AbstractDispute mediation is a practice in which third-neutrals (mediators) help conflicting parties to resolve a dispute in civil cases such as divorces, child custodies or in the workplace for example. Mediation is becoming a major dispute resolution process in most countries; for instance, calls to mediation services are increasing, and many countries make it mandatory to resort to mediation before going to court. This is because it presents many advantages over traditional litigation: it is quicker, cheaper and less stressful. This growth has led scholars to carry out various types of research with the aim of discovering the characteristics of discourse in mediation. As a result, theories based on systematic analyses of mediation dialogues are appearing, which offer novel insights and valuable data. As many research works have shown, argumentation deserves a particular attention in mediation since mediators must, at the same time, make sure that disputants effectively argue to reach an agreement, and preserve their neutral role.
The increasing visibility of mediation and the growing number of investigations on the topic offer new opportunities to provide mediation professionals with support tools which the process lacks when compared with other dispute resolution procedures such as traditional litigation. The research reported here therefore proposes to advance theoretical knowledge of the dialogical and argumentative activity in mediation in order to deliver practical applications to support mediation training.
To achieve this goal, this work relies on argumentation theory applied to discourse studies and computational models, namely Inference Anchoring Theory (IAT). This framework has already been successfully applied to other dialogical contexts (radio debates) in order to study argumentation. It has been shown that its main advantages are its flexibility regarding annotation schemes and its ability to elicit nonobvious argumentative structures which can then be easily modelled thanks to detailed analyses of dialogical dynamics (see e.g. (Budzynska et al., 2016)).
As a first step, a close analysis of transcripts of mediation sessions with IAT allows exploring the link between dialogical and argumentative dynamics, and revealing their patterns. Once modelled, these patterns are used to define rules which are then specified in the form of a dialogue game: the Mediation Dialogue Game (MDG). MDG rules are defined after in-depth empirical studies and statistical analyses. They reflect therefore mediation participants’ actual behaviours; they can also be regarded as normative rules since any mediation dialogue can be compared with MDG rules. The game can also be played in conversational support systems to enable trainee-mediators to practice their skills and techniques in a computational environment replicating mediation dialogues, in the same way as role-plays, the basis of mediation training. Though the aim of this work is to provide a tool for mediation training, the different contributions of this work also represent a first step towards the development of a tool which mediators could use during sessions.
To verify the quality and reliability of MDG, actual mediation dialogues are compared with the rules of the game, thus leading to a revision of some rules for a more accurate dialogue protocol. It is then shown that the revised version of the game, MDG’, fairly matches mediation interactions, and can be further developed as a fully-fledged tool for mediation training. The game represents therefore an empirically based normative tool which finds practical applications.
The evaluation process reveals some limitations of MDG’. Meta-discourse, in particular, plays a major role in mediation dialogues which the game fails to capture. The necessity for potential users to use meta-discursive moves in MDG’ in order to have a greater impact on the direction and content of the dialogues is hence highlighted, and a method for the analysis of the role and function of meta-discourse in mediation is proposed. This first-ever study of meta-discourse in mediation dialogues represents the foundation of a wider account of mediation discursive and argumentative characteristics.
As a conclusion, the research presented here stands as a novel approach of argumentative dialogues in mediation and explores the relationship between dialogical dynamics and meta-discourse. It relies on in-depth investigations of a corpus of mediation dialogues in order to explain the link between dialogical behaviours and argumentative dynamics. These theoretical findings are then used to develop a practical tool intended for mediation training. This work brings new findings in argumentation theory and discourse studies, advancing theoretical knowledge and creating an opportunity for the support of mediators’ training in a context of growing interest in alternative dispute resolution procedures.
|Date of Award||2017|
|Supervisor||Chris Reed (Supervisor) & Katarzyna Budzynska (Supervisor)|