, a negative communication atmosphere, little support from colleagues (Zapf, 1999), authority-based methods (Vartia, 1996) and rigid hierarchical structures (Crawford, 1997).

On the other hand, the managerial culture, i.e. the way managers are encouraged to behave (Ashforth, 1994; Koubi, 2004) can stimulate or inhibit harassment. According to Einarsen (2000), it is the lack of “constructive leadership”, i.e. an approach neither “laisser-faire”, nor authoritarian, which opens the door to harassment.

Finally, Weathley (1999: 79) mentions strong links/relationships between the culture of an organization and bullying. In other words, a culture in which the subject of conflict is taboo constitutes a risk factor (Walter, 1995).

Group Level

The main hypothesis concerning the group level is that anxiety, conflict and instability within a group are redirected towards a single individual, who plays the role of the scapegoat (Garcia and Hue, 2002). In other words, violent actions towards an individual appear when the group, after going through a period of frustration, designates one person as “responsible for all of their troubles” (Hirigoyen, 1998). A group in difficulty maintains its balance at the cost of a reinforcement of its norms and cannot tolerate the fact that an individual deviates from them (Leymann, 1996a). If this is the case, these deviations precipitate the rejection of the individual in question (Zapf and Einarsen, 2003). This theory has been most often proposed in cases of harassment from a group of individuals towards a single individual, and is close in perspective to work on bullying at school (Olweus, 2003).

Interpersonal Level

The justifications for the interpersonal level approach are based on the study of relationship modes that exist between the harasser and the harassed individual. Two hypotheses have been put forward at this level.

The first hypothesis postulates the existence of a perverse seduction mechanism from the harasser towards the harassed individual (Hirigoyen, 2001), which would take place in two steps: (1) the harasser exerts seductive forces on the victim and the victim's surroundings; and (2) the harasser puts the victim under his/her hold. It is during this second stage that harassing and repetitive hostile behaviours take place (Hirigoyen, 1998). The relationship process consists of the harasser progressively taking away the identity and the freedom of the harassed individual: it starts with a seduction mechanism, followed by influence and manipulation, which grows stronger and stronger, finally reaching the stage of fascination of the victim for the harasser. Slowly, a real domination process establishes itself, within which the victim is reduced to the status of an object. At that point, the victims are so intellectually or morally dominated that their ability to show opposition has become so non-existent that they become abettors to what oppresses them (Hirigoyen, 1998).

The other hypothesis relates to the degeneration of an interpersonal conflict which, by means of not being correctly managed, turns into the harassment of one protagonist by the other (Leymann, 1996a; Zapf, 1999). Harassment would therefore be the step that follows the degeneration of a conflict.

Personal Level

In general, research studies have demonstrated no convincing association between personality traits and whether or not an individual is harassed at work (Hoel, Rayner and Cooper, 1999). Nevertheless, certain authors have identified risk profiles. According to these studies, harassment victims are, in comparison with control groups, more submissive, conscientious, traditional, dependent, quiet, reserved, anxious, and sensitive (Coyne, Seigne and Randall, 2000). Furthermore, they would rather avoid conflict, and they exhibit difficulty in developing efficient adaptation strategies (Coyne, Seigne and Randall, 2000); they would exhibit symptoms of anxiety or depression, and would exhibit social ineptitude (Zapf, 1999); they are hypersensitive, suspicious, depressed, and have a tendency to express their suffering through psychosomatic symptoms (Matthiesen and Einarsen, 2001), or they suffer from low self-esteem and a high level of anxiety, and exhibit an introverted and submissive nature (Vartia, 1996). On the other hand, victims may also be too direct, frank, or too involved in their work (Balicco, 2001; Zapf, 1999). Yet, it is necessary to remember that these studies cannot determine whether a trait or symptom is a cause or a consequence of harassment (Schuster, 1996; Leymann, 1996a).

Methods

Mode of Participation in this Research

The detailed clinical case study described here is based on the story of one subject, Lucie, who agreed to participate in this research, the objective of which was to gain a better understanding of psychological harassment in the workplace. What will be examined here is thus the view of this case, as the individual, who sees herself as a victim, tells it. The analyses must therefore be considered in light of the subjectivity inherent in any clinical analysis. Her case appeared to us as a particularly good illustration of the ways processes intertwine within situations of presumed psychological harassment in the workplace. This is because the individual's narration brings to light, despite the small size of the organization, processes that appear at different levels, including dynamics of both the “conflictual” and “harassment” types. We have also chosen to present a situation that unfolds over a long period of time, allowing an observation of the effect of time on the evolution of the situation.

Data Collection

The following analysis is based on the study of six interviews, which lasted between one and one and a half hour. The process of data collection took place in two steps: (1) the individual was invited to narrate her story in a general sense; and (2) she was questioned according to interview guidelines, which followed the analysis model (see Figure 1). Technically, the interviews were conducted with the use of specific prompts (directional actions that initiate a conversation on a specific subject) and of reminder boosters (subordinate actions that refer to the previous statement of the interviewed individual) (Gighlione and Blanchet, 1991). The interviews occurred in the presence of a silent observer. The presence of this silent observer allowed for a comparison and a consolidation of the analysis. Between each interview, the interviewer and the observer were able to exchange impressions and clinical analyses, and to go back to the individual's actual words in order to consolidate and support their observations.

Data Analysis

From the transcription of the interviews, we proceeded through a qualitative analysis of a socio-semantic nature, i.e. through a division of the corpus into pertinent significance units, with an intended a posteriori encoding during which the researcher reads the text and marks himself the significance units of the corpus (Jenny, 1997: 67). The significance unit is therefore the core of the analysis and results from the division of the text into layers which have, on their own, a single global meaning (Mucchielli, 1994: 32). Our methodological approach therefore consisted of grouping together any word, sentence or fragment of a sentence having a complete significance by themselves (L'Écuyer, 1990: 62), applying the principles of thematic analyses and categorical organization.

Next, we carried out a thematic analysis in order to perform a classificatory study, which consists of establishing descriptive categories referring to text variables that are exclusive of one another and which fragment and organize the speech (Poirier, Clapier-Valladon and Raybaut, 1996). The descriptive categories are defined by the analysis model presented in Figure 1.

Figure 1

Analysis Model

Analysis Model

-> Voir la liste des figures

Analysis Model: A Process-Based and Interactive Model

The analysis model used to study Lucie's case is derived from the work of Ardoino (1969, 1996). It aims at identifying the processes occurring at each level as well as the interactions taking place between these levels (represented by arrows in Figure 1).

Our concept of process originates from the idea of stochastic process, defined as a random sequence of events that are associated with a selection process such that only specific results will have a significant probability of remaining (Wittezaele and Garcia, 1999: 422). As such, the concept of process is associated with a form of repetition due to the fact that some answers become more probable than others. A process therefore implies redundancy, constraints, and pattern (Bonami, 1993: 20). The objective of our analysis model is to extract those patterns, those redundancies, those processes. In this case, we are studying behavioural patterns, i.e. repetitive sequences of behaviours between the actors.

In the frame of the study of harassment, the uniqueness of this model is two-fold: on the one hand, it is process-based and views situations according to patterns and not factors, and on the other hand, it is interactive because it takes into account several levels and allows the study of interactions between these various levels.

Study of the Case of Lucie

Summary of the Case[1]

This case takes place in the “Jovanovic and sons” firm, a gardening family business. This firm performs two activities: the sale and delivery of flowers and plants and on-site gardening services. The firm was created 32 years ago by Robert Jovanovic, now deceased, and his wife, Marie Jovanovic. At the time, Robert used to drive to his clients in his personal van, while Marie took care of the administrative tasks for the firm. Marie and Robert had two sons: Grégoire and Michel. The brothers took over the management of the family business when their father passed away. They each own 50% of the shares of the business, as stipulated in their father's will, and are the business directors. Grégoire manages the Flowers and Plants Department and Michel manages the Gardening Department.

The business now employs 11 people (Figure 2): Grégoire, Michel, four administrative employees (Joanne, 45 years old, who is Grégoire's wife; Lucie, François and Lucien). Their tasks consist of answering the phone, dealing with the planning for the labourers, managing the bills and the purchases and managing the overall administration. Joanne and Lucie specifically deal with the flowers and plants, while the other two deal with the gardening. There are also five manual workers.

Figure 2

Main Protagonists of the Case

Main Protagonists of the Case

-> Voir la liste des figures

The person we met, Lucie, is 36 years old. She has been working for the Jovanovic's firm since she was 21. The business is prosperous and Lucie is experienced as an accounting secretary, for which she is known for her competency and her efficiency.

Yet, when she came to our consulting session, she was in a situation of great suffering. She had been on sick-leave for three months and could not bear the idea of going back to work. For the last two years, she had been suffering from acute digestive problems, which prevented her from sleeping and from planning any activity during her free time. The origin of these problems, according to her doctor, is clearly associated with her professional situation. Unable to take it any longer, she has decided, with difficulty, that at the end of her sick leave, she will sign on as unemployed and search for another job.

Lucie explains that she had been having a particularly difficult relationship with Joanne, the wife of her boss, Grégoire. Everything had gone well for the first two years. Lucie was happy at work and had developed a very close relationship with Joanne. Later, things began to fall apart between them. Although Lucie has no understanding of how this happened, their relationship worsened to the point of becoming completely unbearable for Lucie. Lucie then went through what, to her, constituted harassment: hurtful and repeated remarks about a pronunciation problem, public humiliation, systematic denial of her as a person (speaking of her as if she was not there, pretending not to see her), etc. Lucie explains that these behaviours happened regularly and repeatedly. She considers that she was a victim of these behaviours over the course of 11 years. The frequency of these behaviours was irregular. Some periods were quieter (during a few months, she was the victim of one or two such behaviours per week); other periods were characterized by a much stronger hostility, including several such behaviours per day.

Three years ago, after 11 years of such treatment, Lucie “exploded”. She told Joanne everything “that was on her mind”. Their relationship worsened further and the described behaviours intensified with no more periods of respite and Lucie experienced continued hostility or denial of herself as an individual every day.

Analysis by Level

Organizational Processes

Culture. Lucie speaks of the “weight of the family culture on professional function” (L3) and the “confusion between the professional level and the family level” (L3). Indeed, the Jovanovic firm was initially a family business which, while expanding, slowly integrated outside elements. The family culture and the business culture have always remained tightly associated and the familial-based functioning of the firm has remained, despite the arrival of “strangers” into the family.

At Jovanovic's, we like to say that we are one big family. And it is true that we are welcomed with open arms. In the evening, we have meetings at the manager's house and Marie makes pies for us. Meetings at work can extend into the evening.

L2

It appears in the analysis of Lucie's tale that the distinction between work and family time and between work and family space was very difficult to make.

Another significant element of the family culture was set as a norm for the organization: a systematic denial of any form of conflict. The parents have always pushed their children not to talk about conflict. This continues to be an important norm in the organization, and we can therefore speak of conflict as a taboo subject in the Jovanovic business. This norm is constantly reinforced for the common good: “Grégoire always says: `we cannot allow ourselves the luxury of being in conflict with each other. We have to stick together. It is the only way that a small business such as ours can survive.'” (L2)

Finally, Lucie also describes the “macho” (L3) culture of the family and the business. Women are used for more secondary tasks in an essentially male-driven context, and gender-based derogatory comments are very common: “There are calendars of naked women in the office and pictures of the employees are put on the heads, for example.” (L3)

Women are also thought of as more “sensitive” or “prickly” (L4), which leads to the tolerance of Joanne's aggressive behaviour as a gender trait: “They said: `Joanne, it's normal, she is a woman. She is sensitive. Don't take offense from it.'” (L3)

Management. One of the factors described in the literature relates to inadequate management of social processes: “For Grégoire, human relations, it's zero, he does not understand. He covers his ears. The work is well done; he doesn't care about anything else.” (L3)

Moreover, the management team, represented here by Grégoire and Michel, allowed Joanne's aggressive behaviours towards Lucie to develop; maybe they even encouraged Joanne's domination over Lucie. For example, when Lucie rebels against Joanne, she is called into Grégoire's office: “After I yelled at her, she went home. My boss called me in to ask me `What did you do?' You see, they make me feel guilty this way. . . . Michel also, of course, he lets things go.” (L2)

On the other hand, Joanne never has to face comments from the management when she exhibits hostile or aggressive behaviour towards Lucie. As such, despite a globally favourable pattern of behaviour towards Lucie, in cases of clear conflict, Grégoire will favour Joanne. As Lucie says: “The boss is caught between a rock and a hard place: I know that he has a lot of respect for me but, on the other hand, he lives with his wife” (L2). And in the end, Grégoire's decisions rule in the organization.

Work organization and task characteristics. The organization is also characterized by a lack of any form of institutionalized organization: no formal policies regarding schedules or regarding the quality or quantity of work, no precise work rules. Everything is regulated informally, through small comments or observations: “No one will ever tell you: you didn't put in your hours. They say: `Huh? You are leaving?' I always had the impression that I wasn't doing what I had to. Every evening, I wondered, can I leave?” (L5)

The absence of work policies leads to poorly defined tasks, without quality standards. The four employees thus do “a little bit of everything”, even though two of them are supposed to work on deliveries and two of them on maintenance. This situation results in a need for constant communication and the feeling of never mastering a work process. Everything is dealt with a kind of urgency and the work appears to be somewhat disjointed: employees switch from one thing to another depending on their colleagues' requests. Quantitative overload, lack of clear instructions, required cooperation and weak time control are all risk factors identified by Leymann (1996a), Zapf (1999) and Vartia (1996).

Group Processes

In Lucie's case, the hypothesis of the scapegoat, so dominant in the literature, does not seem pertinent. First of all, virtually all attacks come from Joanne. Next, if some of the workers are disagreeable towards her, they only represent a proportion of the employees, while others, by contrast, support her. Instead of describing a group dynamic, such as those found in the literature dealing with a scapegoat (united group, strong identity, constraining norms, etc.), Lucie believes that this dynamic revolves around two sub-groups, of which she describes not only the composition but also the values and behaviours. These supporters share common values: they defend quality of work, the setting of limits between private and professional lives, the respect of employees and the adoption of more precise work policies.

Within the framework of her difficult relationship with Joanne, Lucie views three individuals who appear to be “on her side” (L6) in the business: the two workers responsible for the flowers and plants, as well as François, all support and encourage Lucie.

There are some of us who think that it is going too far. They say that the treatment that Joanne is subjecting me to is really unfair, that she is a frustrated and mean person. . . . Sometimes, they even make fun of Joanne.

L6

On the other hand, Michel, Lucien and the other workers are hostile towards Lucie. According to them, she is too weak to hold her place in the business.

Michel, he was always terse and disagreeable. Lucien also, he was telling me, you must be able to take some of it on yourself. The day that I burst out against Joanne, he looked at me in such a way! And when Joanne would attack me, they looked as if they thought she was right.

L5

They also point out Lucie's “uptight” tendencies and evoke her physical problems as the proof of her inability to resist stress. Lucie is therefore the target of mockery and small hurtful devaluing comments from this subgroup. In addition, Lucie describes a conflict within the group.

There are two sides: the `family' and the `professional' sides. The former wants to give everything to the business; the latter believes that rules are needed that will make everyone more efficient and better off overall.

L6

As far as the functioning of the group is concerned, we can see a conflict-based cleavage appear between the two subgroups within the organization. Each subgroup has its own logic, its own way of thinking, its network, its values, and it maintains a different position on the relationship between Lucie and Joanne.

Interpersonal Processes

By analyzing the relationship that has developed between Joanne and Lucie, we can find the ingredients that make up a behavioural process of perverse seduction (Hirigoyen, 1998, 2001). In the first phase of their relationship, i.e. during the first four years, the two women were very close. Lucie admits having been fascinated by Joanne, who she thought was “beautiful” and “really classy” (L1). Joanne fairly quickly took Lucie under her wing. At the time, Lucie was a very young woman of 21, very happy to have been hired for the first time. Unused to luxury and wealth, she used to visit expensive stores in the nearby town with Joanne, share private conversations with her, etc. In other words, Lucie was “seduced” by Joanne. Joanne, on the other hand, seemed elated by Lucie's admiration for her and admits to her that she is somewhat neglected by her husband and has few girlfriends. Yet, following the return of Lucie's sister from a trip, phenomena reflecting the second phase of perverse seduction appear.

When my sister came back from Brazil, after being gone for five years, the relationship with Joanne deteriorated. I was spending a lot of time with my sister and Joanne could not bear it. It is true that at that point, I was less available and I was spending a lot of time with my sister, who was depressed at the time. From then on, Joanne's attitude towards me changed.

L1

Progressively, Lucie is put into a situation of uncertainty: Joanne alternates between friendly and hostile attitudes and puts Lucie in the position of the object. Joanne ignores Lucie, despises her and is aggressive towards her, including when Lucie tries to rebel against this treatment.

She became mean . . ., contemptuous . . ., mocking towards me, but I was letting her. And then, one day, I exploded. I let everything out. I told her that she was frustrated, that she was attacking the weakest and that she needed help. I meant it. She made me pay for that later.

L1

The imbalance between the resources invested in the relationship between the two protagonists is noteworthy. Imbalance is characteristic of harassment type situations (Einarsen and Skogstad, 1996; Keashly and Nowell, 2003; Salin, 2001). Joanne, although not hierarchically superior, develops a domination relationship based on the valorization of her more extensive professional experience, on belonging to the family that owns the business, and on a psychological domination over Lucie, in whom she arouses a mixture of admiration and fear.

Personal Processes

Lucie, who was the victim of abuse by her uncle during her teenage years, believes that that situation has impacted the way she builds relationships now. She has a very strong need to be appreciated, resulting in her entering into exclusive relationships as soon as she feels that someone is interested in her. Later, she finds herself trapped in a relationship from which she does not know how to escape: “When I like someone, it's a passion. Then people sometimes have a tendency to abuse this. I can't say no.” (L4) She also has a tendency to put other people's desires before her own. As such, when someone helps her or supports her, she tends to easily feel indebted to them, even if this comes at a high price: “I always want to please people. I want people to like me.” (L5)

Randall (2001) demonstrated that people who are victims of psychological harassment in the workplace often exhibit a personal history containing parental dysfunction towards them. One case he described is that of a person who learned early to be submissive in order to avoid confrontation with her parents, and who functions similarly at work. In this case, Lucie held out for eleven years before reacting.

On the other hand, Lucie touches on elements that are linked to the way Joanne functions as a person. According to her, Joanne is a woman who has a strong need to be admired and recognized. On this subject, Lucie mentions the case of another employee who left the business six years ago, and with whom she has been in contact. This person was also very close to Joanne until the situation deteriorated for her as well.

Interaction between the Different Levels

After demonstrating that the different levels of understanding contain “risk” processes, we will now attempt to see whether interactions between the different levels, which give the situation its destructive potential, become apparent from the analysis. It is appropriate here to remember our analysis strategy for clinical cases. Indeed, we are dealing with a meta-analysis, built upon a first analysis of the facts by the victim, which is itself inferred from “reality”. Obviously, in order to investigate the interactions between the different levels more deeply, a cross-checking of reports from different individuals having witnessed the same situation would be appropriate. Nonetheless, we deemed it interesting to proceed with our approach to investigate whether the interaction between levels would allow a deeper understanding, even if only partial, of the situation.

Interaction 1: Personal Process – Interpersonal Process

Lucie's personal process interacts with Joanne's personal process in the framework of an interpersonal process based on seduction and domination, “kind” at first and abusive later on. The interpersonal process will be maintained through the personal functioning of the two protagonists who will, for many years, maintain an unsatisfactory relationship. Their specific way of interacting with each other, with Lucie always wanting to satisfy Joanne, and Joanne wanting to be “admired”, will perpetuate the interpersonal process, even after the relationship has become unsatisfying for both of them.

Interaction 2: Personal Process – Group Process

According to Lucie, the group process is characterized by a conflict within the group. It is possible that Lucie, in her suffering, has over-emphasized the group level and the opposition between “family” and “professionalism”. Nevertheless, if Lucie is correct, we can hypothesize that the two subgroups facing each other have a tendency not to express directly the object of their disagreement, especially because of the norm of non-expression of conflict within the organization. Therefore, the relationship between Lucie and Joanne is an opportunity to express the conflict between the two subgroups. Yet, Lucie's personal functioning pushes her to search for approval, to want to be liked by all and to try to satisfy each person. As an outsider to the family, she will therefore represent a particularly well suited target for tensions because she is very sensitive to remarks addressed against her.

Interaction 3: Personal Process – Organizational Process

The organization is characterized by the strong presence of the family culture. When starting her professional career, Lucie possessed little experience of a work environment and was strongly affected by her family history. Her integration into this system was quick, and this favours unconditional affiliation. In these conditions, few boundaries can be drawn in order to avoid abuse, especially because the system is strongly informal, with no precise work instructions or organized schedule, for example. Pushed by a strong desire to integrate, Lucie experienced a lot of difficulty in setting limits in a system which, itself, has few. The fact that the norm in the organization is to not express conflict also interacted with Lucie's difficulty in expressing her own needs and limits, in other words, in delineating her conflict zones.

Interaction 4: Interpersonal Process – Organizational Process

The relationship between Joanne and Lucie is favoured by the organization's mode of functioning, which is based on the absence of regulation concerning relationships and the ideology of the family unit. This is true in both steps of the evolution of their relationship. The organization allows the establishment of a relationship of privilege and seduction but also the development of attacks and abuse by Joanne towards Lucie.

Interaction 5: Group Process – Organizational Process

The inter-group conflict process and the organizational process, both involving a conflict at the head of management but also its denial, interact to create an environment where the conflict is simultaneously central and yet absent from managerial discourse. While the conflict seems to characterize the group system, we can hypothesize that, at the end of the day, the organization has to affirm its unity. In this case, all those who have supported Lucie might feel even more uncomfortable expressing their objections about the aggression of which Lucie is the object.

Conclusions on Lucie’s Case

Lucie's case demonstrates the importance of analyzing on several levels, and surprisingly of going beyond a strictly interpersonal study of situations of psychological harassment, including in situations showing the characteristics of a primarily “psycho-relational' problem. Indeed, Lucie's situation first appears as an essentially interpersonal problem, fed by a problematic personal functioning. This situation presents all the characteristics of an interpersonal relationship of perverse seduction. It is therefore very tempting to limit the study to this level. Yet, Lucie's case demonstrates the importance of extending the analysis to other levels and of studying the way in which these levels can interact with each other.

This case demonstrates the existence of processes relating to both conflict and victimization within harassment situations. Although conflict and harassment tend to be viewed as opposed in the literature, we observe that victimization processes and conflict-related processes can co-exist.

It is easier to understand this coexistence of processes of a different nature when we realize that they co-exist at different levels of analysis and reinforce each other, which certainly opens the way for investigating the delicate question of the links between conflict and harassment. We observe indeed that conflict-related processes, far from being an alternative to non-egalitarian processes or occurring from limiting unbalanced interpersonal relationships, actually contribute to the support of domination and victimization relationships at the interpersonal level. Despite the fact that Joanne and Lucie are both potentially encouraged by their presumed “supporters”, the relationship between the two women will remain asymmetrical. Lucie's “supporters” are helping her to “hold on” more than they are helping her to re-balance the relationship with Joanne.

Generally speaking, the destructive potential of the situation stems more from the interaction between the levels than from the action of any particular risk factor. We can therefore speak of interaction rather than of a simple influence of the context on the harassment relationship because the analysis of the interactions between the levels demonstrates the presence of a synergistic link between the psychological interpersonal harassment relationship and the functioning of the group. By contrast, in the literature, group and organizational factors are generally viewed as elements influencing or conditioning the interpersonal harassment relationship, but not the contrary. Yet, we see here that the interpersonal relationship is not only made possible by organizational or group factors but rather that the group factors and the interpersonal relationship act synergistically such that the organizational and group processes are reinforced by the interpersonal processes as much as the interpersonal processes reinforce the organizational and group processes. For example, while the culture of negating conflict reinforces the harassment relationship, by taking away from Lucie the legitimacy to express her difficulties with Joanne, the harassment relationship will result in a reinforcement of the denial of the culture of conflict within the organization. In this way, Lucie will ultimately be excluded from the system after her “blow up” against Joanne, thereby confirming and reinforcing the interdiction to point out conflict in this organization. Through this exclusion of Lucie, the norm of non-expression of conflict was both expressed and reinforced within the organization.

This case also demonstrates that the organizational, group and interpersonal processes that are part of situations of presumed psychological harassment in the workplace can evolve. Indeed, it appeared in this analysis that the period of time set the reference conditions for the way we understand the situation. For example, Lucie identifies several key moments (for example, the return of Lucie's sister, Lucie's “blow up” towards Joanne), which impact the interactive processes at the different levels. It is particularly striking for the relationship between Lucie and Joanne, which goes from seduction to perverse aggression, to a momentary hostile “explosion” before returning to an interpersonal relationship of domination and submissiveness. This means that, depending on the time period at which Lucie's case is studied, the conclusions that can be drawn about the processes at hand will be very different, while the risk factors, which evolve to a lesser extent, will not have changed.

Finally, this analysis demonstrates the importance of distinguishing, on the one hand, the group and organizational processes that impact directly the interpersonal relationship which is presumed to be one of harassment (for example, the support of the group towards Joanne), and, on the other hand, the processes that define relationships within the group or the organization in general, independently of the harassment relationship (for example, the conflict within the group), and that act upon it via the first category.

Discussion

Through an approach of investigation and systematic information processing, the study of Lucie's case allowed us to shed light on several unrecognized characteristics of psychological harassment. Specifically, this approach revealed the fact that processes that appear at different levels, at different times and according to different relationship types, intertwine, interact and reinforce each other to create a more complex and particularly inextricable situation.

Since this represents a unique situation, the model we are proposing in this report would need to be applied to a larger population in order to be validated. As a clinical case study, this contribution therefore has an exploratory status because of the limitations inherent in this type of approach: unique and possibly not suitable for generalization, subject to the contingencies of face-to-face interviews, and presented by a protagonist who transmits her own subjective view. Studies based on a larger sampling would allow us to observe the effects associated with specific organizations, as for example in this case, the effect of the fact that the situation happened within a small family business. Similarly, studies based on reports from several witnesses would allow the integration of several points of view and a better understanding of all the micro power relationships, which are woven between all the partners, and not only those involving the presumed victim.

This approach brings complementary information useful to the current state of research. Indeed, to date, investigations on risk factors, although useful, are limited, as observed by Di Martino, Hoel and Cooper (2003), specialists of large-scale statistical approaches: “no profile defining the author of harassment/persecutor or the harassed/victim has been uncovered, no unambiguous profile about situations susceptible to generate violence has been found, no high-risk organization or group profile could be established” (p. 23). Yet, thanks to these studies on risk factors, we were able to build an analysis grid in order to go “inside the black box” of harassment, the behaviours and consequences of which are now well known.

The other usual limitation of case-study approaches is the subjectivity of the analyst. Although we were unable to eliminate this aspect, its effect was limited by the fact that the interviewer was assisted by a silent observer.

Perspectives

The analysis of Lucie's case was performed based on a model relating the different relationship processes, organized by level (Figure 1). Thanks to the conclusions drawn from this study, we have built an enriched model, which allows a finer description of the complexity of situations of presumed psychological harassment in the workplace. The purpose of this new model is to reflect the way the different processes intertwine, and specifically, the way that processes involving the organization, the group and the individuals impact the relationship between the victim and the presumed harasser. On the other hand, this model aims at describing the connection between victimization and conflict processes, in order to describe how they all interact with each other.

Our proposal is to switch from a process-based and interactive model, such as the one we have used to study Lucie's case, to a process-based, integrative and dynamic model. By an integrative approach, we mean one that takes into account the way the different analysis levels (personal, interpersonal, group and organizational) influence each other. By a dynamic approach, we mean one that must be able to describe the balance of forces within relationships and their victimizing or conflictual nature. Finally, the process-based approach is the same as in the first model.

The new analysis model we are proposing includes two axes. The first axis defines four levels of analysis (personal, interpersonal, within a group and within an organization) and is based on the psycho-sociological approach of the levels of intelligibility of reality (Ardoino, 1996). The model allows the differentiation between interactions involving the perpetrator and the victim of the harassment (interpersonal level), interactions within the work group (group level) and interactions within the organization (organizational level). The integrated representation of these levels allows us to enrich the reading of the effects of simple interactions, unit by unit, in the same way as we proceeded with the first model.

Figure 3

Proposed Process-Based, Integrative and Dynamic Analysis Model

Proposed Process-Based, Integrative and Dynamic Analysis Model

Figure legend

1

Personal processes of the harassed individual

2

Personal processes of the harasser

3

Interpersonal processes

4

Group processes towards the harassed individual

5

Group processes towards the relationship of psychological harassment

6

Group processes towards the harasser

7

Relationship processes within the group of co-workers

8

Organization processes towards the harassed individual

9

Organization processes towards the relationship of psychological harassment

10

Organization processes towards the harasser

11

Relationship processes within the organization

-> Voir la liste des figures

The advantage of the new model is that it illustrates how processes, which appear at different levels, interact with each other. In this representation, we see that each level contains its own zone, with processes defined independently of the relationship of psychological harassment or of its protagonists. As such, one zone includes the relationship processes taking place within the organization (11), one zone includes the relationship processes taking place within the group (7) and two zones include the relationship processes that are personal to each of the protagonists (1 and 2).

Interaction zones also appear: Zone 3 represents interpersonal processes; and Zones 4, 5, 6 and 8, 9, 10 represent the group and organizational processes involving the protagonists of the harassment relationship (for example, the organization encouraging harassment behaviours, or the groups supporting one of the protagonists).

Zones 4, 5 and 6 together therefore represent the group process towards the psychological harassment relationship (which includes behaviours towards the harassed individual, the harasser and the harassment relationship). Zones 8, 9 and 10 together represent the organizational process towards the psychological harassment relationship (which includes behaviours towards the harassed individual, the harasser and the harassment relationship).

The second axis allows the distinction between conflict-based processes and victimization processes, originating from the distinction between complementary schismogenesis and symmetrical schismogenesis (Bateson, 1972, 1977). According to the author, there are two types of relationship: complementary schismogenesis, as exemplified by the relationship between a “tormentor” and his “victim” and symmetrical schismogenesis, exemplified by the conflict situation. Characterizing each zone according to this dimension will demonstrate how conflict-type (symmetry) and harassment-type (complementarity) relationships connect in the situations under study. Furthermore, determining whether a situation can be characterized by symmetry and rigid complementarity or not (Watzlawick, Beavin and Jackson, 1972), or as positive or negative (Orgogozo, 1998) would allow a deeper understanding of this distinction. We would then be able to assess to what level a relationship mode contributes to a person's well-being or whether it should rather be considered as pathological (Wittezaele and Garcia, 1999).

As a result, an integrative, process-based and dynamic model is therefore now available for future analyses, which will allow the description of how different levels intertwine and how conflict and victimization connect in situations of presumed psychological harassment.