Navigation – Plan du site

AccueilTous les volumesVol. IMediation and Domestic Violence

Mediation and Domestic Violence

Jacques Faget
Cet article est une traduction de :
Médiation et violences conjugales [fr]


This paper analyzes the potential and limits of recourse to mediation for regulating domestic violence. On the basis of an empirical study of its implementation in France and of the existing academic literature, it shows the existence of two types of practices reflecting two conceptions of mediation and more generally, two conceptions of the articulation between social and penal regulation

Haut de page

Texte intégral

1The question of domestic violence has come to occupy considerable space in public debate in the Western world. The phenomenon itself, which probably has always existed, was first brought to the public eye by the feminist movements of the 1970s, then by the struggles for human rights from the 1980s on. The development of public policies facilitating the denunciation of domestic violence (more informational and counseling structures, more humane reception by the police and so on) and the action of victims aid groups have greatly helped to lift the taboo. The taboo does seem to persist, however, if we are to believe a victimization survey showing that relatively few complaints are lodged with the justice system, since these only cover 22% of cases mentioned to investigators (Pottier, Robert, Zauberman, 2002).1 The problem is even more spectacularly masked in countries where women are in a subservient or servile position for cultural, economic or religious reasons. Apparently, there are statistically more denunciations, by women, of violence in the Northern countries where women have a much better social status and are more often financially independent.

2The obstacles to divulging this type of violence in Western countries may be explained by the influence of a political philosophy consistently extolling the separation between public space and private space. In this conception the State should not interfere with people’s privacy, so as to respect personal freedom. The intrusion of government agencies in the family unit, what the French call “la police des familles”, or “policing family life” (Donzelot, 1977) only came with the Welfare State, and only to a limited extent. Violence between spouses remained outside the perimeter of the law for a long time, and it was only the protection of the child’s interests that motivated increased government control of the private sphere (Faget, 1996).

3The present public debate on domestic violence raises the issue of how it should be handled. Since community-based social control seems unable to prevent these offences, it is up to the institutions, and to the criminal justice system in particular, to take over. Now, given the mediocre achievements of the criminal justice system (I), people are increasingly turning to new types of regulation, and particularly to mediation (II). However, this option elicits many controversies (III).

I. The limits of penal responses

4The feminist movements view domestic violence as a crime akin to violence between strangers, and therefore demand stringent punishment and praise the symbolic and dissuasive effects of penal action. This demand for penalization, along with demands for the criminalization of rape, seems paradoxical when the same movement, at the same time, accuses the criminal justice institution of participating in the reproduction of social and gender inequality. At the end of a decade of observation, the strategy of penalization has shown its limits. Despite the increased number of complaints, it is said that too few cases are prosecuted and the sentences pronounced are too lenient (Gauthier, Laberge, 2000). More generally speaking, two series of factors hinder any efficient action by the criminal justice system. The first, the bureaucracy with which the system is laden (A), does not suffice to account for the conceptual inadequacy of penal practices with respect to the objectives pursued (B).

A. Structural limits

5Overloaded courts are a major problem in every Western country. There is a spectacular increase in social demand for law, for all sorts of reasons. One effect of this is the paralysis of the justice system, which lacks the human and material means to cope with this influx. This is true generally speaking, but the phenomenon is particularly evident in family affairs, given the breakdown of the family model and the significant rise in the number of divorces and separations since the 1980s. This overload, added to the habit of not interfering in this domain, produces a high rate of dismissals in France, since it is up to the public prosecutor to decide whether or not to prosecute.

6Despite this consequential elimination, further reinforced by the development of what are known as “troisième voie” (third-track) regulatory mechanisms2 left in the hands of the public prosecutors’ offices, judicial work is criticized on three counts:

  • It is said to be too slow, owing to the number of case files accumulating in the court clerks’ offices, which are insufficiently equipped and under-staffed in secretaries, clerks and judges. In an extremely mobile, increasingly unstable society, this snail-paced conception of judicial time seems anachronistic;

  • It is allegedly very expensive, despite the considerable extension, in several countries, of the field of application of policies giving the most underprivileged access to law and justice by making legal services totally or partially free for them. In family suits in particular, the professionalization of conflict resolution (when a lawyer is required) and the cumbersome procedures (social investigations, expert psychological evaluation and so on) account for their cost;

  • With the exception of trials for serious crimes, in which procedure is painstaking and the ritual takes its time, and of complex cases involving an investigational procedure, judicial work is said to be bureaucratic and heedless of individual differences. The claim is that the development of fast-track treatment procedures encourages recourse to mechanical sentencing logics, thus limiting the ability to collect information on individuals and on the context in which the offence was committed.

B. Conceptual limits

7What is challenged, behind this structural screen, is the whole conception of the functioning of judicial work. The claim is  that the institutional achievements of the justice system in family conflicts are mediocre, be it with respect to reparation for victims, prevention of recidivism or conflict appeasement.

8Historically speaking, the criminal justice system was set up to punish the guilty. Victims were completely overlooked in criminal trials (Cario, 2000), inasmuch as crime was considered above all as an assault on the values of the community. Although their rights receive more protection than in the past, they still are often neglected, and reparation for the material losses or moral wrongs suffered is viewed as secondary.  Hearings are neither places in which individuals can express themselves, their subjectivity or personal suffering, nor places for debate. Judicial time is strictly limited and professionals speak in lieu of the victims. The outcome is that victims have the feeling their conflict has been confiscated (Christie, 1977) and that the penal actors make decisions that go against their own interests and what they want.

9Prevention of recidivism among people who commit domestic violence is problematic on two counts. Firstly, theories of stigmatization have shown that a conviction may have negative effects on a person’s career (Becker, 1963). Rather than transforming the person’s behavior it may reinforce the very features it condemns. Secondly, the conviction itself does not seem to be dissuasive. Most of the time it only feeds and worsens the conflict. “It throws oil on the fire”, in the words of many mediators, and only treats the symptom—the offence—without making any attempt to get to the roots of the conflict.

10The criminal justice system is under fire doubly, as esoteric and inefficient. First of all it constructs situations abstractly, analyzing them using a legal and procedural grid that is incomprehensible for the parties involved, neither of which can identify with its portrayal of their situation. Secondly, in a large number of cases the conflict continues after the trial. “When relations between the spouses are conflicting at the start, they tend to deteriorate during the proceedings and turn out to be even more difficult after the trial.” This is true independently of how serious the conflict is. Actually, it seems that “the more serious the conflict the less the judicial solution seems to be adequate to the situation of the two parties.” (Noreau, 2001)

11Lastly, many criminologists and criminal justice practitioners feel that the use of repressive responses in regulating conflicts between people who are in permanent contact (affective or family relations, neighbors or on-the-job relations and so on) probably serve to temporarily satisfy social demand but cannot constitute a rational response to the complexity of the conflicts involved. This is particularly so since the intervention of the penal institution, which reduces the problem to a matter of individual responsibility, is a far cry from meeting the objective of denouncing the structural nature of violence against women (Gauthier, Laberge, 2000). In fact, a Canadian study shows that judicial actors, although rarely inclined to let juridically qualified offences out of their hands and generally convinced of the efficiency of criminal justice action in combating crime, are extremely split on this question, since 50% of public prosecutors and 54% of counsels for the defense view judicial intervention as ineffective in domestic violence cases (Noreau, 2000).

II. The search for new kinds of regulation

12The reason why a trial rarely solves the conflict is that its function lies elsewhere. It consists of pointing up the social prohibition, reasserting the rule of law by condemning relations based on sheer force and acting pedagogically by asserting the rights of the weak. This symbolic role leaves the question of conflict appeasement unresolved, however.

13The material and moral crisis in the judicial system forces it to look for new modes of action. Thermodynamics teaches us that every system goes through periods of structuring and de-structuring, of balance and imbalance, and is constantly in search of the best option for maintaining itself. Systems in a state of entropy have to open up, but the resulting exchange of energy produces areas of uncertainty since the expected revitalizing outside negentropy is also, simultaneously, a disorganizing factor. So we observe the emergence of a new reformative philosophy known as restorative justice (a), with mediation (b)—the most often-practiced expression, along with  sentence-setting circles and community conferences—contributing the degree of negentropy needed by the criminal justice system to ensure its ongoing legitimacy (Faget, 1993).

A. Restorative justice

14All in all, there are four models for dealing with penal conflict; these are not mutually exclusive, and may be used in more or less complementary combinations (Faget, 2002). Punishment- and restitution-based models are often intertwined, so that restitution is one dimension of every restorative process. Their objectives differ somewhat, however. Repression of domestic violence tends to focus on retribution if the aim is to punish the offender, on therapy or rehabilitation if it is to sooth the traumatized victim or to treat the offender’s psychological or psychiatric problems or social maladjustment, on restitution if emphasis is placed on reparation for the victim and on restoration if the aim is to treat the interpersonal context surrounding the conflict.

























Models for dealing with penal conflicts

39The principles behind restorative justice (or reparative justice according to the terminology popular in Quebec), introduced by Howard Zehr (Zehr, 1990) and John Braithwaite (Braithwaite, 1989) have been repeatedly reworked, and their unity is undermined by political/cultural and institutional disagreements between English and French-speaking academics (Jaccoud, 2003) as well as by widely diverging conceptions. The restorative model is torn between two positions. One tendency, which may be diversionist or maximalist and more or less institutional (Walgrave, 2002), centers its concerns on victims, and calls for a modern version of a logic of restitution (redress, compensatory benefits for victims or the community and so on). A second fraction is more inclined to reconstruct social ties (through mediation, community conferences, etc). The present analysis belongs to this current, sometimes also called comprehensive justice (Bonafé-Schmitt, 2003). It does not view restorative justice as a predominantly penal dynamic for making punishment meaningful again and bringing victims back to a central position in judicial concerns but rather, in a broader political perspective illustrated by mediation, as “the emblem of a new model of negotiated justice” inducing “a transformation of our relation to norms and authority” (Cartuyvels, 2003).

40This restorative model, which may conceivably be implemented up- or downstream of the justice system aims (when applied in the penal framework) at promoting new judicial responses contrasting with earlier logics in that it is “a play on meaning and not on the rules” (Milburn, 2002). Its tenets may be briefly summarized as follows:

  • we must change the image of crime: it must be viewed essentially as an offence against individuals, not against the State;

  • we must change the goals of judicial action: rather than focusing on the offender’s guilt, it should be concerned with people’s needs;

  • we must change the type of legitimacy of the penal response: rather than being a monopoly in the hands of the government and legal professionals, it should involve citizens and the community in the construction of the penal response;

  • we must change our conception of punishment: instead of viewing it as punishing past action it must be seen as preparing the future. The wrong done to the victim should not be counterbalanced by making the offender suffer but by having him accomplish a positive, reconstructive compensatory act. Judicial rites of exclusion should be replaced by rites of inclusion based on respect for the individual and commitment of the offender’s community;

  • we must give back to the actors the control of their own conflict, the suffering of victims must be acknowledged, the offender made responsible, in a process involving cooperation and communication between the parties to the conflict, who meet and talk.

B. Victim-offender mediation practices

41Victim-offender mediation (generally known as médiation pénale in French-speaking countries) developed in North America in the 1970s, then in Great Britain, France and Scandinavia in the 80s, and later in the rest of Europe. Its goal is the same everywhere: better consideration of victims’ problems, making offenders feel more responsible, giving the community a greater role in conflict regulation (Faget, 1997). In France it is only applied to adults and in the pre-sentencing phase, whereas in other Western countries it is mostly used for juveniles, and many provisions make it applicable throughout the penal process, including after sentencing.

42In France this practice, made official by a bill passed in 1993, has been rather successful since it is increasingly used, reaching 40.000 cases in 2003 (Annuaire statistique de la Justice). A national evaluation in 1998 provides information on practices (Faget, 1999). Although some habits have changed since then, the issue of the cultural tension between the two conceptions of mediation is more topical than ever.

1. Unveiling mediation practices

43Recourse to mediation; as mentioned above, it is the Public Prosecutor who receives complaints and has the option, among others, of suggesting mediation to the parties to the conflict. Two figures give a good idea of how mediation is actually used. For one thing, mediation is mostly used when the parties are acquainted (72% of cases). Secondly, mediation is mainly recommended for violent offences (physical and moral abuse). Not all of these violent acts are committed in a family context, however.

44According to our study, domestic violence represents about 40% of the cases handled by mediation (33% for physical abuse and 7% for moral abuse). But there are no figures for the relative proportion of complaints for domestic violence that are dropped, oriented toward mediation or sent to court. Interviews with judges show that they presently view mediation as the best response to this problem.3 Every social class and type of family is represented in cases of violence sent for mediation. However, we find a growing number of conflicts involving recomposed and multicultural families, and of immigrant-origin families in which the husband does not accept his wife’s desire for emancipation.

45The mediation process: mediators usually see the individuals separately. Then, if both agree, they organize one or several encounters during which both parties may give their viewpoint and discuss possible solutions. Some skip the first phase and organize encounters immediately. Others, a minority, tend to be wary of encounters and practice indirect mediation.

46It should be remembered that according to the prevailing codes of ethics,4 mediators are independent and impartial and have no decision-making power. They are neither judges nor arbitrators. Their role is simply to enable the parties to communicate and to help them uncover a mutually satisfactory solution. Their only power consists in the ability to interrupt mediation in case of violence or lack of compliance with the rules of mutual respect between the parties. They may also refuse to go along with an agreement they view as particularly unfair or inequitable. Either party may decide to interrupt the mediation at any point.

47The mediators: firstly, we should note that in France, police officers and active judicial professionals are not allowed to act as victim-offender mediators. This reflects concern with introducing a different approach to conflicts. The judicial authorities may give mediator accreditation to any person viewed as competent. Assessment of this competence is subjective and in fact challenged, but serious training is increasingly required. Mediators may be professionals (about 2/3 of cases) or volunteers (about 1/3), and most work within a non-profit organization. The professionals often have training in social work, law or psychology, whereas the volunteers are often retired jurists. The professionalization process will probably be accelerated by the creation, in 2003, of a national diploma in family mediation. These family mediators (80% are women, as against 40% of victim-offender mediators), competent in civil affairs, receive much longer, more in-depth training than victim-offender mediators and are now demanding a role in family conflicts of a criminal nature.5

48Outcome of mediation: the need for mutual consent is not a vain claim, since about 40% of protagonists turn down mediation. Either the parties are not present at the appointments set by the mediator, or they explicitly reject mediation at the beginning of the process or later on. Sometimes the complaint is withdrawn, putting an end to mediation. When the process does get under way, 76% of mediation attempts reach an agreement. But this figure varies with the offence: it is 60% for domestic violence, the origin and nature of which are often complex. However, signature of an agreement is not the only criterion for a successful mediation. Some settlements are concluded somewhat formally, without any real acceptance by the parties, whereas conversely, some mediations transform the relations between the individuals, although no agreement is reached. The few available qualitative investigations measuring post-mediation satisfaction show that most people who agree to respond are satisfied. Firstly, they appreciate the fact that they could tell their spouse their own viewpoint, and secondly, the fact that they have a better understanding of what happened. A few studies show that expression of the emotional dynamics, bridled in the traditional judicial process, is a necessary requisite for an effective restorative process, for both victims and offenders (Harris, Walgrave, Braithwaite, 2004). Still and all, these dynamics must not lapse into violence and the settling of scores, but must be the moving force in a process of meaning-making, through increased awareness of one’s own responsibility, discovery of otherness and a change in the way the parties view the conflict. In this respect mediation may be said to have therapeutic effects, although it is not specifically aimed at therapy. But it is not easy to discuss the efficiency of mediation in general terms. In reality, the impact of the process rests on the mediator’s talent, theoretical and interpersonal resources and ability to set up a reassuring context, to create a climate of empathy and comprehension as well as to elicit thoughtfulness and creativity in the search for solutions.

2. Tension between two mediation models

49Mediation isn’t always a true alternative to the judicial model, and the participation of volunteers or non-judicial professionals doesn’t necessarily ensure a change of logic. In fact, an analytic approach shows there are two models for this practice, which may overlap and be conflicting, especially within non-profit organizations torn between the need to be productive, which reinforces their institutional dependency and the desire to respect the ethics of mediation and of the mediator’s position.

50The first model, which we will call a judicial model, is implemented on justice system premises and marked by judicial logics. Mediators are usually jurists with no serious, specific training in mediation. The parties, designated as offender and victims, are summoned, the criminal case record is used to orient the discussion, the mediator, investigator-like, tries to determine who is right and who is wrong, and an agreement is generally reached after a single meeting, since the mediator is directive and suggests solutions. The mediator’s directive approach is based on the application of a problem-solving logic aimed at settling the conflict and producing agreements, which accounts for the high percentage of agreements achieved. This leads some people to say that this should be called conciliation rather than mediation.6

51There is another model, which differs with the judicial logic and is more respectful of the ethics of mediation. Mediators are mostly social workers or psychologists with varying degrees of training in mediation. Usually the parties are designated as offenders and victims by the public prosecutor’s office (a few experiments have succeeded in introducing the more neutral expressions of claimant and defendant) and invited to see the mediator, only what the people have to say (and not the penal dossier) is considered, emphasis is mostly placed on communication, several encounters are generally needed, the mediator is non-directive and helps the individuals to formulate their own proposals for settlement. The theoretical framework behind these practices is “transformative mediation” (Bush and Folger, 1994), which views the offence as a mere symptom of a deeper problem. Since the goal is to get to the roots of the conflict, the mediator is not so much concerned with reaching an agreement as with transforming as deeply as possible the way in which the individuals perceive the conflict and will behave in the future.

52Judicial model

53Restorative model


55Courthouse or “maisons de justice

56Non-profit organizations, town halls




60Professionals or occasional workers paid by the act

61No or inadequate specific training

62Social workers and psychologists

63Working people

64Professionals and volunteers

65Specific training






III. Conflicts between penal and mediation logics

71The existence of these two models shows that the institutionalization of victim-offender mediation creates a cultural conflict within the criminal justice system, opposing supporters of two antagonistic conceptions of justice, one based on the highly juridical conception of conflict regulation, the other guided by a more flexible conception. This conflict, which has considerable repercussions on choices as to how to regulate domestic violence, is expressed on the cultural, ideological and economic registers.

A. A cultural conflict

72What is at stake in victim-offender mediation is not simply the improved efficiency of the judicial institution. The aim is to transform the justice system model, and thereby, the professional identities of all actors within it.

73The juxtaposition of judicial and mediation models brings out the tensions between them. The judicial model is based on the principle of vertical power, directed toward the production of legal truths, defends a collective conception of public order referring to a transcendental norm to settle conflicts. Conversely, the mediation model offers a space for democratic debate in which the goal is to unravel conflicts and restore the lost balance, for the sake of the individuals and the community, by producing subjective norms established by the persons themselves.

Criminal justice


Vertical logic

Search for the truth

Public order

Principle of rupture

Reference to law

Horizontal logic

Search for balance

Private needs

Principle of ties

Reference to equity

74The question is what happens when these two logics meet. Does one prevail over the other? Does the judicial logic force the mediation logic to conform to legal and institutional rationales? Does the mediation logic succeed in turning authoritarian relations into contractual ones and in transforming the way State authority is enforced? Do the two logics interact, and how?

75The way in which mediation transforms professional practices and identities yields a partial answer to these questions. For instance, the judge’s role is increasingly limited to referring cases, thus producing a procedural framework. Judges make increasingly fewer decisions: they check the validity of the agreements reached by the parties and guarantee they will be enforced (Théry 1993). This trend, even more evident in civil law than in criminal law, makes judges more like guardians than like arbitrators (Garapon, 1996) and illustrates the gradual shift from imposed justice to negotiated justice. Lawyers too find their work habits challenged. According to the law they cannot be mediators and only play an optional role in the legal monitoring of the process. Indeed, mediation refuses the principle of representation by a lawyer, preferring personal commitment and direct, verbal expression. Furthermore, the replacement of conflict-based logic by consensus-based logic leads them to drop the dialectic principle of judiciary confrontation so as to adopt a more pedagogical, constructive attitude. This upheaval in practices and professional identities explains why lawyers viewed the arrival of mediators on the judicial scene unfavorably, since they were seen as threatening their traditional monopoly on the judicial regulation of conflicts. While many are still hostile to victim-offender mediation, a growing number are adjusting to the trend by participating in other types of mediation.

B. An ideological conflict

76Some lawyers, as well as academics, advance arguments that are less professionally interested and more ideological, denouncing the risk represented by mediation for public liberties, or warning of the alleged dangers of the denial and privatization of conflicts.

77The claim that mediation threatens public liberties is often heard. It is said to illustrate a more general process of penalization of social life (Mary, 2004). Since it only deals with not very serious acts, it makes for the judicial treatment of situations that were previously regulated socially. Moreover, in many countries mediation completes the reinforcement of the power of public prosecutors, to the detriment of judges. Lastly, the judiciary ideology is so powerful that one doubts that the system can be humanized, and recourse to grass roots volunteers aggravates that risk, since they are more vulnerable to its influence than professionals, who are more organized and possess a shared culture and identity.

78The social discourse produced on the judiciary scene is generally reinterpreted so as to further the reproduction of the institutional logic. It is emptied of its substance and reconstructed to fit the functional needs of the repressive system (Teubner, 1989). Social actors who collaborate with the criminal justice system on the long term undergo a process of cultural colonization (Faget, 1992) leading them to unconsciously integrate the judiciary ideology (its binary logic, the legal grid of analysis of reality, assimilation of language and behavioral stereotypes and so on). Instead of being tools for the socialization of justice, they are artisans of a “judiciarization” of the social sphere. The actors’ dependency increases in proportion to the symbolic benefits they draw from their collaboration with the penal system. The risk of dependency is considerable if victim-offender mediation is the only way for volunteers to get social recognition or for professions to earn their living. The risk increases with the development of a managerial type of justice whose main concern is to handle as many cases as possible at the lowest possible cost (Crawford, 2000).

79Other researchers, who do not elude these risks, believe that mediation introduces a new “communicational rationale” (Habermas, 1987) in the judicial system, which transforms its modes of action and signification (Bonafé, 1998). They are categorically opposed to the idea of a conspiracy against public liberties, because judicial strategies are not rational, but rather they are the outcome of a series of not really controlled restructuring processes. In this view, mediation is not an alternative to the justice system model but a complement, taking into consideration the emotional and human dimension which the system is unable to handle (Milburn, 2002).

80A second criticism stresses the danger for human rights represented by the adaptability and flexibility in conflict regulation introduced by mediation. The lack of a legal framework and of legal guarantees is seen as inducing the implementation of a lower, lawless and judgeless form of justice. In particular, one alleged effect of handing penal conflicts over to mediators would be to bar access to the courts for the underprivileged, and to privatize the regulation of their conflicts. This risk is especially denounced by women’s rights movements who are opposed to mediation for family violence. They contend that mediation is a way of denying the seriousness of this social problem and of disqualifying women. The social combat against their domestic oppression must be fought on the public scene, represented by the justice system. Penal punishment of the guilty serves to clearly proclaim the social prohibition, whereas mediation, by trying to reach an agreement, dilutes responsibilities. Mediation is said to reinforce the unequal status of men and women by allowing men to escape punishment (Digneffe, Parent, 1997).

81These arguments introduce an important debate, of course, as to the types of conflicts susceptible of being handled by mediation. But this all depends on the legal framework defined for mediation. Mediation may be viewed as a derivative solution within the justice system, and it can then act at various stages in the penal proceedings, but prior to any decision. Or it may be seen as a solution integrated in sentencing, and it then becomes a probationary measure, guiding the judge’s decision. Or lastly, it may be organized after sentencing, acting as a complement to conviction. This combination of logics should satisfy all those who object to the orientation toward victim-offender mediation in cases such as domestic violence in which the rights of the weakest deserve to be asserted publicly. The declamation of rights has educational value. It proclaims the social prohibition loud and strong while setting up a process, more finely tuned than is usually the case in the judicial system, with redress for victims and the socialization of offenders.

C. an economic conflict

82The development of alternative dispute resolution (ADR) modes (Chevalier, Desdevises, Milburn, 2003) opens a new market in which lawyers’ offices and mediator organizations compete. This is particularly true for family mediation, where jurists intent on preserving their monopoly try to control mediation, arguing that one must be a jurist to understand the stakes of conflicts. In fact, mediating ability seems to be grounded more in competency in interpersonal relations, “an ordinary, practical know-how, one that is not established as esoteric professional knowledge” (Milburn, 2002). Mediators with non-juridical training actually use that argument to disqualify jurists as blind to the psychological and social stakes of the situations they handle.

83The outcome of this competition is to make the participation of volunteers increasingly marginal. Professionals feel that too many volunteers would threaten their monopoly in the judicial field. They also fear that trained, competent volunteers may jeopardize their professional identity.


84The achievements of the justice system with respect to domestic violence often seem inadequate. The legal vision of the world is too narrow to enable it to deal with these cases as tactfully and in depth as it should. But the choice of mediation by default would not be very pertinent. For as shown by the criticism mentioned above, mediation is not applicable to all situations. It is not meant to solve the social problem constituted by domestic violence, and should be conceived as one element in an overall prevention and control scheme. It cannot replace either the need to proclaim the law, pedagogically, in the most emblematic cases, or the therapeutic responses required when violent behavior or a victim posture is of a pathological nature. It may be fully effective provided (the consensual character of the process is guaranteed, and when the parties are in full possession of their intellectual and psychological capacities, when the conflicts have not yet hardened, when recourse to violence is not structural but contextual—that is, a reaction to a situation felt to be intolerable—and last, when the mediators are unquestionably competent. We must be cautious, however: although relatively serious empirical observation attests to the potential of mediation, we still lack reliable elements of evaluation for measuring the middle and long-term effects in terms of conflict appeasement and the transformation of violent behavior.

Translation : Helen Arnold

Haut de page


Becker H., (1985) Outsiders. Etude de sociologie de la déviance, Paris, A.M Métailié, (ed. originale 1963)

Bonafé-Schmitt J.P, (2003), Justice réparatrice et médiation pénale: vers de nouveaux modèles de régulation sociale? In Justice réparatrice et médiation pénale. Convergences ou divergences ?, Paris, L’Harmattan.

Bonafé-Schmitt J.P., (1998) La médiation pénale en France et aux Etats-Unis, Paris, LGDJ.

Braithwaite J., (1989) Crime, shame and reintegration, Cambridge University Press.

Bush B., Folger J., (1994) The promise of mediation, Jossey-Bass Publishers, San Francisco.

Cario R., (2000), Victimologie. De l’effraction du lien intersubjectif à la restauration sociale, Paris, L’Harmattan.

Chevalier P., Desdevises Y., Milburn Ph., (2003) Les modes alternatifs de règlement des litiges : les voies nouvelles d’une autre justice, Mission de recherche Droit et Justice, Paris, La Documentation française.

Christie N., (1977) Conflict as property, British Journal of Criminology, vol.17,1.

Crawford A., (1997) The local governance of crime: appeals to the community and partnerships, Oxford, Clarenton Press.

Digneffe F., Parent C., (1997) La médiation face aux situations de violence contre les conjointes, Politique et justice au bord du futur, Bruxelles, Bruylant.

Donzelot J., (1977) La police des familles, Paris, Seuil.

Faget J. (1992) Justice et travail social. Le rhizome pénal, Toulouse, Erès.

Faget J., (1996) Conflits privés, pudeurs publiques. Le traitement des plaintes pour violences conjugales, Les Cahiers de la Sécurité Intérieure, n°28, p.101-112.

Faget J., (1997) La médiation. Essai de politique pénale, Toulouse, Erès.

Faget J., (1999) La médiation pénale. Evaluation nationale, Fonds national de la vie associative/CLCJ, GERICO, n°11, synthétisé in .La double vie de la médiation pénale, Médiations et sociétés, n°4, 2003 et Le tensioni della mediazione penale. Valutazione delle pratiche francesi, Dei delitti e delle pene, n°3, 2000, p. 75-92..

Faget J., (1993) La médiation pénale : une dialectique de l’ordre et du désordre, Déviance et société, n°3, p.221-233.

Garapon A., (1996) Le gardien des promesses. Justice et démocratie, Paris, Odile Jacob.

Gauthier S., Laberge D., (2000) Entre les attentes face à la judiciarisation et l’issue des procédures : réflexion à partir d’une étude sur le traitement judiciaire des causes de violence conjugale, Criminologie, vol.33, n°2, p. 31-53.

Habermas J., (1987) Théorie de l’agir communicationnel, Paris, Fayard.

Harris N., Walgrave L., Braithwaite J., (2004), Emotional dynamics in restorative conferences, in Theoretical criminology, London, Sage, p.191-210.

Jaccoud M., (Dir.), Justice réparatrice et médiation pénale. Convergences ou divergences ? Paris, L’Harmattan, 2003.

Mary Ph., (2003) Insécurité et pénalisation du social, Bruxelles, Labor.

Milburn Ph., (2002) La médiation : expériences et compétences, Paris, La Découverte.

Noreau P., (2000) Judiciarisation et déjudiciarisation : la part de la poursuite et de la défense, Criminologie, vol. 33, n°2, p.55-79.

Noreau P., (2001) La superposition des conflits : limites de l’institution judiciaire comme espace de résolution, Droit et société, n°40.

Pottier M.L., Robert Ph., Zauberman R., (2001) Les victimes et le sentiment d’insécurité en île de France, Paris-Guyancourt, CESDIP.

Teubner G., (1989) How the law thinks : towards a constructivist epistemology of law, Law and socity review, 23.

Théry I., (1993) Le démariage, Paris, Odile Jacob.

Walgrave L., (2002), La justice restaurative et la justice pénale: un duo ou un duel? in Victimes : du traumatisme à la restauration, Paris, L’Harmattan.

Zehr H., (1990) Changing lenses. A new focus for crime and justice, Scottdale, P.A : Herald Press.

Haut de page


1 The low rate of denunciation is also tied to the victims’ representations of the criminal justice system, and more particularly to their perception of whether the available responses can meet their needs.
2 The first track is penal prosecution, the second dismissal and the third, prescribed by articles 41-1; 41-2 and 41-3 of the criminal code, is the sum of a hodge-podge of measures such as “rappel de la loi”, obligation of redress, health, social or occupational guidance, legalization of a situation constituting an infringement, compensation for damages resulting from the offence, victim-offender mediation and “composition pénale”.
3 An official instruction dated March 16, 2004 now recommends mediation for offences involving intimate relations, in family contexts for desertion (non-payment of alimony) and failure to present a child (one parent’s hindering the other parent’s right to visit). However, it excludes serious, repeated violence within the family.
4 Despite a few tangible nuances, there is a shared moral basis to the codes of ethics established in France by the Association for Family Mediation, the Citizens and Justice Federation and the National Institute for Aid to Victims and Mediation.
5 The hiatus between family conflicts of a civil or criminal nature is a relative one. It often depends more on the strategy of the actors than on the actual facts. Many family mediators point out that they handle the civil aspects of conflicts involving physical or moral abuse to which husband and wife admit, whereas those handled by the penal system involve alleged abuse which turn out not to be grounded in the end.
6 Conciliation being defined by its objective (conciliare meaning to reunite), whereas mediation would be defined by its method (mediare meaning being in the middle). Conciliators would to be more directive, then.
Haut de page

Pour citer cet article

Référence électronique

Jacques Faget, « Mediation and Domestic Violence »Champ pénal/Penal field [En ligne], Vol. I | 2004, mis en ligne le 10 juillet 2005, consulté le 07 décembre 2023. URL : ; DOI :

Haut de page


Jacques Faget

Chercheur au CNRS (CERVL) - Institut d’études politiques de Bordeaux

Articles du même auteur

Haut de page

Droits d’auteur


Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search