Computer Supported Argument Visualisation: Modelling Wicked Problems

Abstract

This project studied an instrumental case to investigate the potential contribution of computer supported argument visualisation (CSAV) for the representation and analysis of public discourse in so-called wicked problems. All societies face complex problems where stakeholders hold heterogeneous viewpoints, agendas and voices, built on disparate and often contradictory logic. These require multi-criteria analysis and policy decisions that will affect many stakeholders. Such problems have been termed wicked problems by Rittel. Research suggests that there is a need for a means to facilitate and clearly represent discourse and argument in these complex settings. In the 1960s Rittel designed a system, ontology and approach to aid complex decision making in wicked problems: the ‘Issue Based Information System’, ‘IBIS ontology’ and, ‘the 2nd generation systems approach’. Yet even though Rittel’s notions and related ontology are well regarded in academia and practice, there is a dearth of research that seriously engages them on a theoretical basis. Hence, there is both a theoretical and a practical need to investigate this problem and to develop effective methods and tools to aid dealing with them. Argument visualisation is a technique and tool that has the potential to address many of the needs inherent in wicked problems. It comes from the emerging field of computer supported argument visualisation. Technological advances since the sixties have led to computers that can produce and process greater volumes of data. This is pertinent to consultative forums because of the potential for creating a volume of information exceeding that which Rittel’s original system could cope. Accordingly, this project asks the question: “What contribution can computer supported argument visualisation make to representation and analysis in wicked problems?” The primary argument visualisation software used in this project, Compendium, is a derivative of the IBIS ontology and grammar. The development of this software has occurred through expediency and convenience without adherence to rigorous theory development methods. This relatively ad hoc approach to development has served practice well, but it nonetheless represents a potential disconnect from its foundational ontology and notation. This project aimed to bridge this gap. The investigation used the draft South-East Queensland Regional Plan Consultation as an example of a wicked problem. The case has issues of global importance, such as atmospheric pollution, climate change, water management and transport. Many international jurisdictions share these issues. The Queensland Government is among the world leaders in consultative democracy. Brisbane, its major metropolitan area, has the world’s largest local government jurisdiction by land mass, and at times during this project was the second fastest growing city in the western world, implying a plethora of major infrastructure and social issues. These factors come together to present a challenging level of complexity for policy planning. To apply computer supported argument visualisation in such a complex and high volume forum, an appropriate design model is required. To date there has been no in-depth research on how acceptable the use of CSAV is for providing visualisation of discourses on evidence-based policymaking. Accordingly, the Project investigated this matter, which resulted in the development of a design model. The research methodology used was descriptive, and was based on an interpretive epistemology and a single in-depth, instrumental case study. The findings were elicited from respondent data using qualitative research and a constructivist grounded theory method. This facilitated theory building, where emerging concepts, propositions and theory were contrasted with extant literature. This thesis contributes in theoretical and practical ways. Firstly, a detailed generic model for representation and analysis of consultation discourse in wicked problems was developed and evaluated. This model provided a basis for evidentiary discourse analysis and policy development on any scale. Secondly, the notional ideas from Rittel were theoretically reified following a grounded theory method, and the project validated that all theoretical constructs of the wicked problem notion were required and applicable to a situation. Practical contributions included using an extension of the 2nd generation systems approach to represent and manage multifarious discourses in wicked problems, and explaining how CSAV and consultation mapping can be used to support and enhance Weick’s (1995) sensemaking theory. Recommendations were developed for the execution of consultative policy development programmes and the use of CSAV in public engagement, submission analysis and policy development. Generic theoretical requirements for software support in representing and analysing problems of arbitrary complexity were also specified. Commercial software developers have since adopted many of these specifications.