How can tool use improve the process and outcomes of human deliberation on paths of action?

In deliberation dialogues, participants collectively decide on which action is to be taken. The outcomes of such dialogues can be enormously impactful. Examples are rife across varied domains, such as policy, industrial strategy, academia, disaster response, and governance. Therefore, improvements in the quality of deliberative outcomes and the satisfaction rate of non-functional requirements of the deliberative process could deliver important, scalable improvements in the results of these critical decisions.

In this project, we approach the question "How can tool use improve the process and outcomes of human deliberation on paths of action?"

My proposed research consists in two broad phases. Firstly, an analysis of deliberation dialogues, to find features of real-world deliberations which tend to produce the best outcomes and processes. Secondly, the construction of a tool to specifically nurture and promote these features in users who are participants in deliberations.

The first phase will aim to answer the following questions:

- To what extent do existing formal models of deliberation capture actual human-human deliberations?
- Which modifications will allow the most promising models to more fully describe real deliberations?

These are open questions because the original design of most of the formal models was not intended to model human-human deliberations. However, by expanding them in this direction, the powerful insights of argumentation theory can be applied to critical deliberation processes.

Once deliberative features that promote good outcomes and processes have been found, the next phase of the proposed research is to design and build a tool that promotes them. Due to its reliance on the results of earlier sections of the research, the precise shape of this work is harder to map out at present. At this stage, it is prudent to remain neutral on what a tool to support deliberation might be like and how it will be realised, because these facts should be fully determined by the previous research, and thus a pre-commitment to some set of specifics at this stage is unlikely to select the best course of action. Our initial work has focussed on the promising method of argument mapping, and has involved building a framework to capture the mental operations involved in externalising ideas to an argument map.

By supporting the features of the best deliberation dialogues, our produced tool will aim to improve the outcomes of critical real-world deliberation.