Reasoning about causes and evidence law

Abstract

This is a study on causal reasoning in legal argumentation. It starts by identifying the core problems of determination and evaluation of cause-in-fact in legal cases, including the challenges of responsibility attribution in multi-party cases, and issues arising from hidden assumptions. The thesis introduces the Framework for Causal Argument (FCA) that uses artificial intelligence methods to identify, model, and evaluate causal arguments. The FCA aims to represent the arguments by expert witnesses and judges in a structured manner by modelling each of the reasoning steps in a logic-based language. The argument models are then evaluated using the argument scheme from-effect-to-cause and an updated list of critical questions. Moreover, with the developments in the science of causality, some new insights are directly integrated into the evaluation criteria. This thesis provides a general framework for causal argument analysis and presents its use through case studies of vaccine injury and asbestos exposure related lung cancer cases with complex evidential settings. It concludes with a discussion of the role of causality and evidence in the burden of proof debates, and also proposes a new approach to the application of the standard of proof exceptions in specific areas of torts.