Epistemological Reflections on the Evaluation of Restorative Justice Practices1

Article excerpt


This paper presents a reflection on the reasons which motivate the evaluation of restorative justice practices. In a social and institutional domain crowded with believes and symbols it debates what this need for understanding means. With the help of simple questions: who evaluates, for what reasons, what is evaluated and how, it proposes to researchers a small exercise in scientific hygiene useful both for improving their necessary distance and for understanding better the difficulties of evaluative reasoning within the field of restorative justice.

Key Words: Epistemology, Restorative Justice, Evaluation, Criminal Justice system, Methodology, Mediation


One might legitimately think that the best way of developing restorative justice is to provide proof of its effectiveness (although it would be necessary to define the kind of effectiveness one is talking about). But this idea is somewhat naïve. No researcher has ever been able to prove the effectiveness of prison in preventing recidivism, yet imprisonment has universally been adopted as a sentence. It is unlikely that it would suffice to prove the effectiveness of an experiment in order to gain the support of political and judicial decision-makers or of the public.

It is important to keep in mind that the function of the penal system is more symbolic than instrumental. Let us not forget Durkheim's words from a long time ago saying that punishment "does not, or only secondarily, serve the purpose of punishing the guilty or deterring their potential imitators: from either point of view its effectiveness is dubious and in any case poor. Its real function is to keep social cohesion intact by lending the collective social conscience energy." Punishment thus exercises its function of social defence in reviving collective feelings, the communion of spirits around these feelings, and it does so all the better as it is expiatory. So one can say without paradox that punishment is above all destined to put pressure on honest people (Durkheim, 1893).

It follows that it would be pointless to wear oneself out trying to provide proof of the instrumental effectiveness of mediation or of conferencing without taking into account the fact that criminal justice is replete with symbols whose political function is basically the production of images of a certain conception of social order. Even when the judicial apparatus puts the resolution of penal conflicts in the hands of community authorities, accepting a more democratic logic, it does not negate its institutional role.

This introduction is not meant as an indictment of evaluation as pointless. As a researcher, I am very attached to any reasoning concerned with bringing into daylight the institutional shade and unveiling what hides behind appearances. It is simply that we need to ask ourselves about the reasons that make us want to evaluate. One good reason is expressed in Recommendation R (99) 19 of the European Council of Ministers, which states that "the governments of member states should promote research and evaluation on mediation in penal matters". One can understand the concern of the European governments to develop a public policy which rests upon tangible elements. Yet in most of the countries which have widely developed mediation, Norway, France and others, this has been done without prior scientific investigation. Most mediation practices have been institutionalised (Faget, 2006) on the basis of a belief which has no empirical foundation. Consequently, the a posteriori concern with evaluation and, especially, the desire of militant supporters of mediation to subject it to the rigours of evaluation, should probably be seen as a need of recognition and a search for legitimacy.

Thus, what I want to argue here is that the motivation to evaluate is not as rational as one might believe. One should never forget that social practices - especially those which are the most institutionalised - have at the same time both an instrumental and a symbolic impact. …