What a systemic reading brings

The phenomenon is not limited to the stereotyped figure of the “narcissistic pervert”: it can occur in many everyday contexts, at any age (Hirigoyen, 2005). Once installed, coercive control has potentially serious consequences for the mental health of the person subjected to it (Hirigoyen, 1998).

The systemic approach, made known by the work of Gregory Bateson and his colleagues in Palo Alto in the 1950s and 1960s, holds that individual behaviours take on meaning inside circles of interaction. With the theory of the double bind, Bateson et al. (1956) showed how paradoxical communication patterns could trap an individual in an inextricable situation: two contradictory messages issued at two levels — a verbal content and an opposing non-verbal message — leave the person unable to respond correctly without “doing wrong”.

This mechanism transposes directly to situations of coercive control: the manipulator alternates compliments and criticism, affection and threat, creating a climate of confusion that paralyses the other. Such paralysing contradictory messages have been described as central to coercive control (Hirigoyen, 2005), because they blur the critical judgement of the person subjected to them.

A frozen complementary relationship

Paul Watzlawick and his collaborators formulated several axioms of human communication that shed light on power dynamics (Watzlawick, Beavin & Jackson, 1967).

In a symmetrical relationship, each partner tends to act as an equal, sometimes in competitive escalation. In a complementary relationship, by contrast, the roles of dominator and dominated take hold: one adopts a “one-up” position, the other a “one-down” position. This pattern is not pathological in itself — it is found in many social interactions. But when it rigidifies to the extreme, with the one-up position always occupied by the same person, with no possibility of reversal or negotiation, it opens the way to abuse of power. Coercive control corresponds precisely to this frozen, deregulated complementary relationship.

Watzlawick et al. (1967) also noted that participants in an interaction punctuate the sequence of events differently: each has their own interpretation of who causes what. In a relationship of coercive control, the manipulator and the person subjected to it may each see themselves as reacting to the other’s behaviour — “I have to control her because she is too unstable” against “I give in because he frightens me” — which feeds a self-justifying vicious circle.

Circularity and homeostasis: why it is so hard to undo

Another useful systemic concept is that of feedback and circular causality. Unlike a linear view in which A causes B, systemic thinking sees the two protagonists influencing each other in a loop. Even when the balance of power is unequal, the dominated person takes part in the cycle in spite of themselves by adopting behaviours that confirm the domination — trying to appease the aggressor, for instance, may reinforce his sense of power.

The maintenance of the relational system’s homeostasis also explains the difficulty of changing the situation: every system tends to resist change in order to preserve its balance (Jackson, 1957). In a family where coercive control reigns, any attempt to assert oneself or to leave can trigger redoubled reactions of control or guilt-inducing designed to restore the status quo.

The mechanisms of coercive control

How, concretely, does one person come to hold such sway over another? The mechanisms combine techniques of mental manipulation with strategies of isolation and progressive domination. Clinical literature and accounts generally identify several stages in this process (Hirigoyen, 2005; Payet, 2013).

1 — Seduction and initial buy-in

At the start of the relationship, the future abuser often appears charming, attentive, “too good to be true”. This love bombing consists in showering the target with compliments, gifts and intense affection so as to create rapid attachment (Sweet, 2019). The manipulator knows exactly what the other needs or wants to hear, displays a false empathy and presents himself as ideal. This phase builds a nascent affective dependency: fulfilled and under an illusion, the person commits emotionally and lowers their guard.

2 — Confusion and guilt

Once trust is won, the manipulator gradually introduces controlling behaviours. It is often subtle at first: criticism disguised as advice, hurtful jokes followed by “I’m only joking, you’re too sensitive”, unpredictable mood swings. The disorienting alternation of tender moments and psychological violence creates cognitive blurring (Hirigoyen, 2017). Contradictory messages — “I love you”, then “you are pathetic” — lead the person to doubt their own perceptions.

This mechanism is often called gaslighting (Sweet, 2019): the aggressor imposes his distorted version of reality, denies the facts (“I never said that, you’re making it up”), minimises his wrongs and makes the other out to be mad or at fault. Little by little, the person subjected to control internalises the dominator’s discourse and loses confidence in their own judgement. They feel permanently in the wrong: he makes them feel guilty for supposed failings (“it’s all your fault if I lose my temper”) and shifts onto them the responsibility for his own acts.

This role-reversal tactic was theorised by Freyd (1997) under the acronym DARVO (Deny, Attack, Reverse Victim and Offender): when challenged, the aggressor denies the facts, attacks the other and casts himself as the victim, turning the situation to his advantage.

3 — Isolation and control

To consolidate his hold, the manipulator seeks to isolate the other progressively from any outside influence. He may sow discord with relatives, criticise friends or family, create conflicts, or set himself up as the only trustworthy person. Often the person reduces contact of their own accord, either to avoid displeasing a jealous abuser or out of shame about the situation. Deprived of others’ gaze, they lose the social supports that could serve as bearings.

In parallel, control extends to every aspect of daily life: monitoring of movements, control of money, decisions about going out, spying on the phone. Forms of deprivation appear: withdrawal of affection through punitive silence, sleep deprivation through interminable night-time arguments, direct threats (“if you leave me, I’ll destroy you”, “you’ll lose the children”). Fear settles in, and psychological violence sometimes gives way to more explicit violence. At this stage the person is cut off from their free will, “unable to live without the other or to make the slightest personal decision” (Psyliège, 2021). Their identity erodes, their reality is colonised by the other.

Power, manipulation and toxic relationships: contemporary debates

For a long time, classical systemic approaches — focused on circularity and neutrality — were criticised for their relative silence on the moral and asymmetrical dimension of these situations (Goldner, 1998). Treating a violent relationship as a system in which everyone “contributes” can indeed give the impression of equally blaming the person who suffers, or of minimising the abuser’s intent. Voices, notably feminist ones, were raised from the 1980s and 1990s onwards to recall that in domestic violence there is a fundamental inequality of power and responsibility.

Systemic practitioners have partly integrated these criticisms: one can analyse the interaction circularly while clearly naming the abuse and demanding the immediate cessation of violence. As Goldner (1998) writes, recognising psychological interdependence does not mean conferring equal responsibility on the perpetrator and on the person subjected to violence: safety and equity must remain the priority. Current approaches combine a relational reading with an ethical position — the therapist steps out of benevolent neutrality to adopt zero tolerance for behaviours of domination.

The label and the clinic

At the same time, public discourse has seized on the theme of “toxic relationships” and “narcissistic perverts”. The concept of narcissistic perversion was introduced into psychopathology by Racamier (1986) to describe manipulative personalities who use and devalue others in order to sustain a failing self-esteem. In the 2000s, the expression met with dazzling media success. While this popularisation had the merit of bringing psychological violence into the light, it also led to excesses.

Clinicians have denounced an over-extensive use of the term, to the point where “narcissistic pervert” has come, in everyday language, to designate any unpleasant partner or tyrannical superior (Seguin, 2014). Yet the notion remains contested: it does not appear as such in the official diagnostic classifications (DSM-5, ICD-11), and many psychiatrists and psychologists remain reserved about its validity. Speaking of it as a separate category tends to mask the fact that these individuals often present already identified disorders.

Seguin (2014) points to an ethical risk: a therapist who hastily diagnoses an absent third party certainly gives their patient a simple explanation, but at the price of a fixed and hopeless view that can limit the person’s own work on themselves or obscure other issues.

This obviously does not mean that manipulation does not exist, nor that everyone subjected to coercive control is “co-responsible”. There undeniably are individuals who are highly toxic and destructive to those around them (Racamier, 1986). The debate concerns the need to distinguish the clinical reality of coercive control — a complex, progressive phenomenon involving a particular relational interlocking — from its sometimes caricatural collective imagery.

The reflection now extends to institutional power and to collective forms of coercive control. The notion of coercive control in Stark’s sense (2007) has enriched the view of domestic violence by insisting on the global hold — psychological, economic, social — exercised day to day. Parallels are drawn with sectarian indoctrination, workplace bullying and abuses of power in religious and educational institutions. In all these cases we find a gradient of power and techniques of manipulation that create dependency and submission.

Practical applications in therapy and institutions

Family therapy

The systemic approach has given rise to intervention models suited to situations of coercive control within families. The behaviour of one member — the tyrannical adolescent, the ultra-authoritarian parent — can only be understood by examining the implicit relational rules of the system. The therapist decodes the messages exchanged, brings the vicious circles to light and reintroduces flexibility into roles. Faced with a parent exercising control over a child, the work will involve the whole family in order to restore healthy boundaries and an appropriate balance of power.

The Palo Alto school demonstrated the effectiveness of certain paradoxical interventions for undoing coercive control: pushing the logic of the behaviour to its extreme can provoke an awareness. Likewise, reframing the situation from a new angle allows what is being lived to be re-read — pointing out that “loving is not controlling”, or that “giving in for the sake of peace” is understandable but involuntarily sustains the problem.

Couple therapy: an indication to be weighed

Treating relationships of coercive control in couple work is delicate. It is generally accepted that where physical violence is present and active, classical couple therapy is not indicated as long as safety is not assured: a temporary physical separation, individual follow-up of the perpetrator and support for the person subjected to the violence will be preferred.

For couples caught in a dynamic of psychological domination without imminent danger, certain integrative approaches have shown results. Goldner (1998) proposes a treatment in which both partners are heard together, but with a therapist who is highly directive on questions of respect and equality. The aim is to allow the perpetrator to become fully aware of the impact of his acts, by hearing the other’s experience in a secure frame, while exploring how he came to behave this way — often a history of narcissistic wounds or a violent family model, none of which excuses anything.

The simultaneous presence of the person concerned, supported by a therapist who bears witness to their suffering, creates a moral imperative (Goldner, 1998) and blocks the usual evasions. This work demands great tact, and not all professionals agree on its suitability: some hold that joint therapy is only possible once the perpetrator has begun a process of personal questioning, failing which the therapeutic frame itself risks becoming a site of control or retaliation.

Institutions and network practice

The mechanisms of coercive control are not confined to the family behind closed doors: they are found in groups and organisations wherever a hierarchy or a strong ideology is at play. A cult illustrates the process in exacerbated form: the charismatic leader uses seduction, imposes contradictory dogmas and collective surveillance, and isolates followers from the outside world (Lalich & Tobias, 2006). The techniques used there are very similar to those observed between two people — simply deployed on a larger scale. A toxic manager can likewise install a culture of fear and dependency; a carer can abuse their authority over vulnerable patients.

In organisations, one now speaks of systemic dysfunction when a manipulative climate persists: beyond the harassing individual, the rules of the system must be analysed — an ultra-competitive culture that turns a blind eye as long as results follow, a blurred organisation chart that lets an informal leader rule through fear. Interventions then bear on structural changes: clarified reporting procedures, training for managers, monitoring of psychosocial risks. In care or residential institutions, the emphasis is on relational ethics: avoiding excessive paternalism, promoting the autonomy of service users, setting up cross-checks rather than a single all-powerful person in charge of a case.

In institutional contexts — child protection, responses to domestic violence — the systemic approach has led to network practice: the person and the perpetrator are not treated in isolation, the extended entourage, social workers and, where relevant, the justice system are involved in a coordinated way. The idea is to reweave a system of support to counterbalance the coercive control, offering other anchor points than the dominant party’s voice. In parallel, the perpetrator can be engaged in a process of accountability that helps him step out of omnipotence.

Finally, the notion has entered the legal field: in some countries, coercive control within a couple is now recognised as an offence (Stark, 2007). The most effective prevention nevertheless remains education: learning early to recognise the signs of manipulation, encouraging equality and respect in relationships, valuing critical thinking. Increasing efforts are being made to train doctors, teachers and social workers to detect coercive control behind vague symptoms — depression, social isolation, post-traumatic stress — and to direct people towards support services.

What this reading implies

Seen through systemic thinking, the mechanisms of coercive control appear as the product of a pathological encounter between a need to dominate and a relational context that permits it. Neither simple individual manipulation nor simple weakness: an interactive process that locks two actors into a destructive script, and that continues as long as it is not interrupted from the outside. Understanding this dimension leads to four concrete orientations.

• Supporting the people concerned by breaking their isolation, deconstructing with them the narrative imposed by the manipulator, restoring their autonomy step by step.
• Intervening on the relational dynamic rather than demonising the perpetrator alone: where it is possible and ethical, working on the bond makes it possible to change behaviour durably, with a strong requirement of non-violence.
• Training people to spot the signs of pathological power in any human collective — so that neither the family, nor the workplace, nor the school becomes the silent breeding ground of coercive control.
• Qualifying popular concepts: naming suffering is useful, but the label is enough neither to explain nor to heal.

Systemic psychology invites us to look beyond the fixed figures of torturer and victim: it shows a dysfunctional system to be realigned. Unpicking the threads of coercive control takes time, patience and often professional support, but it is possible. Many people who have come out of toxic relationships testify that by understanding its workings they were able to rebuild their psychic freedom. That understanding carries hope: what was built by the relationship can also be deconstructed — through new interactions, marked this time by respect and empathy.

References

Bateson, G., Jackson, D. D., Haley, J., & Weakland, J. (1956). Toward a theory of schizophrenia. Behavioral Science, 1(4), 251-264.

Freyd, J. J. (1997). Violations of power, adaptive blindness, and betrayal trauma. Feminism & Psychology, 7(1), 22-32.

Goldner, V. (1998). The treatment of violence and victimization in intimate relationships. Family Process, 37(3), 263-280.

Hirigoyen, M.-F. (1998). Le harcèlement moral: la violence perverse au quotidien. Syros.

Hirigoyen, M.-F. (2005). Femmes sous emprise: les ressorts de la violence dans le couple. Oh! Éditions.

Hirigoyen, M.-F. (2017). Emprise, soumission et violence dans le couple. Femmes Diplômées, (262-263), 65-76.

Payet, G. (2013). Emprise psychologique. In M. Kédia & A. Sabouraud-Séguin (Eds.), L’aide-mémoire de psychotraumatologie en 49 notions (pp. 114-120). Dunod.

Racamier, P.-C. (1986). Le génie des origines. Payot.

Seguin, É. (2014, March 4). Les “pervers narcissiques” ou le triomphe d’un concept flou. Le Monde (opinion).

Stark, E. (2007). Coercive control: how men entrap women in personal life. Oxford University Press.

Sweet, P. L. (2019). The sociology of gaslighting. American Sociological Review, 84(5), 851-875.

Watzlawick, P., Beavin, J. H., & Jackson, D. D. (1967). Pragmatics of human communication. W. W. Norton.

Watzlawick, P., Weakland, J., & Fisch, R. (1974). Change: principles of problem formation and problem resolution. W. W. Norton.

This systemic reading of coercive control is taken up and illustrated in the video below — in French.