Volume 3, No. 1, Art. 16 – January 2002

Victim-Offender-Mediation in Domestic Violence cases—A Comparison of the Effects of Criminal Law Intervention: the Penal Process and Mediation. Doing Qualitative Research

Christa Pelikan

Abstract: Research on the efficacy of criminal law intervention involves social processes and due to its nature, the processes require the use of qualitative methods. It is about processes of change, about development, or fixation and about the influence, various criminal law interventions exert onto these processes. In addition there is the content of the subject matter that is conducive to the application of qualitative methods: violence in intimate relationships constitutes intricate and complex fabrics of power and love, of dependency and sexuality.

The instruments of collecting data consisted of observing of criminal processes and mediation procedures that provided access to the parties and the opportunity for intensive talks with men and women as the core piece of research. They were complemented by expert-interviews with judges and mediators. The analysis of the data involved using an "ideal type"-process analysis guided by a triangulation of perspectives as presented by the different actors.

The qualitative, "interactive" method of doing research on processes of change calls for continued intervention in criminal policy—something that goes beyond presenting evaluation results. I will present and analyze the difficulties encountered when using this approach.

Key words: victim-offender-mediation, domestic violence, "ideal type"-process analysis, triangulation of perspectives, research into intervention and research as intervention

Table of Contents

1. Research Topic and Research Goals

1.1 Women's politics and criminal policy. Challenges and dangers of conducting research in domestic violence

1.2 On client-orientation and on subjective meaning

1.3 On feelings, on passions, and on states of mind

1.4 On processes

2. How We Proceeded

2.1 The research design—position and importance of a "framing" quantitative data assessment

2.2 Triangulation of perspectives

2.3 Steps of assembling data

2.3.1 Observation of processes

2.3.2 Clients' talks

2.3.3 Experts' talks

2.4 Qualitative data analysis

2.4.1 The structural analysis of processes

2.4.2 Building a typology

3. On Research as Interaction and Involvement







1. Research Topic and Research Goals

1.1 Women's politics and criminal policy. Challenges and dangers of conducting research in domestic violence

The theme of domestic violence and the use of criminal law interventions have from its beginnings sparked off controversy and so has the research thereupon (PELIKAN & STANGL, 1994). Also the research project presented in this contribution resulted from a veritable criminal policy controversy. The core piece of this controversy was whether the instrument of Victim-Offender-Mediation (VOM) should and could be applied to cases of domestic violence, more precisely in cases of violence in an intimate relationship. In Austria, the establishment of a pilot project "Victim-Offender-Mediation in General Criminal Law", i.e. for adults, which took place in 1992, had brought about the referral of a considerable number of cases of minor assault and of dangerous threat that had occurred within the family or a man-woman relationship to the agencies of the so-called Außergerichtlicher Tatausgleich (Out-of-Court-Offence-Compensation), Austria's version of VOM. Right from the start, this practice faced critique and the opposition of the activists of the women's shelter movement.1) As a consequence, these cases did receive special in the course of the concomitant research that accompanied a pilot project. The research project "On the efficacy of criminal law interventions in domestic violence cases" that started in 1998 was intended to further challenge this critique in a constructive way and to produce empirical data on the effects and the efficacy of VOM on the one hand, the criminal court procedure on the other. [1]

One of the main arguments, the protagonists of the women's movement had led against the use of VOM concerned the supposed lack of ability (due to insufficient training and qualifications) of social workers and mediators to uncover the (male) perpetrator's defences and excuses as well as the "false" readiness of women to accede to a settlement; in other words: the VOM-workers' inability to recognise and consequently denounce "gendered" power relationships. Precisely the same kind of suspicion, or even accusation was brought forward with regard to social scientists that refrained from overtly "taking sides" by attributing greater or exclusive credibility to female victims. The "experts on violence" ("GewaltexpertInnen") had come to claim a monopoly on interpreting what is violence against women. This is quite understandable taking into account the history of the reactions of the law vis-á-vis violence in the private sphere: For many years the attitudes and—following from that—the strategies of the agencies of criminal law, first of all the police, had been characterised by looking away, by belittling and downgrading signs and traces of private violence and by general male chauvinist camaraderie (PELIKAN & STANGL 1994, p.53). [2]

Therefore the method of scientific reflection and interpretation had to live up to the challenge of confronting the doubts and meeting the demands of the experts of violence. [3]

1.2 On client-orientation and on subjective meaning

Client-orientation and focussing on the voices of the men and the women who have suffered from violence and those who have acted violently are thus one of the most salient features of this research approach.2) Concerning the research methods applied, client orientation becomes apparent as the reconstruction of subjective meaning. The active and the passive experience of violence as well as the experience of the criminal law intervention can be approached most adequately by starting to comprehend the respective subjective meaning of these experiences. Reconstruction of subjective meaning constitutes the first step in the course of the qualitative analysis. [4]

To achieve this, we have to take seriously what we are told by the persons who become law's beneficiaries ("Rechtsbenefiziare"). But starting from these narrations and going through several turns of interpretations, their meaning becomes typified and systematised. This process of analysis is an open one as the researchers will not replace or "improve" the interpretations that can be provided by the interviewees, but they have at their disposal a knowledge of theories on the subject matter and a variety of subject related facts that are used to deepen and to systematise the understanding of the narrations. The process of analysis can thus dig up a reality that reaches beyond the surface accounts of events, or points to operational mechanisms that lie underneath these accounts. Both elements, taking the narrations of those concerned seriously, and elaborating upon a scientifically guided and informed interpretation of these narrations should provide the appropriate foundation for a quality research, i.e. its cognitive value ("Erkenntniswert"). [5]

Client orientation and reconstruction of subjective meaning as means of making the voices of the parties themselves audible and strong constitutes one of the elements of a qualitative approach, of qualitative data collection and data analysis. We faced the task of apprehending (re-producing) as precisely and as differentiated as possible, the subjective perception of an event and the subjective meaning it has for that certain individual. [6]

1.3 On feelings, on passions, and on states of mind

The experience of violence is highly charged with emotions and with passionate impulses such as rage and pain, with anger, fear, and desperation and thus any account of these experiences is strongly and inevitably shaped by these emotions. [7]

This does not in itself demand the use of qualitative research methods. One might as well attempt to find out about emotions and impulses by means of a questionnaire. This might have the advantage of allowing the persons questioned to answer more freely and in a more distanced way and could eventually produce answers that are exact and truthful. How does it come that we consider themes like divorce, or private violence that are fraught with mostly negative emotions almost inevitably require the application of qualitative methods of research, the interview, or the open talk? Very likely it is the trust in the explorative power of direct interaction, and in the narration as a means of transportation, or as a "carrier" for complex and vague states of mind, for multi-layered and deeply-rooted feelings that makes for this choice—not inevitably but arguably grounded. [8]

Linked to the role of direct interaction holding and carrying subjective meaning that is tinged by emotions and impulses, there is an ethical question which researchers have to deal with; this questions became rather pressing also in the course of this research project. How intrusive can you be when talking to men and women about their experiences in intimate relationships? Does the scientific interest, the gaining of empirical knowledge justify our digging up feelings that have been painful and will be painful to recall? The justification used frequently by researchers—and we did so too—namely the benefit other people in similar situations could possibly derive from the findings of social science, in other words: the appeal to altruism, albeit not futile, appears hollow. On the other hand, for quite a few of the interviewees, the interaction turns out to be a fruitful experience; they meet someone with a keen interest in the stories they have to relate, they get a chance to clarify for themselves certain aspects of what has happened, and sometimes by talking about their experiences they succeed in seeing through confusions and entanglements that went along with these events. Nevertheless, the researcher might still have to deal with a certain amount of ambivalence and uneasiness when working in this field and in this way.3) [9]

1.4 On processes

The most important feature of the subject matter at stake that is conducive to the choice of qualitative methods is its quality as a process—in a twofold sense: On the one hand, there is mediation and the criminal procedures, we have observed and analysed, being processes by definition; on the other hand, we have come to understand the effects of these criminal law interventions—within the context of other variables exerting an influence—also as processes. One could contend that the temporal structure of the subject matter finds its correspondence in the temporal structure of the research tool. The single, non-recurring assessment of static conditions, of the state of a person's mind and of a relationship, cannot really grasp the essence of these processes. (It might serve as a valuable contribution to the research design though, in the way of a triangulation of methods.) [10]

The process mentioned first, i.e. the mediation or the court procedure served as a starting point for data collection and as a focus of the "quest for knowledge and cognition" (Erkenntnisinteresse). The various research instruments used such as observation, experts' interviews, and clients' interviews were all centred around this process. [11]

The processes set in motion by the interventions of criminal law were reconstructed through the intensive talks with the clients and especially through follow-up talks after several months. What the men and women told us during these talks/interviews presents their subjective "personal" interpretation of these processes of change or of persistence; it is their subjective meaning that comes to the fore that is of importance and that calls for a scientifically guided interpretation, more concretely the use of qualitative data analysis. [12]

The effort of interpretation constitutes the core of the scientific endeavour. It is a difficult and it is—because of the nature of the subject matter—a politically sensitive piece of work. [13]

One more clarification concerning the theme of research seems necessary: The subject of research was not "Domestic Violence" or "Violence in Intimate Relationships". Concerning the phenomenology of violence one can only point to the vast amount of the relevant literature on that topic; meanwhile several scientific journals are dedicated exclusively to it (for a selective overview: cf. PELIKAN 1990). The research project presented, concentrates on the interventions provided by the criminal law: starting with police being called in, the "negative" discretion exercised by the state procurators and finally the actions of the judges and of mediators, in Austria: the social workers of the VOM-services. The latter are also the partners of an interactive research approach, to be outlined in the third and last section of this contribution. [14]

2. How We Proceeded

2.1 The research design—position and importance of a "framing" quantitative data assessment

The research design consisted of three steps: 1) observation of procedures, 2) clients' interviews (talks) and and follow-up talks after several months, and 3) interviews of the professionals handling the case. [15]

This qualitative data collection was supplemented by a "framing" collection of data from files of the ATA-bureau in Vienna. These files contained information about all the cases pertaining to violence in partner relationships that had been referred to the ATA by the Vienna state prosecutors' bureau between January 1st 1998 and April 30th 1998. The information regarding the type of the case, the personal characteristics of the alleged offender, as well as the victim and the way the case was handled and settled by the mediators, and lastly by the state prosecutors was derived from these files by using a short assessment instrument.4) (The ATA-files contain copies of the most important parts of the state prosecutor's files, to start with the protocol of the police investigation.) There were altogether 54 cases. Using the same approach, we collected the respective information from the court files of the cases we had observed at Viennese courts (mostly the Higher Criminal Court—Landesgericht fuer Strafsachen) during three months (April 28th 1998 until July 31st 1998). Out of 38 cases we had observed, the files for 31 cases were available. The ATA-data as well as the court data were subjected to a simple "linear" data computation and an interpretation. [16]

This so-called "framing assessment of quantitative data" was done to get an impression of the overall statistical distribution of characteristics of domestic violence cases that are handled by way of VOM. We arrived at the decision to do this kind of quantitative assessment, because the first cases we had observed appeared to have very outstanding and extraordinary features. On the other hand, for the sake of a comparison, we wanted to get some information on the characteristics of the respective group of cases of partnership violence where court proceedings had taken place. Albeit one has to bear in mind that the overall proportion and the characteristics of the cases, handled by VOM as well as through a criminal procedure, cannot be assessed in a definite way, without having information on the respective characteristics of all cases that came to the notice of the police and of which, according to our knowledge, a majority is disposed of by just dropping the charge. To achieve this, a prospective data assessment procedure ought to be designed with the state prosecutors systematically noting the features of all cases received, prior to making a decision which line of handling to follow: dropping the charge "a limine", referral to VOM or going ahead with an indictment. [17]

2.2 Triangulation of perspectives

The three steps of data assessment I have been talking about could more adequately be described by a concept—or rather by the metaphor—"triangulation", as referred to it by Udo KELLE (2001). Predominantly the talk is about a triangulation of methods. According to Norman DENZIN (1978, p.304) it is a complex process of mutually confronting different methods and consequently maximising the validity of field research. But KELLE has argued that triangulation could also pertain to the description of different aspects of the same phenomenon by using quantitative and qualitative methods, depending on the aspect it is focused upon. Altogether he discerns three different meanings, or types of triangulation: one is about achieving mutual validation of different methods applied, the second is about gaining a more complete picture of a certain phenomenon and the third is concerned with the need for bringing together quantitative and qualitative methods aiming at an explanation of sociological phenomena. In the case of the research presented, still another version or metaphorical application of the concept of triangulation that first found its use in trigonometry comes to the fore.5) [18]

The triangulation of perspectives that was already applied within an earlier "accompanying research" project on "Family Mediation" is about a methodological approach that attempts to take various points of observation (in a broad sense) aimed at getting a perspective on certain processes and events. It intends to achieve greater variety and greater density of description and starting from that a higher degree of complexity in grasping and understanding the relevant phenomena—in this case the efficacy of criminal law interventions. But this choice of a methodological approach is also inherently recommended for the subject matter of this research project: what else should one do when trying to learn about violence in intimate relations and about the way mediation and the court procedures affect these experiences, than bringing together and confronting the experiences and the perceptions of all the actors involved? [19]

2.3 Steps of assembling data

2.3.1 Observation of processes

The first step—in chronological order—implies also the type of triangulation Uwe FLICK refers to when talking about a triangulation of perspectives (cf. Note 5). The perspective of the researcher that is observing court or mediation procedures can bring to the open the power relationship underlying the communication of a man and a woman, as well as the power relationship determining the communication between judges and mediators on one side and the parties on the other side, i.e. the structural aspects of the matter at hand. The interventions set by social workers and/or judges will be described from the distanced point of view of an outsider. This outside position of the researcher adds a perspective to those of the persons involved. Moreover—it is not just an addition, but owns the quality of introducing a meta-perspective, and this means, a perspective located at a different level. As with any scientific observation it can be characterised as "second order observation" as referred by Niklas LUHMANN (1993, pp.403f.) This does not "automatically" imply the claim to and the ownership of some superior truth; the specific quality of this standpoint has to be taken care of in the course of the data analysis though. [20]

To be more specific: The observation of both court procedures and VOM procedures has a double purpose: it ought to provide access to the (ex-) partners, where an incident of violence had come to the notice of the police and consequently the state prosecutor, and it ought to supply us with an additional set of data, i.e. offer an opportunity to hear and to see how the parties interact with each other and how they interact with the agencies of the criminal law, including the set-up of VOM. [21]

We concentrated on observing the actual mediation sessions (involving the partners and the social workers) at the Viennese ATA bureau in Vienna and in Salzburg. One has to know that the VOM procedure usually starts with the state prosecutors referring a case, they deem appropriate according to the exercise of their discretion, to the ATA-bureau, that contacts, usually by letter, the alleged offender and then the victim—this being the sequence followed in most cases. It can be reversed though, as it happens in quite a few cases of domestic violence. Depending on the consent of both parties, first individual talks are arranged. The method followed in Vienna in cases of domestic violence has been predominantly that of the so-called "mixed double" and it implies joint responsibility and cooperation of a male and a female social worker. Very often the clients, man and woman, are asked to come to the ATA-bureau at the same time, and there the female social worker talks to the woman in a separate room and the male social worker talks to the man. In these "single talks" they ask the parties about their concrete experience of the incident that was reported,—or led to the police being called. Previous incidents and the whole state of the relationship are further explored, and also, last but not least, the expectations concerning an agreement, including the content of material and non-material compensation, the intentions and concrete proposals of the offender for reparations or compensations and the victim's demands and wishes concerning the future of the relationship on the other, i.e. either the conditions for separation or for staying together. When both partners have finished, the four of them: the partners and the two mediators get together for the mediation session ("the talk of the four", "Vierergespräch"). The session with four participants is the core element of the whole procedure. Its course follows a rather sophisticated and elaborated professional design that aims at bringing into effect the two main working principles of mediation: recognition and empowerment. At the beginning of this session the sitting arrangement is as following: The mediators sit opposite facing each other with their respective clients on their side. The mediators tell each other what they have heard during the previous single talks: the story of the relationship, the account of suffering and of acting violently, of threatening, hitting, constraining the other's freedom; they tell what the man and woman have done, what they have seen, what they have felt. The partners are asked to listen without interfering, and only afterwards they have the opportunity to comment, to correct, and to modify the rendering of their story told by the mediator. This is also the beginning of the immediate exchange of the partners—about their perceptions, and their expectations. [22]

I cannot get into the ways the mediators facilitate the active exchange between the partners in more detail.6) Only that the "distancing effect" (or alienation effect) achieved by hearing one's own and the partner's story told by another person, this "changing of lenses", promotes "re-cognition"—(literally, getting to know again) of one's own standing, and re-cognition of one's own needs and interests, beyond those of a formal legal position, or standpoint. Having found understanding and en-cognition (an attempt at translating the German "Anerkennung" or "Würdigung") proves a prerequisite for empowerment—empowerment that aims at balancing existing power imbalances, and thus lends support to the weaker party. The core elements, or working principles of any mediation process: recognition and empowerment are thus displayed in a particularly strong way through the professional setting designed and applied (for a more detailed account, cf.: PELIKAN 1999 passim). [23]

Since the setting of mediation is apt to produce the effects described, the task of research consists in following the ways this happens and in delineating the long-term effects the interventions exert on the lives of men and women. More specifically: observation of procedures meant sitting in the background and listening while—first the single and then the mediation sessions took place. (We had first procured the permission of the parties to do so which was denied in only one case.) We observed individual talks with women as well as with men. In the beginning we had some doubts whether the presence of a woman researcher might hamper, or even distort the flow of communication. But finally we did both and decided whether to go with the female or the male social worker according to the impression we got from studying the files. The difficulties we faced when observing the VOM-sessions stemmed from the fact that the presence of the researcher is sometimes strongly felt in a small room with only two people talking to each other; this could only be dealt with by carefully explaining the role and the task of scientific research. In that respect the observation of court procedures posed less of a problem, the criminal procedure being open to the public and the courtrooms designed to accommodate a larger audience. [24]

In the end, we had assembled 36 observation protocols of court procedures and 25 protocols from VOM-sessions, comprising individual talks and mediation talks. These were hand written notes taken during the sessions and typed out afterwards. They follow as closely as possible the course of the procedure. In addition we attempted to characterise the setting and the performance, i.e. the mode of acting of the participants. The protocols contain verbatim statements and comments of the acting persons and often a few concluding remarks from the point of view of the researcher. [25]

Concerning the second task of observation: for recruitment of clients for the intensive interviews planned, we faced considerable difficulties—this time more frequently at the courts. We had decided to approach the parties right after the procedure was closed. But it soon became obvious that it was almost impossible to contact both the victim and the offender and to ask them whether they would be prepared to have a more lengthy talk at some other time and place. This would have afforded the presence of two of us, preferable a man and a woman. We therefore restricted the establishment of contact to the woman, who was more often, albeit not always, in the role of the victim. Only in few cases, mostly those where the relationship continued and the partners left the courtroom together, we succeeded in making arrangements for talks with both the man and the woman. Since the setting of VOM usually implies man and woman coming together, it was easy to establish contact and to make the necessary arrangements for future talks. We refrained from this attempt in a few cases where the strain exerted on both partners during the mediation sessions was very heavy and any request for further communication would have put too much pressure on them. Twice it was the parties who told us that they did not want to talk at length and in the frame of a special interview about their experience with the VOM-procedure. In both cases we had asked them on the spot and very briefly about their appreciation of the course of the intervention and its outcome and about their expectations before and after the VOM-procedure. [26]

2.3.2 Clients' talks

Final appointments for these talks were made by telephone. (To ask for the client's telephone number is the most, I would even say, the only reliable mean of securing contact. The law of action lies with the researcher and it is up to him/her to call and arrange a meeting. Once this is achieved, continuation of contact usually works very well.) Interviews took place in clients' apartments, at the research institute or in coffeehouses or pubs; they lasted from three quarters of an hour up to more than two hours. With the permission of the clients they were tape-recorded and transcribed—though not as a specific linguistic transcript—partly completely and partly as an excerpt. General reflections concerning techniques of carrying out talks of this kind and of stimulating narrations had already guided our work in the context of the research project on "Family Mediation". This time we used a modified version of a point of departure that Aaron CICOUREL had kindly—and personally—recommended: The evocation of a concrete event in the course of the criminal law intervention and the concomitant feelings and apprehensions should serve as trigger for the stream of memory and set in motion the narration of past experiences: "Waiting outside the courtroom/the inner rooms of the ATA, what did you think, will happen? What were your hopes? What were your fears?" Thus very often these were the starting questions. [27]

In addition we had a kind of checklist, i.e. a compilation of themes to be covered, to guide us through the talks. The talks went on in an open and associative way, following as closely as possible the stream of the narration of our counterparts. This accounts for the great differences in length of time, the interviews took. [28]

The courtroom- or VOM-experience constituted the node of events; from there the narration of the clients unfolded both into the past and into the future. Especially for the women who had experienced violence, the incident, or several incidents of violence formed the centre of their narration. Further focal points to be covered, were the mobilisation of the police and the way they reacted, starting with taking down the report and interrogating the suspect, up to making the offender leave the premises or placing a "restraining-order" (according to the "Law on Protection from Violence") or an arrest-order. Other topics covered in the interviews included: the events that took place in the aftermath of the police intervention, the considerations that influence the decision to hold up an accusation in the court or to step back (which is possible in cases where the suspect is a relative of the witness), then the acceptance or non-acceptance of the offer to participate in a VOM-procedure, finally the ways the court procedure itself or the VOM-procedure, the individual sessions and the mediation sessions were perceived and experienced. [29]

The ways the criminal law intervention imprints upon the future of men and women, forms the other important focus of the research effort; there the difference between material or outcome-oriented justice on one hand and procedural justice on the other, received special attention. In addition, we applied a broader concept of efficacy that was not restricted to the preventive effect as indicated by recidivism, but attempted to perceive and to understand the whole life-situation and the configuration and (re- or de-) construction of the relationship. At the end of the talks, we had therefore asked the clients permission to call again and to ask how they had got along and also, how they judged the court or the VOM-intervention after some time had elapsed. [30]

Also for the follow-up interviews we had designed a guiding checklist; it was discussed and reviewed by a group of experts consisting of representatives of the ministries that had commissioned the research project (the Ministry of Justice, the Ministry of the Interior and the Ministry of Family and Youth Affairs) the bureau of the Minister of Women's Affairs and of the so-called "Intervention Agencies" (that have been established to implement the "Law on Protection from Violence"). Several of these follow-up interviews were done by telephone—depending on the client's preference. The problem of research as an intrusion became once again relevant. In several instances we met a rather defensive reaction like: "Why are you bothering us again, why are you meddling into our affairs again?" Partly this resulted from the fact that notwithstanding our assertions to the contrary, we were seen as part of the controlling apparatus of the criminal law. It proved difficult to really explain the role of research within this field of action. On the other hand, the opposite kind of reaction was quite frequent and it came from the women: those who expressed gratitude that there was somebody who took an interest in their life and the way they had tried to handle a difficult situation. [31]

A remark yet on interviewing men. How to speak—as a female social scientist—with men and to men who have acted violently against their wives? Can one expect anything but strategies of defence, of avoidance and of self-justification? The Scottish social scientists Kate CAVANAGH and Ruth LEWIS (1995) have dealt extensively with the difficulties of doing this sort of interviews and/or talks. But they refer to a research project that investigated the phenomenology of violence against women, i.e. it was intended to find out about the circumstances and about the psychological and the social dynamics that instigates male domestic violence. Our theme was—as we have emphasised above—a different one. Also the fact that we had been present during a procedure in whose course at least one incident of violence had been scrutinised to some extent, created a different point of departure. It was the interventions of the agencies of criminal law and its representatives that we concentrated upon and therefore the talks with men proceeded without any difficulty. But we met with very aggressive behaviour in some cases where we tried to contact men after a criminal procedure right outside the courtroom. This tells of course something about the constellation of male violence as well as of the court intervention. On the other hand, sometimes it also happened that in the course of the talks, a movement toward something new surfaced, a tiny bit of an insight that could bring about change. [32]

As already mentioned, the clients' interviews as well as the follow-up interviews were transcribed. Finally we had altogether transcripts of 76 clients' talks. In only one case this material was complemented by an interview with the psychologist of the shelter where the woman had resided for some time. According to our original research plan, talks with representatives of institutions that provided help to women and men in situations of violence would have been an important addition to the information gained by the clients' talks. This proved impossible to achieve. [33]

2.3.3 Experts' talks

We did two kinds of experts' talks (or interviews). One was the type of the case-related interview, done with the social workers/mediators immediately or shortly after they had conducted the mediation session we had observed. We had very few of these case-related talks with the trial judges or the responsible state prosecutor. There were organisational obstacles we could not overcome. These case-related interviews are an important contribution to the proceeding of triangulation. They introduce one of the perspectives that have to be brought together with the perspective of offenders and victims and with the researcher's perspective as it emerges from the observation protocols. Since the mediation and the court procedure as an intervening process were at the core of our research interest, the way the professionals saw themselves and their actions and the way they appreciated and positioned their importance, constitutes an indispensable part of the whole picture. [34]

In addition, we conducted a series of more lengthy talks with judges and with state prosecutors—experts' interviews in a more narrow sense—about the problem of criminal law intervention in domestic violence cases. Following a short questionnaire, we covered a range of topics. With the state prosecutors it was their discretionary powers concerning the indictment, and concerning the referral to VOM that we mainly talked about. One has to bear in mind that one of the intentions of the research project consisted of looking at the development of criteria for the referral practice of the state prosecutors, i.e. criteria for referring a case to VOM, for dropping the charge or for making up an indictment. I will come to the respective results of this research in the following section. [35]

Apart form this, the talks produced important statements that complemented the framing quantitative research and the analysis of the VOM-cases. These statements related to the modes of cooperation between the courts and the VOM-agency. One of the pivotal conclusion to be drawn from these interviews concerns the need to provide extensive and intensive information to judges and state prosecutors with regard to the function, the potential effects, and the potential value of VOM. [36]

2.4 Qualitative data analysis

2.4.1 The structural analysis of processes

This is a proceeding of qualitative data analysis put forward by Uta GERHARDT (1991) and we did apply it in the course of the research, accompanying a pilot project on "Family Mediation". In its course, we made quite a few adaptations that appeared necessary and inevitable considering the subject of this research: the legal termination of partnerships, of love-affairs and sexual relationships. We had also developed a model or flow-chart depicting the diachronic course of events, but also the various forces of influence that came to bear upon the use, or the mobilisation of the agencies of law and of the mediation procedure, and on their preventive and "peace-making" effects on the lives of men and women. Within the family mediation project, the transcripts of the clients' talks as well as the case-related talks with the judges and mediators constituted the starting point of the analysis. A few cases had first been interpreted by the members of the team of researchers in a holistic way, and guided by a few variables (or "dimensions") taken from the literature on mediation. In the course of regular team sessions, the different "attempts at reconstruction" of cases were discussed. After a first series of sessions and with the material piling up, the opportunities for comparing and for contrasting cases increased rapidly and consequently so did the possibility for an investigation of the explanatory potential of the differences contained in the dimensions first chosen, until we arrived at an extension and a regrouping of these dimensions to be moulded into the model of mediation effects. The procedure of mediation was the focus of our attention. What happens within this process, we perceived as depending on the kind of communication that is taking place between the partners, the communication at the time of a divorce or a separation being shaped by the ordinary everyday mode of communication as it was characteristic in the relationship. Following from that, the model came to serve as a guiding chart for the interpretation and the analysis of the whole of our material. It took the place of what Uta GERHARDT had attributed to the construction of "ideal-type case sequences" ("reine Fallverläufe"), a step we could not follow, due to the tremendous complexity of our subject matter. Ultimately, the model of the sequence and the effects of mediation can assist in the understanding of structures. [37]

Also in the research project on criminal law interventions in domestic violence cases, the narrations of clients constitute the core element and the starting point of the qualitative analysis. Unlike the analysis within the accompanying research on family mediation, we had at our disposal an additional set of data, namely the observation protocols: the method of a triangulation of perspectives could therefore be already used at the step of reconstructing and interpreting single cases. This triangulated reconstruction was concerned with reflections on the text, i.e. on different kinds of talks and the narrations produced, it looked into interviewer-influences and into the specific temporal and social placement of the talks. We also directed our attention to the fact known from socio-linguistic and ethnological research that what is told in an interview is to convey a certain message to the listener and to create a certain image of the person, be it that of offender-victim, weak victim, or fighting-back victim; that of victimised offender, of helpless offender or of reformed offender. Rendering the narration, implies the construction of an identity; it is part of the "politics of constructing an identity" (Identitätspolitik). (GOTSBACHNER 1999) [38]

By confronting and bringing together different perspectives, simultaneously a multi-layered picture with more of a contour emerges. With the help of the model of mediation sequences and effects, which was only slightly modified, we arrived at an understanding of underlying structures of processes of change, and of the influence the criminal law interventions exert. Finally we could draw up a typology of these effects and influences. [39]

We did concentrate on the following factors, or forces of influence: the severity and the type of violence brought to the attention of the police, the constellation, or the balance/imbalance of power inside the relationship of the partners; their individual socio-economic and personal resources (qualifications, communication skills), and finally, the dispositions, attitudes and coping strategies acquired in the course of the life, (the socialisation process), especially the attitudes toward state agencies and authorities. [40]

Within the diachronic model altogether three sets of variables were represented: 1) constellations of conditions (as listed above), 2) sequences (courses) of events, and 3) interventions and effects/outcomes. [41]

As already explained, the analysis began by considering single cases, followed by comparing cases and by contrasting cases alongside the different aspects or variables. Within each of the three sets of variables, we introduced relevant differences ("differences that make a difference") and from there—by establishing correlation between the differences of conditions, of sequence and of effect—we arrived at building a typology. [42]

2.4.2 Building a typology

It needed, in fact, several attempts at comparing and at contrasting cases to dig up viable difference on which to build the "typology of the restorative process" that is presented below. [43]

I have to explain that in the beginning, we, and even more so the Ministry of Justice, had hoped for the research resulting in a typology of cases of domestic violence. We had anticipated that it would serve as guidance, or even as instruction for the selection and placement of the different types of cases of domestic violence either with the ATA, or any other of the diversionary measures set out in the "Criminal Procedural Law Reform 1999". It would also help to decide about dropping the charge or going ahead with a full-fledged criminal procedure. Due to reasons that are related to the content of our findings this was not achieved. Particularly, the analysis of the clients' talks, especially those with battered women provided us with new insights into the meaning and importance of relational power structures, the role of violence and the dynamics of change. These insights led us to modify earlier perceptions that were built exclusively on the reports of the mediators (that constituted the material of the accompanying research on the pilot project "VOM in General Criminal Law"). It became evident that the decisive element in assessing the feasibility and suitability of mediation lies not with the power position and the behaviour of the violent partner, but rather the—inner and outer—situation and disposition of the woman and whether she can muster resources to break free from an oppressive relationship or alter it fundamentally. Sometimes this happens even after a long time and a long history of violence. But women whose position inside the relationship is a strong one, are readier and less reluctant to take concrete steps to counteract violence, including the mobilisation of the criminal law, i.e. as a first step, calling the police. Anyway, these facts and conditions are almost impossible to assess beforehand and just based on reading the files. [44]

We have therefore arrived at proposing a different strategy, namely the "procedural placement or diagnosis of cases' to be done by the state prosecutors with the assistance and cooperation of mediation services and victim support services. This is based on the research finding that it needs a whole range of interdependent features and conditions to become known, and be brought to the surface, to find the best way of handling the case and procuring further support. The assessment/diagnosis itself should therefore be perceived and organised as a process. In its course—step by step—the array and the sequence of interventions have to be decided. [45]

Concerning the efficacy of VOM we found the following types of cases:

  • VOM as reinforcement of change; here we find two sub-types:

    • VOM reinforcing change as a mutual effort, initiated by both partners;

    • VOM as a reinforcement of change, enforced by the woman who gets confirmation of a her claim to freedom from violence in the private sphere in the course of the VOM-procedure;

  • VOM as the beginning of reformation (the individual prevention effect—a rather rare constellation);

  • VOM supporting separation; and finally

  • VOM at its limits; i.e. cases where VOM's effort proved futile and where violence occurred again. [46]

Taking into consideration the criminal law intervention investigated, the criminal court procedure as well as the VOM-procedure, we came to realise that these procedures are effective mainly as reinforcement of dynamics already set in motion, i.e. of changes and of efforts that were brought about either by both partners or by the woman only as a consequence of the occurrence of violence that was made public by calling in the police. This is more pronounced with the out-of-court procedure of mediation, a procedure that is apt to address deeper-lying relational power structures, to make them visible and to reinforce their transformation. Only very rarely does something new start, and a conversion, or a reformation of the alleged perpetrator takes place; this holds true for both the criminal procedure and for VOM. [47]

Where we find an instrumental use of the criminal law interventions, i.e. where the appeal to the agencies of criminal law is intended to signalise a warning to the partner that has acted violently, both—the formal and the informal procedure—are quite well suited to fulfil the function of "drawing a line". In other words the function of an affirmation of the norm can be achieved quite well by the court procedure, but also by the VOM-procedure. [48]

3. On Research as Interaction and Involvement

The larger legal and social policy oriented goal of this research was to find out and further develop modes of intervention, i.e. modes of support and assistance in the widest sense that will counteract violence in the private sphere, especially violence in intimate relationships. These modes of intervention aim at immediate protection of a partner that is threatened, as well as at the prevention of future threats and acts of violence directed against an (ex-) partner. One has to bear in mind that the criminal law intervention constitutes just one of the wide array of instruments and measures that ought to come into effect and ought to mutually complement and enhance each other. [49]

It is—as we have outlined above—research dealing with processes and with the efficacy of processes (therefore with processes in a twofold sense). The subject matter of the investigation in conjunction with the overall goal of the research endeavour suggests a research attitude that reaches beyond a mere presentation of results, in this case beyond the elaboration of a typology of sequences and effects of criminal law interventions. The typology of the restorative process we have presented needs to become exposed to a process of exchange: These processes need to be scrutinised by the social workers whose performance we have observed, and about whose effects we have learned from the clients. Put differently: This typology and the contentions it carries ought to be used as indications or "pointers' that will help to focus and to frame the attention of the social workers. To achieve this, an additional sequence of interpretation would be necessary. There the social workers would meet the challenge, to comment on the interpretations and the typologies suggested by the researchers. They should tell to which degree they appear plausible, in which way they correspond their perception, where they contradict them and what additions or alterations they would propose and why. We are proposing to set up of recursive and circular sequences (or turns) of interpretations, or to put it even more strongly: processes of permanent (ex-) change. [50]

The same requirement holds for the proposal of a "procedural diagnosis or placement of cases" we have put forward in a previous section and presented in our list of recommendations to the ministries. It also rests on the condition of an ongoing process of exchange between the professions involved: the state prosecutors, the judges, the mediators and the social workers in victim assistance institutions. Again, our typology offers just a few hints on how to commence. [51]

This is not to downgrade and devaluate the importance and the potential of social research. Scientifically systematised interpretation and outlining a typology, grounded in this kind of interpretations, is fundamentally different from the partial interpretations of various actors involved as it is free from the burden of active involvement ("handungsentlastet" according to Jürgen HABERMAS [1973]). But to make it relevant, it has to be brought back into the arena of action. [52]

Besides: a typology is derived from new clusters of differences being invented and introduced, and some of them have never been used by the actors. "In this way", says Niklas LUHMANN, "the observer becomes a protagonist of enlightenment, albeit enlightenment works only insofar he/she uses differences that the one to be enlightened is able to take on board" (1987, p.654). We might add that taking on new differences should happen as result of discussion and of persuasion and must not be imposed upon the actors. [53]

But this has proved hard to implement—very often our call for commenting and discussing, what we had presented, when the task of compiling and analysing the material was finished, was and is met with some uneasiness. There prevails the opinion that the researcher is the one to know, the one to tell what is "fact" and to tell what would be the right path to follow. We are confronted with an attitude that is a mixture of a grudging trust in the authority of science and a readiness to push aside the results of the same science whenever they appear inconvenient; they are then reproached for missing the reality. Confrontation and mutual exchange does not take place. [54]

Anyway, in regard to the research project presented here, we did not succeed in that respect. In a pilot project "Out-of-court conflict resolution in juvenile justice" established in 1985, we had loops of feedback between researchers and other professionals, especially the social workers were an integral part of the research project. And at that time we succeeded in starting and in keeping up an intensive and fruitful exchange of the sort we have outlined above; accompanying researcher functioned as an instigator for systematic scientifically guided ongoing reflection. In retrospective, two factors can be discerned that were influential in establishing a positive exchange: 1) this was an innovative, even an avant-garde project; (almost) everything was new and the actors saw themselves as avant-garde; 2) a hierarchy of knowledge and competence was not (yet) existent because everything was new to everybody. Researchers were co-actors and the VOM-workers were co-researchers. One remains doubtful whether this constellation, or any functional equivalent can ever be reproduced to realise the ideal put forward, the ideal of research that takes part in criminal policy efforts. Maybe it needs another attempt at a new and innovative criminal policy to make it come true. [55]


I thank those women and men that have talked to me. I still nourish the hope that my work, either immediately or in the long run, proved useful to them.


1) I wrote more extensively about the different aspects of this critique in the introduction of an "expert's opinion" I did for the Hamburg Senate for Gender Equality (PELIKAN 1999). Put in a nutshell, the critique of VOM contains firstly, the reproach for insufficiently asserting the norm of non-violence in the private sphere, a norm not yet generally accepted, secondly, the accusation of being apt to exacerbate power imbalances due to the inner structure of mediation that might provide a forum for exerting individual (male) power and enables batterers to turn to all kinds of defence strategies; thirdly there exist apprehensions because of the quality of VOM as a short-time intervention that disclaims to be held accountable for any further incidents and developments. <back>

2) At the Institute for the Sociology of Law and Criminology exists a long-standing tradition of research on the institutions of the law, especially criminal law ("Justizforschung") focussing on the beneficiaries of the law, the consumers and clients of legal procedures. As early as 1981 Gerhard HANAK has observed criminal procedures at magistrate's courts and established empirical findings concerning the management of conflicts by the agencies of the criminal law. This laid the foundation for the "Ethnology of Conflict Management" (HANAK 1987), that was further expanded in the course of a rather large empirical research project whose results were presented in the book "Ärgernisse und Lebenskatastrophen" (Everyday Nuisances and Life-world Catastrophes, HANAK, STEHR & STEINERT 1989). This approach was further inspired by research in the field of contemporary ethnology, and more precisely the ethnology of law by Sally ENGLE MERRY (Getting Justice and Getting Even, 1990) and by John M. CONLEY & William O. BARR (Rules versus Relationships, 1990). In the context of the research activity by the Institute for the Sociology of Law and Criminology on the application of alternative (mediation) procedures, we emphasised on consumer orientation and asked about the concrete experiences of parties (and clients) who had been the beneficiaries of these procedures. We asked them about their satisfaction and the way this experience affected their affairs. Another opportunity for this kind of approach was available when we did additional research on a pilot project on family mediation, comprising divorce mediation as well as mediation in cases dealing with problems of custody and access. Austria was a laggard in this realm (not in terms of VOM where it could be characterised as vanguard). Within this project we compared couples, those who had been through mediation to those who went the conventional way of the court procedure only. Contact with mediation couples was established through the mediators, while we had to use the court to send out letters to our sample of "court cases'. The intensive talks with the men and women were repeated after about one year to up to one and a half years. In addition we did case-related interviews with the mediators and we got at least short comments from the judges, who had handled the cases we had investigated. It was essentially the same design we applied to the research project: "The efficacy of criminal law interventions in domestic violence cases." <back>

3) As early as 1980, Klaus WAHL, Greta TÜLLMANN, Michael-Sebastian HONIG and Lerke GRAVENHORST had addressed this problem while doing a research project at the Munich "Youth Institute". It was on the lives of lower class families and was supposed to provide basic knowledge and give support to a new family policy. In the introduction to the book that was titled "Familien sind anders" (Families are different) the authors expressed their doubts and apprehensions—despite the fact that their methodological proceedings had proved quite successful as well as widely satisfactory. Thus there summing up comment reads: "What gives us a bad conscience is the prevailing tendency to get close to the lives of individuals and to get it into the administrative grip by increasingly subtle methods of comprehension and by continuously developing means of data processing." About ten years later I found out about another attempt to tackle the problem. In Norway, a group of feminist sociologists of law associated with Anne HELLUM linked their research on the family situation of women and their participation in the labour market by giving free legal advice. I wonder whether this is manageable, perhaps only in rare circumstances. <back>

4) The framing quantitative data assessment was done by Bernhard HÖNISCH who cooperated in this project. The observation of procedures as well as the talks with clients and experts I had to manage without any help. The interpretation of data was again a joint effort. <back>

5) FLICK understands triangulation of perspectives as the use of various approaches within a comprehensive qualitative research design, i.e. the combination of interpretative and reconstructive methods. We found still yet another field of application of the concept of triangulation—one that is of special interest in this context: At the interfaces of psychoanalysis and systemic family therapy the phenomenon of triangulation with respect to parent-child-relationship denotes the fact that the father-child relationship contains its meaning and importance not only in itself, but also with a view to the mother-child relationship. This could be called a "triangulation of effects"! <back>

6) A more detailed description of the procedure applied within the Austrian ATA and of the special methods that have been developed to deal with the problem of domestic violence can be found in the research report at the Institute for the Sociology of Law and Criminology: "The efficacy of criminal law interventions in domestic violence cases" (HOENISCH & PELIKAN 1999, pp.157ff.). <back>


Cavanagh, Kate & Lewis, Ruth (1995). Interviewing violent men. Challenge or compromise? In Mariana Valverde, Linda MacLeod & Kirsten Johnson (Eds.), Wife assault and the Canadian criminal justice system: issues and policies (pp.86-112). Toronto: University of Toronto Press.

Conley, John M. & OŽBarr, William (1990). Rules versus relationships: The ethnography of legal discourse. London/Chicago: The University of Chicago Press.

Denzin, Norman K. (1987). The research act. A theoretical introduction to sociological methods (2nd edition). New York: McGraw Hill.

Engle Merry, Sally (1990). Getting justice and getting even. Legal consciousness among working-class Americans. London/Chicago: University of Chicago Press.

Figdor, Helmuth (1994). Kinder aus geschiedenen Ehen: Zwischen Trauma und Hoffnung (4th edition). Mainz: Matthias Grünewald-Verlag.

Flick, Uwe (1991). Triangulation. In Uwe Flick, Ernst von Kardorff, Heiner Keupp, Lutz von Rosenstiel & Stephan Wolff (Eds.), Handbuch qualitative Sozialforschung. Grundlagen, Konzepte, Methoden und Anwendungen (pp.432-434). München: Psychologie Unions Verlag.

Gerhardt, Uta (1991). Typenbildung. In Uwe Flick, Ernst von Kardorff, Heiner Keupp, Lutz von Rosenstiel & Stephan Wolff (Eds.), Handbuch qualitative Sozialforschung. Grundlagen, Konzepte, Methoden und Anwendungen (pp.435-439). München: Psychologie Unions Verlag.

Gotsbachner, Emo (1999). Ausländerbilder als symbolische Ressource in Tatausgleichsverhandlungen. Jahrbuch für Rechts- und Kriminalsoziologie, 99, 189-239.

Hanak, Gerhard (1982). Alltagskriminalität und ihre Verarbeitung durch die Strafjustiz. Kriminalsoziologische Bibliografie, 9, 117-136.

Hanak, Gerhard (1987). Ethnographie der Konfliktverarbeitung. Kriminalsoziologische Bibliographie, 14, 9-34.

Hanak, Gerhard; Stehr, Johannes & Steinert, Heinz (1989). Ärgernisse und Lebenskatastrophen. Über den alltäglichen Umgang mit Kriminalität. Bielefeld: AJZ-Verlag.

Hellum, Anne & Fastvold, Marianne (1988). Money and work in marriage. WomenŽs perspective in family law (unveröffentl. Manuskript, Universität Oslo).

Hönisch, Bernhard & Pelikan Christa (1999). Die Wirkungsweisen strafrechtlicher Maßnahmen bei Gewaltstraftaten in Paarbeziehungen. Forschungsbericht des Instituts für Rechts- und Kriminalsoziologie, Wien.

Kelle, Udo (2001). Sociological explanation between micro and macro and the integration of qualitative and quantitative methods. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research [On-line Journal], 2(1), Art. 5. Available at: http://www.qualitative-research.net/fqs-texte/1-01/1-01kelle-e.htm

Luhmann, Niklas (1984). Soziale Systeme. Grundriss einer allgemeinen Theorie. Frankfurt/Main: Suhrkamp.

Luhmann, Niklas (1993). Das Recht der Gesellschaft. Frankfurt/Main: Suhrkamp.

Pelikan, Christa; Hanak, Gerhard; Pelikan, Johanna & Schandl, Heinz (1996). Familienmediation. Forschungsbericht des Instituts für Rechts- und Kriminalsoziologie, Wien.

Pelikan, Christa & Stangl, Wolfgang (1994). Das Strafrecht, die Konfliktregelung und die Macht der Frauen. Jahrbuch für Rechts- und Kriminalsoziologie, 94, 47-74.

Pelikan, Christa (1995). Partnerschafts- und Familienkonflikte im Außergerichtlichen Tatausgleich. Neue Praxis, 25, 151-166.

Pelikan, Christa (1999). Gutachten zur Entwicklung eines Projektes "Mediation bei Gewaltstraftaten in Paarbeziehungen" im Auftrag des Senatsamt für die Gleichstellung der Freien und Hansestadt Hamburg. Hamburg: Dokumentation des Senats für die Gleichstellung.

Pelikan, Christa (1999). Über Mediationsverfahren. Jahrbuch für Rechts- und Kriminalsoziologie, 99, 11-28.

Pelikan, Christa (2000). Victim-offender-mediation in domestic violence cases – A research report, http://www.restorativejustice.org (broken link, FQS, May 2003).

Wahl, Klaus; Tüllmann, Greta; Honig, Michael-Sebastian & Gravenhorst, Lerke (1980). Familien sind anders! Wie sie sich selbst sehen: Anstöße für eine neue Familienpolitik. Reinbek: Rowohlt.

Watzke, Ed (1997). Äquilibristischer Tanz zwischen Welten: Neue Methoden professioneller Konfliktmediation. Bonn-Bad Godesberg: Forum-Verlag.


Christa PELIKAN is a researcher at the "Institute for the Sociology of Law and Criminology" in Vienna. She has been doing research in the fields of family law and of criminal law. During the last years her main involvement was with Victim-Offender-Mediation and Restorative Justice and with research projects accompanying the pilot projects introducing these modes of criminal law reactions in Austria.


Christa Pelikan

Institut für Rechts- und Kriminalsoziologie
Postfach 1
A-1016 WIEN

Phone: 0043-1-526 15 16
Fax: 0043-1-526 15 16-10

E-mail: christa.pelikan@irks.at
URL: http://www.irks.at


Pelikan, Christa (2002). Victim-Offender-Mediation in Domestic Violence Cases—A Comparison of the Effects of Criminal Law Intervention: the Penal Process and Mediation. Doing Qualitative Research [55 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 3(1), Art. 16, http://nbn-resolving.de/urn:nbn:de:0114-fqs0201169.

Copyright (c) 2002 Christa Pelikan

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.